This has got to hurt
Just when the experts were willing to say Mazzaro could possibly be a number 3, along comes this from Gammons:
'Two scouts watching Oakland's Vin Mazzaro on Thursday agreed that they think he is "a young Chris Carpenter." One added, "I love Rick Porcello, but Mazzaro is the best young pitcher in the American League."'
**************************************************************************
Again, I am sure that there will be the reactive replies (if any) that focus on me and my continued focus on this, but the point is that Mazzaro was a huge miss for the prognosticatorrs (sp?) on this site and I think there is an opportunity to look at the story surrounding Mazzaro and learn a little bit.
For example, I think that because of the small amount of information that is available on these prospects, when the coaches of a league (guys who actually saw Mazzaro pitch all year) look at a 21 year old and say he was the best pitcher in the league, maybe next time this should be considered a little more in the evaluation process.
232 comments
|
4 recs |
Do you like this story?
Comments
yawn
This is starting to get a little old . . .but that’s fine.
Yes, he could possibly be a No. 3 starter. There are certainly those who think that he could be more or less than that, though. This should not be surprising to you, nor should it be offensive to you.
Yes, there is one scout who said that he thought that Mazzaro was better than Rick Porcello. One scout. In your last thread I said that I think you’d find something on the order of 8 out of 10 GMs would take Tommy Hanson over Mazzaro, and I’m guessing it’s probably in the 7-8 range for Porcello as well. That the minority opinion was the one that was published does not make the majority opinion irrelevant. Often the point of news blurbs is to express interesting and/or distinct information. Certainly the vast majority of fans, being more familiar with Porcello, would be inclined to think that Porcello is the better player . . .this news blurb, however, does not challenge the idea that Porcello is an outstanding young pitcher, only that you can find SOMEBODY (one guy, which actually underestimates my personal prediction) who thinks that Mazzaro is better at this point.
The community list had Mazzaro at No. 98, which was perfectly defensible and said a lot about his merits. The 98th best prospect in baseball is still a REALLY good prospect with a good chance of having a quality MLB career. I think you make the mistake of not understanding how fluid the idea of ranking prospects really is . . .it doesn’t take much for a guy to, say, jump 20 places in somebody’s mind, or to fall 40 places . . .but does that alone tell you anything about how good the guy is, or even necessarily what you think of him?
Ranking lists are subjective . . .for instance, in 2006 Matt Cain had just finished his age 20 season, in which he reached the majors and pitched 46 innings of 2.33 ERA ball, had struck out 176 guys in 145 PCL innings with a solid-for-the-league 4.39 ERA, had unbelievable stuff . . .and yet was “only” the No. 10 prospect for BA. Hell, Cain was only the 4th best pitching prospect in baseball that year, which actually was a DROP for him from 2005 in which, despite being the No. 13 prospect in baseball, he was the 3rd best pitching prospect. In other words, his ranking had a LOT to do with what BA thought of his peers, regardless of his ability. And beyond this point, there are many more than 100 good prospects in baseball.
There’s really not much to learn here, other than a reinforcement of the usual caveat that with prep draftees you often have to compare their development with their make-believe counterpart that went to college. Mazzaro would have been very comparable to many high draft picks out of college who, after underperforming their first couple of years, took off in their junior season . . .and if he had been a first rounder coming out of college with the kind of performance he’s had the last couple of years, he might actually be OVERRATED compared to what he’s likely to do. But in any case, Mazzaro was, regardless of how he was ranked, considered by virtually everyone to be a talented young pitcher with the ability to succeed in some capacity at the major league level.
Pretty much spot on
Maybe if it’s repeated enough it will finally sink in that having a prospect rated 98th who gets off to a great start in the majors and has the potential to be a very good pitcher is not a huge surprise and does not constitute a “miss” by the ranker(s).
The best lesson to take away from this is that the community prospect lists are a composite opinion of a bunch of random people with various biases and different levels of knowledge so you probably shouldn’t let your feelings get hurt so much if a guy you like is ranked lower than you want him to be.
I appreciate the thoughtful reply
However, I will respectfully disagree .
A 21 year old who was the AA pitcher of the year had both the age-performance one looks for in the top 10 prospects and the 92+ MPH fastball with wicked movement movement also rates very well.
While the exact location is debateable, even John agrees that Mazzaro was clearly missed.
Oh wow AA pitcher of the year!?!? How mistaken were we!?!? Your argument is stupid. His 2008 season was stellar, but his 07’ season was quite shitty. It would be absolutely stupid to argue that because of his 08’ season alone, he was a better prospect than the many prospects around the league who had the same, if not an even higher level of success for a longer time. You’re clearly an A’s homer trying to justify why you were so high on your own prospects. Get over yourself.
Semi-proud adoptive father of Scott Barnes.
are you sure
that this thread is about any prospect in particular, be they in the A’s organization or not?
Saying something is stupid does not make it so. One could easily have made the argument that plotting the year to year improvement from 2007 to 2009 on a graph would actually have projected this years performance quite closely.
I don’t propose that method, but it is a viable method.
Reading 101
he didn’t need to make your argument stupid by saying so
by RedSoxFaithful on Jun 21, 2009 11:35 AM EDT up reply actions
do you know what it means ...
… to reject the null hypothesis?
I dont
but I dont care. Your argument is stupid and you know it. And as for this…
OneAn A’s fan could easily have made the argument that plotting the year to year improvement from 2007 to 2009 on a graph would actually have projected this years performance quite closely.
There ya go. All fixed.
Semi-proud adoptive father of Scott Barnes.
There's plenty of A's fans here
And until this off-season (well, during last season, but this off-season as far as rankings go), I don’t think any of us considered Mazzaro more than the interesting-prospect-with-potential type.
Don’t lump Daaron in with the rest of us who post here. That’s not fair.
by thejd44 on Jun 22, 2009 2:32 AM EDT up reply actions 1 recs
You're right
my bad.
An A’s fanDaaron could easily have made the argument that plotting the year to year improvement from 2007 to 2009 on a graph would actually have projected this years performance quite closely.
Fixed again.
Semi-proud adoptive father of Scott Barnes.
Congratulations
You have a PhD in statistics and a kindergarten-level education in getting along with and influencing people.
why don't you answer the question I proposed?
It is not a trick question
Because you're the kid at the playground no one wants to play with
And you brought it all on yourself.
I know what a null hypothesis is...
what are you declaring is the null hypothesis in this case?
what is the point of all these?
If people don’t like Mazzaro and think he is going to flop, do you really think you can convince them otherwise? All you are doing is getting them to gravitate further away from you, not closer. People have a right to have their own thoughts on prospects… if some didn’t like him, that’s fine.
Adam Dunn: Proof that even sabermetrics doesn't have it right.
Can't you see?
Daaron is performing a public service with his insistence on getting to the root of how the prognosticatorrs missed so badly on Mazzaro. The key which unlocks the entire mystery of prospect evaluation is there within Mazzaro’s story. He will not rest until that key is found.
not the entire mystery
but there is the possibility of education.
Not everybody likes education.
Remember all the Bonifacio articles at the start of the year??
yeahhhhh
Remember: baseball guys... baseball...
way, way too early for some sort of ‘I told you so’. I hope Mazarro succeeds, but I also hope you face the music so publicly if he doesn’t. But you won’t.
Prospective parent of new pick, Zack Wheeler. Projectable Righty stolen from the braves. Of course, I stalk my son's myspace: http://www.myspace.com/zackwheelerbaseball
"Obviously I’m not doing things like going toe-to-toe with a ninja. Find me a ninja, for one."--Brian Wilson
Thanks
I can now safely say that you have zoomed into the lead for the biggest tool on this site and I’ll be sure to skip reading any and all of your posts in the future. You are the weakest link, goodbye.
I reject your reality and substitute my own.
by WayneCampbell08 on Jun 20, 2009 6:21 PM EDT reply actions
Look, I agree with you
But “You are the weakest link, goodbye”?
Seriously?
by RedSoxFaithful on Jun 20, 2009 6:58 PM EDT up reply actions 1 recs
LOL
+1
Wow that’s gangsta right there pulling out those retro quips.
by AthleticsReign on Jun 21, 2009 1:11 AM EDT up reply actions 1 recs
Jeez
Now I’m praying for Mazzaro to fail miserably, not because I give a shit about Mazzaro but because I just want this guy to shut up about him.
G G G E-flat_______ F F F D__________....
+1
Daaron, I’ll ask it again: please post your top 100 before next season so we can all see how perfect you are.
Jagoff.
And by the way
The teams Mazzaro has faced would be, in order of OPS:
White Sox (25th)
Orioles (13th)
Giants (28th)
Dodgers (12th)
Just because you’re such a tool, Daaron, I eagerly await Mazzaro’s starts against the Yankees (1st), Rays (2nd), Red Sox (3rd), Jays (5th), Rangers (6th) and Twins (7th). Is this the reaction you were hoping for?
can I respond with a question?
In any given part of the country, how often do you think I would be considered right when asked what the weather will be like tomorrow, and my reply is: the same as it is today?
Would you say I would be right a high percentage of the time?
Most of the time?
Some of the time?
Rarely?
I see no one wants to respond
and probably think it is a trick question. The only trick is that there will be a follow on question.
However, until then I will give an anecdote:
I lived in Orlando in the early 90s and the weather forcasters (predicters if you will) had very interesting methods of prediction. In the summer, if you have ever lived in Orlando you would know that it rains in the summer nearly every afternoon, the weather forecasters would prodict that there would be a 50% chance of rain in the afternoon.
The qualification of the prediction was that there was a 100% chance of rain in the Orlando area in the afternoon, however there was only a 50% chance of rain that it would be falling on you if you were walking outside.
Things that make you go hmmmm….
Funny
This whole thing is funny really. The responses and the way you dont respond to the teasing. Way to go man? I may be tired or something but care to explain how your weather analogy relates to Vin Mazzaro? lol Or does it?
intellectual integrity
Tell me something, if you study weather for a living and you tell me that the best you can say is that there is a 100% chance of rain in the Orlando area and a 50% chance of rain that it will actually fall on you if you are outdoors, what kind of value added is that?
No regression analysis on pockets of areas that are much more prone to rain or tornados, no regression analysis on the amount of rain that some areas might get, just 50% chance of rain.
It seems to me that there are people in this community that spend hundreds if not thousands of hours scouring over every possibe aspect of information that they can use to create an opinion and yet they spend zero time attempting any method of determining if they are wasting their time or not.
I would think that a little proactivity on the part of the community would provide a plethora of information that could actually determine ways to be much useful and interesting.
Doesn't that assume that
all questions – or at least these two – can even possibly be answered correctly all of the time?
http://rswanzey.blogspot.com
the goal
would be to answer the question statistically with a certain degree of confidence.
Not perfect I know, but light years ahead of the system currently in place.
I agree
At the same time of the voting of the CP list we are speaking of i proposed a list that would start from the BOTTOM up- I thought it would be interesting and a different way of doing things but it was kinda shot down on here.
i think there are MANY ways of doing things. For one, a way of keeping stats on US would be kinda great. People are always saying when they arte wrong about something that it was something out of thier control- the numbering isnt important- blah, blah, blah- but NEVER…
I just didnt realize how good somebody was, and HERE are the reasons why.
Another thought question
Lets say for example that Mazzaro had a younger brother, call him Van, that is 1 year younger than Mazzaro. His stuff is pretty equal to Vince’s and he was drafted by the A’s 1 year after VInce and his first 2 years he was equally as bad as Vince.
However, for whatever reason, Van puts it all together and becomes the AA pitcher of the year at 21.
Do you think Van would be rated higher number 98 on this communities list and in the top 100 of the BA list and get a higher grade than C+ and be higher rated than 3 stars on the BP ratings?
There is nothing I dislike as much as thought experiments
Stop thinking that the social sciences is a science. Baseball is not a science. Prospects are not a science. These are people, not atoms. Considering physicists can’t even tell you precisely where an atom is, then of course none of us know exactly what a prospect will do. We’re making educated guesses – that is all – nothing more, nothing less.
Statistics are useful for understanding the past, but are of very little use to understanding the future. What is useful for understanding the future? Not very much. Paul Kennedy tried when he decided to stop being an historian and become a social scientist – you know what he discovered? That Japan would have a bigger economy than the US by the mid-1990s. How’s that working out for him?
You know the problem with this whole argument – the phrase thought experiment. You can’t do an experiment with baseball. Do you know why you can’t do an experiment? Because for an experiment to be viable, you need to try it many times. Startlingly, there is no Van Mazarro, and certainly not enough of them to be statistically relevant.
So what have we learned? You like Vin Mazzaro and he pitched 26 good IP. Basically, you made a good educated guess. If you think this is the start of a large scientific theory, you’re wrong. You know why? Because social science is bullshit.
TheSouthWing.com - A Magazine of essays, prose and poems
by OldProspects on Jun 22, 2009 5:52 PM EDT up reply actions 4 recs
I will attempt an answer and hope another follows
I will suggest that if along comes Van, and assuming Vin continues to pitch really well (big if), then Van would be a top 5 prospect. He would be an ‘A’ prospect from John, a 5 star prospect from PB, and in the top 5 in all the lists.
Why the difference?
Because the current system is basically open loop. No method of refining the system with any degree of precision.
So then if Matt Wieters flops in the bigs...
…do we start doubting highly-touted college catchers who perform well through the minors?
Even if Mazzaro has a Hall of Fame career, fact of the matter is that he would be one prospect who’s performance went from pretty bad one year to very impressive the next and managed to continue to improve. Plenty of guys have a bad season and bounce back for a good one. Some of them never have another good season, some of them probably have a handful more good ones after, and maybe some have even had good MLB careers.
Fact of the matter is that if Van ever comes along, yes, he will receive comps to Vin, and yes, it will probably improve his value in the eyes of people who follow baseball prospects. But no, he would not be a top-5 prospect, or probably even a top-50 prospect.
Jesus, I almost want Mazzaro to fail now.
And I’m an A’s fan.
This is really getting ridiculous. I wonder why he isn’t making a fanpost after every outing that Andrew Bailey has.
Solace: Law says he's a fourth OFer
PaulThomas: I think Keith Law is only a fourth analyst
do you really think this thread is about Mazzaro?
knowledge is power
Right, it's actually about you being a worthless troll
by thejd44 on Jun 21, 2009 1:22 PM EDT up reply actions
This crap is deleteable, I think...
Solace: Law says he's a fourth OFer
PaulThomas: I think Keith Law is only a fourth analyst
The thread
This is nonsense
Solace: Law says he's a fourth OFer
PaulThomas: I think Keith Law is only a fourth analyst
Mazzaro is simply a metaphor
for how screwed up the current performance estimation process is.
My observation has been that what little valid data was on Mazzaro was ignored and most people merely regurgitated (sp?) inaccurate data to decide what kind of prospect he really is. As late as 2 weeks ago then was some concession that he had number 3 potential when there was data (not just statistical data) to suggest that the ceiling was much higher.
The question remains: is there a way to bridge the gap in the information age?
A couple of decades ago Bill James started an new age by making decision making information available based on statistical analysis of just the numbers. I would maintain that the next generation of decision making information will include the carefully surveyed observations of people who have to coach against a player and those who have to play against players.
The information provided by scouts is still important, but I would think that the information from carefully crafted surveys of, for example (again Mazzaro is simply a metaphor), the players who played against the guy in the California league and the Texas league and the coaches who coached against him is pretty indicative of the kind of player he is.
I would think that a site like this one could be a trend starter in creating a new age of decision making information for predictive estimation of prospects.
JMO.
Perhaps you could have taken this tack from the beginning
instead of making everything so personal and insulting everyone here.
Aside from that, perhaps you’ll learn one other lesson: There comes a time when you’ve beaten a dead horse so thoroughly that your voice grates on everyone around you and you turn off those you seek to convince. You have long passed that point, and I would suggest that any further posting on your part about this subject will only further alienate people. I mean, are you a stalker in real life? This monomaniacal insistence on having people see things your way, to the point that you drive them away … it’s not productive, not for you, not for them, not for the site.
Letting go … you really should try it.
by Flynn Blake on Jun 20, 2009 10:23 PM EDT up reply actions
so....
Just how heavily do you weigh the opinions of coaches? I mean, Adam Piatt was Texas League MVP or something, wasn’t he? Also, some coach in the Cape Cod League compared Grant Green to Longoria. Should we take that as gospel, even though no one else is making that comparison?
Sure, people who try to predict the future aren’t always right. That’s not exactly earth-shattering news.
good instincts
It is a good question and in fact the study of outliers is often the key to improving any predictive estimation system.
However, I would raise a couple concerns:
1. Steroid era examples will prove difficult in many respects. First, the longer back the example, the less accurate the information. Second, who knows if he was juicing or not.
2. This is actually the negative of the Mazzaro case in the Piatt did not live up to expectations. This is much less of an outlier than a guy who was completely under predicted.
Piatt
I don’t think I understand you. You don’t want to include Piatt because he wasn’t a successful major leaguer. If you want to improve the “predictive estimation system” for baseball players, you would have to have evaluated Piatt before he accumulated any significant amount of data in the majors, just as you’re doing with Mazzaro. If, on the other hand, you are going to look at players from the past to predict the success of future players, you can’t then selectively eliminate players who don’t prove your assumption. The end result will be a predictive system with absolutely no value. Essentially, what you’ll be saying is, “This set of players will be successful, except for those that will fail.”
I understand that you think there’s a lesson to be learned about player evaluation from Mazzaro’s case that can be applied to other players. I disagree. First of all, we have to accept the premise that everyone (except you, apparently) was dead wrong about Mazzaro specifically. That’s just not the case. To make the top 100 list on this site is definitely not a sign of disrespect. Even those who thought he was being overrated were not necessarily wrong. After all, he had one good season after a couple of mediocre ones. A lot of people wanted to see him sustain that success for a longer period of time. That’s perfectly reasonable. There are plenty of players who manage to be successful at some point or another, but fail to remain that successful. Second of all, you are working with a sample size of one. Third, is it really worth it? I mean, if we start concentrating our efforts on players with mediocre track records with potentially fluky performances, what will we gain? The occasional potential mid-rotation starter in a sea of busts? That’s not to say that a mid-rotation starter isn’t valuable. However, you’ll probably find more of them by concentrating on players with a longer history of success. Now granted, you probably think Mazzaro is more than a mid-rotation starter, and the value you place on him increases accordingly. But either way, the burden is on you to show us a group of players with a similar history to Mazzaro, that is so large and has a record of so much major league success as to be worthy of so much attention.
3 points
1 If all you see is a sea of prospects and cant attempt to differentiate them you are kind of excusing yourself from the conversation.
2 If a guy has his best season, at his highest level, at the age of 22. The previous seasons dont matter. It is most likely a new level of ability. It happens EVERY year with good young ballplayers. Its actually very basic stuff man. If you learned that you’d be farther ahead.
3. Mazzaro didnt need to show that he could keep this up at a higher level first. YOU were supposed to look at the info before the season and see that it would happen. That is what some of us are doing- predicting
Its not a strict merit system
good try, but...
1. I think it’s pretty clear that the “sea of” comment referred specifically to players like Mazzaro, who have a much larger record of mediocre performance than of good performance. I lump them together because the OP keeps insisting that he’s not talking about Mazzaro, but about a group of players like Mazzaro. I’m not the one who isn’t attempting to differentiate, the OP is. Try to keep up.
2. The previous seasons don’t matter? You’ll throw out a guy’s entire history based on ONE season? Wow. Now granted, if he makes that improvement at a young age and a higher level, it’s possible that he has improved as a player. But to conclude that such a player has suddenly become one of the best prospects in the game is extreme and unwarranted. I looked up other POY winners in the Texas League going back to ‘97 (I would have gone back further, but I got bored). I’ll grant you that only 2 were as young or younger than Mazzaro at the time. So how are their careers going? I’ll let you look them up. Their names are Bud Smith and Travis Blackley. Granted, SSS issue here. But my sample is twice as big as the OP’s. Yes, every year there is a ballplayer or two who breaks out and becomes better than most people thought. Players like that are the exception, not the rule. They don’t serve to invalidate all existing methods of evaluation. No system predicting human behavior is 100% accurate.
3. I should have predicted Mazzaro’s ‘08 based on his ’07? Are you kidding? Or do you mean that I should have predicted ’09 based on ’08? That’s slightly more reasonable. But you are still wrong. You really don’t think Mazzaro had to prove that ‘08 wasn’t a fluke? Have you never in your life seen a person accomplish something they have never been able to repeat? Have you never seen a baseball player have a season far better than any other in his career? They even gave it a name. It happens that often. Yes, some of us are predicting. Others are a little too excitable.
Look, I like Mazzaro. But I hate that someone is so in love with him, and so obnoxious about it, that I end up looking for reasons why Mazzaro will “only” be a #3 starter. And I hate that you come in here with a condescending tone when your post reads like you didn’t understand 80% of what I said.
excellent try
I think casejud understood you just fine, but it does not look like you understood his argument.
Further, until you understand that this thread is not about Mazzaro, you will never understand my argument.
Finally, are you actually trying to say that because you have twice the sample size that I have, you are able to reject the null hypothesis and I am not? I doubt it.
The good news is that your instincts to get more data are great. The bad news is that the signal to noise ratio in your data is pretty low.
what?
Every time someone disagrees with you, you fall back to your “it’s not about Mazzaro” argument. i would say it was a good try, but it really wasn’t. I never once said that it was all about Mazzaro. The problem is, Mazzaro is the only player you discuss. If you have a larger group in mind, then you need to share that. Bring us a sizable list of similar players that had sudden improvements in performance at a young ARL, and let’s see where their careers went after that. But to simply say that “there is an opportunity to look at the story surrounding Mazzaro and learn a little bit,” does not in any way indicate that you have any more data beyond the one player you say this isn’t about. Do you see what I mean? I just re-read that. I don’t mean to be sarcastic or rude. I mean, have I explained my point in a way that makes sense to people besides me?
I think I understood casejud quite well. There wasn’t much to understand and it wasn’t very complex. On the other hand, he clearly didn’t get my points. And I am willing to admit that it I may not have done a good job making those points, although I’ve read my own post a number of times and I don’t know how to make my points any clearer. Maybe I will make one more attempt:
If a player has a couple of mediocre years, followed by one good one at a relatively young age, it is possible that he has made strides in his performance. It is also possible that the one good year was a fluke. Assuming that such a player has gone from a virtual non-prospect to one of the best in the game is premature without more data. Drawing conclusions about other players based on the example of any one player is unwarranted.
Yes, yes, I know, it’s not about Mazzaro. So make it about a certain set of players. Give us names. Without that, you have no point.
You have been perfectly clear
One question:
If I provided a set of 10 players, do you still think I have enough data and sufficient methodology to reject the null hypothesis?
short answer? no
Long(er) answer:
First of all, we need to define this set of players. Second, we have to define the players they are a subset of. What I mean is, are these the only 10 players that meet our criteria out of all players in history? In that case, these 10 players don’tmean anything. Are they 10 players who were ranked (relatively) low on most prospect charts last year but are now performing well? In that case, they probably still don’t mean much, but I’ll concede taht they may be worth further study.
Let me pose a thought question
If my company, call it company A, can produce a widgit with 100 hours of labor on average and company B can produce the exact same (same quality, same cosyt of overhead, same cost of labor and everything else) widget with 90 hours of labor (all comparisons being apples to apples), do you believe that you have sufficient information to deduce that company B is more productive than company A?
sorry, that was a somewhat trick question
The answer is no because I did not give what the distribution looked like.
For example, I might have said that I have a productivity of 100 hours per widget, but what if the data actually was half of them took 1 hour and the other half took 199 hours.
There is so much noise in that data that very few statistically significant observations could be made with it. This is the fundamental concept behind my replaying the term “rejecting the null hypothesis”. There is a statistics test that would most certainly fail and it is called the student’s t-test, also called the mean difference test. This would say that no meaningful significant statistical conclusion could be drawn from the data.
How this applies in the case of predicting prospect performance is that I have yet to find a prediction system that has any methodology in place to know if the results are statistically significant.
The bottom line is that a semi-trained monkey could probably do as well.
Its interesting
I dont quite grasp it but Id be happy to discuss and help if you need it and are planning anything intersting here.
I think Im better than average in evaluating young players and it is tricky. Its easy to get led astry for this reason or that. Constantly looking at the statistical info but not very in depth.
Just lookig at a lot of “common sense” and trying to see reality through numbers, age, where a guy is playing, etc and get to where you are actually “seeing” things more clearly than before.
Id like to get beyond the level of a trained monkey. Thats a worthy goal :-)
I appreciate the offer
Right now I am trying to find out what it is I don’t know.
All the best.
You seem to really enjoy spinning your wheels about statistics and the null hypothesis without really proving any knowledge of what these things mean.
A t-test (or whatever fancy name you want to call it) is a horribly biased way of comparing the means of two normally distributed data sets with respect to the variance. All it will tell you is whether there is a non-zero difference between groups that fit those assumptions. Trying to sound smart by inserting little comments here and there about analysis that is completely irrelevant to evaluating prospects is a pretty blowhard move.
by assorted small objects on Jun 23, 2009 1:41 AM EDT up reply actions
are you saying
that there was enough statistically significant information in the thought question to say anything about the data?
No, sorry, I should have been more clear (and less obnoxious).
What’s the point of asking the thought question? How does this apply to evaluating prospects? It seems like you brought up a somewhat irrelevant point here. What I’m really asking is — how do parametric statistics apply here?
And, yes, as you pointed out, you can’t evaluate statistical significance without accounting for variability within the populations that you’re comparing.
by assorted small objects on Jun 23, 2009 2:30 PM EDT up reply actions
the point
The point has been all along is that many predictions are made and many long hours spent pouring over data that has no statistical significance.
There are many methods of making non-numerical data statistically significant.
Unfortunately, I won’t be continuing in this thread as it has devolved in honest people calling other honest people liars.
Maybe we can discuss this at an other time.
Fair enough.
Perhaps our definitions of “statistically significant” are different — when you talk about statistical significance (such as with a t-test), you’re saying that the difference between the means for two groups is significant at whatever alpha level you’re using.
How can one set of data (or population) on its own be considered statistically significant? That’s what I’m hung up on.
by assorted small objects on Jun 24, 2009 12:36 AM EDT up reply actions
Mazzrro was ignored?
He was ranked in the top 100 by the community. How many other prospect evaluation systems can say the same? A top 100 prospect is high praise so I don’t really see how we missed on him.
"When a true genius appears in the world, you may know him by this sign, that the dunces are all in confederacy against him."
-Jonathan Swift
by King Billy Royal on Jun 21, 2009 12:14 AM EDT up reply actions
Well . . .
Don’t get me wrong, I’m intrigued. But what you’re proposing is actually a revisiting of “traditional” scouting methods. Statistics, for better or for worse, exist in a vaccum – they are nothing if not objective (although obviously how one manipulates statistical data results in SOME sort of bias, this is inevitable). Scouting, on the other hand, is totally subjective . . .it’s just the opinions of people who are trained to give their opinions on a given player. Intuitively one would be inclined to think that these trained individuals are going to end up giving more accurate (generally speaking) predictions on a player’s future success than non-trained individuals such as coaches, opposing players, etc. but 1) we don’t necessarily know that the determinants of future success that a given scout (or any scout for that matter) looks for in a player are more accurate than what opposing players / coaches are looking for and 2) nobody has done a respectable job of isolating the data, i.e. separating player / coach opinion from scouting opinion.
Another issue is, of course, how a player changes from year to year. In the case of Mazzaro, if we went by coach / player opinion for his first two full minor league seasons, they would have likely described a mildly interesting but somewhat fringy pitching prospect – he was not impressive, nothing wrong with saying that. He’s been a totally different pitcher from 2008 onwards, but that’s something that hasn’t just been backed up by the opinions of his peers, it’s something that’s coincided with both a general improvement in the quality of his raw stuff and a massive improvement in his statistical results. So then we have to ask: just what was it that Texas League participants were impressed by with Mazzaro? That’s the big question, and it’s one that you haven’t asked. The guy led the league in ERA . . .obviously players and coaches would have to be morons to refuse to acknowledge the fact that they couldn’t score off him. But how about what they actually thought of his stuff? How did his fastball rate to his opponents? His slider? In other words, what’s more important: the means of Mazzaro’s stuff or the ends of his statistical record?
The most interesting thing about Mazzaro’s case, IMO, is the massive difference in his league ranking (#4) and his overall ranking (not in BA’s top 100). Now, BA’s league reports are handled a bit differently from the top 100 . . .the league reports have a heavy dose of input from league-specific sources (league scouts, coaches, players), similar to what you’re proposing. The top 100 is developed exclusively by the BA guys themselves, and while they consult those same sources that they used in the league rankings they do not weigh them nearly as heavily. Typically there is not a significant degree of discord . . .Mazzaro is just about the biggest anomaly I can think of under their system. Which ranking should we trust more? Should we trust neither and rely solely on the objective data that is available (not just the simple statistics, but fastball velocity, movement of pitches, swing-and-miss data on pitches, things that in some cases we just don’t have access to currently)? None of these options would seem to produce a totally satisfactory answer.
Basically, subjective opinion (read: scouting) is as equally capable of lying about the chances of future success for a prospect as objective opinion (read: statistics) is; I have seen nothing to convince me that we have refined either to the extent that I should consider it to be of greater importance. In Mazzaro’s case, we have a pitcher who creates cognitive dissonance no matter which way you approach it. His early statistical record (and 30 throwaway innings in the 2008 PCL) is flawed, and even his component statistics when he’s succeeded are good but by no means exceptional. His scouting reports grade him out as a respectable but not elite prospect. The only place he’s achieved notice was in the 2008 Texas League and – I don’t think this can be under-emphasized – even that ranking was in hindsight. People around the league had good things to say about him, but it wasn’t like they were calling their friends up and saying “you gotta see this guy!” – Mazzaro just didn’t produce that kind of reaction in the present sense, and even now in the major leagues we are seeing the same thing – scouts saying that a guy with a 4.9 K/9 has pitched like Chris Carpenter part 2. Now that, my friend, does not make sense.
Bingo - you are on the road to 'getting it'
James, for better of for worse, basically addapted Demming and Shannon statistical analysis to baseball, IMO. However, there is a drawback to this methodology: it minimizes the story behind the numbers.
There are other statistical methods that could provide the story behind the numbers. The medical industry performs statistics on non-numerical data all the time (i.e. on a scale of 1 to 10, how do your feel when you take a certain medication?
Here I think is the place where great strides could be made in the predictive estimation of prospects. For example, how much work would it really take to provide a questionaire that also went with the vote for who was the best pitcher in the league? Maybe the same questionaire could be given to players.
The point being that a site like this one could be used for collating a tremendous amount of information that over time could be analyzed and refined for providing the objective story behind the numbers.
Obviously, I am not even a big enough fan to have all the answers, I am just providing some food for thought on how to improve the predictive estimation methodology.
one more thing
In the world of predictive cost estimation, there are devices called ‘CERs’ (cost estimation relationships). The simplistic example of a CER is a builder telling you how much a building will cost based on it square feet.
An example of software that applies CERs in software cost estimation is called ‘cocomo’. When making complex predictive cost estimations, there are often CERs that are not well numerically defined (requirements volatility for example).
However, they have applied statistical techniques in obtaining non-numerical data and can often answer these kinds of questions.
O HAI GUYS
DUSTY RYAN went 1-2 with a double in his season debut last night and you guys didn’t even rank him on the top 100 list on the site!! WTF??? I knew he was a future hall-of-famer because his Mom said he’s very good at baseball and I even heard he volunteered at a soup kitchen last winter. See, there was data beyond numbers to show his great upside, and only I could see it.
Sincerely,
Bill Bavasi
http://rswanzey.blogspot.com
I am sure ...
that this enough data to reject the null hypothesis.
STOPP
Now, go and create a
Vin Mazzaro + Daaron = ))<>((
blog. I checked, and the domain:
http://vinmazzaro+daaron=))<>((.com
is open.
If you are unsure about the symbol, google [urbandictionary] is your friend.
Adoptive parent of Kyle Nicholson
Wow that's disturbing.
They say nothing is impossible, but I do nothing all the time!
by muffinpryde on Jun 22, 2009 12:52 AM EDT up reply actions
I strongly agree
so it fits perfectly, as these posts are getting strongly disturbing/obnoxious
Adoptive parent of Kyle Nicholson
Gammons The Prognosticator
Peter Gammons is a fine source and knows a lot, but go back and check how many times he lavished praise on George Lombard. And I found this item from 2000 after a little Googling:
- Corey Patterson reeks of star, and has the modesty, intelligence and work habits to build on it. Cubs fans, get ready.
Gammons wasn't the only one...
Patterson was ranked extremely high by Baseball America for 2 or 3 straight years. I would bet that Sickels had him high as well. People need to remember that nobody bats 1.000 when it comes to prospects.
"When a true genius appears in the world, you may know him by this sign, that the dunces are all in confederacy against him."
-Jonathan Swift
by King Billy Royal on Jun 21, 2009 12:39 AM EDT up reply actions
You're welcome
Gammons wasn’t wrong for saying what he did. Patterson had awesome tools, a very nice glove in CF, and the guy really did work hard. The plate discipline issues just swallowed him up, a pretty extreme case of how rushing a prospect who couldn’t afford to be rushed proved to be the undoing of a can’t miss guy.
Really, this one kills me every time. Delmon Young obviously comes to mind as the present-day Patterson, but Patterson just had so much extra room for error with his plus defense in a premium position.
the study of Patterson like outliers
while probably necessary at some point, is also somewhat morbid and depressing and I would rather focus on the more interesting cases of the ones who far exceeded expectations.
For your argument
the Patterson study would be important now. The fact that it is depressing, as you say, has nothing to do with the present discussion, that being “providing some food for thought on how to improve the predictive estimation methodology.”
The problem many are having with your post is that you aren’t presenting anything new, but simply saying I told you so.
There must be a medium between what coaches/scouts say and what statistics present. There is a reason why Patterson was rated so high and Mazzaro wasn’t. Why do you believe that was and how has their success diminished or improved their stock?
Thanks for the thoughtful post, but
I do not believe that I have said ‘I told you so’ so much as ridiculed the status of those who believe themselves experts. A subtle difference I understand, but it is a significant one as I would not proclaim to have found a formula of omniscience in the area.
Further, I don’t have the statistics, but I think anecdotal evidence suggests that there are far more guys who completely fail to live up to expectations than those who exceed expectations by a far degree. This is the reason why I would argue that the Patterson case is much less of an outlier than Mazzaro.
While I am not sure that there is anything new under the sun, I would think that providing a place where the discussion could be about ways to obtain statistically significant non-numerical data is somewhat new.
no illusion of being loved and accepted
My methods were chosen very carefully for my own purposes.
I believe that combining statistically significant non-numerical data with the numerical data (ERA+ etc) is actually the the future of prospect prediction and has even taken place to a minor degree.
Whether or not this community contributes to this advancement remains to be seen.
Hey Gammons...
we’re still waiting for Bobby Crosby to win that MVP award you predicted two years in a row.
I think he only did that for
1 year. But still a LOL. I guess someone could argue he would be a much different player without all those injuries adding up, but the result is still the shell of a major leaguer we see before us now.
by AthleticsReign on Jun 21, 2009 1:18 AM EDT up reply actions
He's pitched 4 games
Lets not get carried away. He’s pitched 25 innings. It’s not like he’s pulling a Dontrelle Willis on us, and look at how that one turned out. Way way way way waaaaaaaay too early for the I told you so, I’m so wonderful post.
Reading 101
VIN MAZZARO IS NOT PULLING A DONTRELLE WILLIS HE IS PULLING A BOB GIBSON
by RedSoxFaithful on Jun 21, 2009 11:33 AM EDT up reply actions
don't sell him short
Vin Mazzaro makes Bob Gibson look like Sidney Ponson.
http://rswanzey.blogspot.com
at this point
this fanpost is just beating a dead horse
This
The 2009 Texas Rangers offense: sigh...
by Kinslerhomer on Jun 22, 2009 2:24 PM EDT up reply actions
lol
This is like the 3rd one of these. I don’t usually follow this site too closely but this is pretty funny.
The artist formerly known as Set-up man
Interesting stuff
Of course, i love controvery so, whatever
A few points….
1 if I make a list of top prospets and I rank somebody 34th and another player 98th. I am saying that I think the player who is 34th is BETTER than the player ranked 98th- that he will have a better career. For someone to say “Hey, there isnt any difference in ranking somebody that high or low” doesnt make any sense to me and is a total difference in how we percieve the rankings and the meaning of them.
2 This fella isnt exactly saying that Mazzaro should have been ranked much higher BECAUSE of his 4 fine starts in the big leagues but that he ALREADY should have.
3. One of the biggest criticisms of Mazzaro I heard over and over again was of his results in 2006-2007. Just a basic thing in evaluating young players is that if they all of a sudden get better at a HIGHER LEV EL combined with being YOUNG – you need both of these elements for it to apply- usaully means he has established a new level of ability. It is much more logical of a conclusion than a ‘fluke" Before this man’s posts I have heard Mazzaro’s AA season described as such many times here.
4. How many people on here who are irate and argue that Mazzaro’s ranking 98th on the CP list was not significant would have felt the same if Travis Snider or Colby Rasmus or Cameron Maybin or whoever else was ranked 97th instead? Not many would be arguing that it is no different than being ranked 7th.
This says nothing about Mazzaro specifically. Just some points I wanted to make.The best thing you can do Daaron, is not rely on anybody here to help you gain a concensus on a prospect or learn anything from that often but, for us to all be in a fantasy league together and trounce thier ass! That would be fun.
Point 2 and 3
There are a couple of flaws with some of your points, flaws that cannot be overlooked as they are pretty important to your central argument:
2. The fact that this “fella” is saying that Mazarro SHOULD HAVE ALREADY been ranked higher rather than saying that BECAUSE OF THIS he should be ranked higher is totally irrelevant. The point is, this guy is basing his opinion of the communities inaccuracy off of a short sample size. Whether this early-season performance is the cause of his move up or merely the proof that he should have already been moved up is totally unrelated to his central argument: That Mazzaro’s couple of starts, in which he has been pretty good, albeit against lesser quality offenses, are some sort of proof for anything. This guy’s ranking on the prospect will either be justified or disqualified off of more than a month or two of performance. For all we know, this performance could have been an outlier and he’ll turn into the next Haydenn Penn (I always use the Orioles examples)
3. Yes. A prospect’s improvement is one of the most important factors in evaluating his future. That does not mean it is the ONLY factor. Just because he improved does not mean he should be higher on the list. Doing better one year than the last should not be the basis for a giant leap up the prospect rankings. Adding two months of performance against low quality offenses to that does not make his improvement a true thing, or even that important. A slow, weak, badly fielding center fielder could improve to just slow and weak and average fielding the next year, but that does not mean he is much better than he was before; there is still the matter of his talent as a whole.
by maneatingbaby on Jun 21, 2009 7:34 PM EDT up reply actions
you, my friend, are a star
Very good points and I obviously could have addressed the particulars of Mazzaro’s ranking, however it would have only fed the bears. I try to observe the ‘do not feed the bears’ signs.
Also, my interest is in investigation the possibility of exploring a different system of evaluation. I don’t have the answers, but believe that there are plenty of others who have the intellectual curiosity to figure out a statistical evaluation process based on non-numerical data.
I interviewed for an executive position not long ago and during the interview I asked the very successful interviewer what was the secret to his success and his reply was ‘interllectual curiosity’. I thought it was an interesting response.
My methods in this thread, as in others, were chosen to identify the signal from the noise.
Thank you
I may not understand exactly what you are saying but I do take prospect evaluation seriously beause, it interests me, I play in a fantasy Diamond Mind Baseball league, and im competitive. The only way to get better at it, or anything, and win- is to LOOK AT YOURSELF- what you are doing- and get better.
A couple things about Mazzaro, completely unrelated to his stuff- scouting- any of that- that didnt make sense in evaluating him in the off-season AND were really not related to examining the statistics in any great detail were…
1) The fact that a guy goes to a higher, much tougher level in 2008 and raised his performance to a great degree is a MUCH stronger indication that he is developing than it is a FLUKE. Yet, WAY more people here have used that against him in evaluating him. That doesnt make sense.
2) Guys who have pitched that well, for that many innings, at that age, in the Texas League have a very good track record.
These are points just related to understanding the CONTEXT and MEANING of the basic numbers. Of course I saw him pitch just the other day and his STUFF and command was very impressive as well.
I guess that is what has dissapointed me about this site in the past. You think you will get this broad source of info on a player- many different views and NOBODY told me how good he was. How OBVIOUS it was that he was good. Thats the appaling part. And nobody cares- just abusive and defensive and personal.
These things are good if you in a league with people like this- you can exploit it. as far as learning together- no help.
common sense
is not very common …
You have been blessed with common sense.
unfortunately
you do not select sides very well.
Can’t you see I am outnumbered 30+ to 1?
Ah yes
But one of the great pleasures in life is knowing something everybody else doesnt. Lonely maybe but, fun nonetheless.
I would not say it so
in that I cannot say I exactly know something that everyone else doesn’t.
I would say that I am interested in learning something that many refuse to learn.
If I had the answers, I doubt I would be here.
True
I didnt mean it to sound so arrogant just that even in posing the QUESTIONS implies an OPENESS to knowledge that is somewhat rare, and there is integrity in that. Which isnt always going to have a consensus behind it. It can be a struggle to convince evn one person of something sometimes, even if it seems logical, simple, and innocent
People are open to learning
They are not so open to being insulted, denigrated, trolled and just generally treated as inferior. Can you not see that your attitude is responsible for the reaction you inspire here? Are you so blind?
He's a moron
And has no right to talk down to anybody else here.
Someone posed a challenge to him a while back, and I’d like to see him take it. How about he makes a top 100 list for next year so we can taunt him and mock him about how he didn’t rank so and so as high as he should’ve.
This is complete and utter bullshit. This type of stupidity and arrogance and sheer idiocy should be punishable by law.
Solace: Law says he's a fourth OFer
PaulThomas: I think Keith Law is only a fourth analyst
NOBODY
hey, he made the top 100, and was discussed and voted on higher than that.
When you say nobody told you how good he is, well, how good is he? Has that been decided? Does it ever get decided? People still debate where Lefty Grove ranks.
Daaron apparently like mazzaro more than most if not all on the site. Good for him. But no one gets them all right.
Mazzaro got ranked in the community list behind Price, Feliz, Tillman, Hanson, Holland, Matuzs, Bumgarner, Alderson, Zimmermann, Parker, Arrieta, Chacin, Anderson, Cahill, Porcello, McDonald, Cecil, Bowden, walden, main, Miller, Wade davis, Carrasco, Gio Gonzalez, Chris perez and martin perez, Aumont, Inoa, Samardzija, Cortes, Jeffress, McGee, Simmons, Poreda and Niese.
I was a little surprised some of those guys got ranked ahead of him (I argued against Samardzija for one), but I still rank a lot of them ahead of him, and Matt Latos was behind him, for example.
I do think the community underrated Mazzaro. maybe we trusted the BA and other analysts too much, maybe we didn’t listen enough to a coach’s poll, maybe we were too worried about small AAA sample, or K rate not high enough, or were too in love with our home team guys (like Niese-btw his era is masking some nice peripherals :)), flavors of the month, or jeffress’s fastball, or over-emphasized Mazaaro’s so-so prior years.
It is likely a combo of all of these things.
But to think the community somehow didn’t in some collective way surmise mazzaro was good or could be very good? 97 rank and discussed well before then suggests people thought he MIGHT be for real and wanted to see more.
I have been patting myself on the back for saying Snider would struggle to oba .300 and slug .400 this year, but I am aware that is a small sample and flukey. i also thought Mike pelfery would be quite good and he has struggled thus far. I overestimated how well Holland would do. No one really thought Kris Medlen would dominate AAA.
We all get some right and some wrong, sometimes not by design.
Given that daaron also predicts Daric barton will be a better overall player than Justin Smoak (conceding the power), i’d like to see a giant mea culpa if Smoak not only has more power, but a higher career on-base average, or batting average, whatever other criteria aside from power, UZR, what have you , before I annoint him more open-minded or a seer compared to the community as a whole.
I see my reply got lost
Here it is again.
I am very surprised my prediction turned out so well.
I spent all of 10 minutes on that post and did little to no research. In all of that, more than 10 specific predictions later and a half of a season, and the best you can come up with is one name?
Any accuracy at all was 100% luck.
Now I ask you, does that prove or disprove my thesis?
that wasn't the only one
just the most glaring.
You mean the research you did that must have consisted of deciding every A’s prospect was better than some Rangers’ prospect of your choosing?
I mean
a semi-trained monkey could have done as good of job.
I used no informatoin that could concieably been construed as statistically significant. Unfortunately, I did what everybody else is doing only worse and probably a little luckier.
So now tell me, do you believe that you base any of your predictions on data that is statistically significant for making predictions?
I was just
voting in a community poll using what info I had. I actually go more on previous season’s stats than most. I would not have voted for a lot of those pitchers ahead of Mazzaro. In fact, if I recall correctly, only about 17 of the 35 would I have, Bowden maybe being the cutoff, and I now may suspect his k rate is too low. McDonald I ranked ahead despite his flyball rate because he pitched quite well in the playoffs. Small pressurized sample biad and visual bias. All I knew of Martin Perez were low minors stats and the bandied about quote, I think from Dewan, the Santana comparison, etc.
Very unscientific. I can’t recall if I knew about the coaches poll.
I don’t know exactly how statistically significant the info I use is. I go by K rates, walk rates, flyball rates, and, wrong as it may be, hit rates, homer rates even e.r.a. somewhat (I try to suppress my learned behavior, and ya know, the fact that e.r.a. really isn’t such an awful predictor-I mean, Roger Clemens career e.r.a. was lower than, say, Jamie Moyer’s), age, league and park effects to extent I know them, I don’t go out of my way to learn the details, what scouts say, sure, what management and coaches say, what fellow amateurs think, age relative to league, hell yeah, opposing coaches if known, I tend to go a lot by the most recent season on a young player, unless injuries cause poor performance, or there is some kinda low k rate low babip flukiness to it.
After that I go with entrails. Tie goes to the runner.
btw
I didn’t think Martin Perez was better than Mazzaro, if I recall correctly. He’s an example of someone I didn’t know enough about and was too far away for me. I could be wrong, as there was a lot of support for him, and BA etc liked him. Samardzija I recall specifically arguing against despite a decent major league debut.
oh
and I go by reports on mph, pitches, etc. And hearsay. I am big on hearsay. If Dewey Finn says the Texas Rag said Holland was at 97, I take it with a grain of salt. As the judge, I decide what evidence is admisible and how much weight it deserves.
Shoud you go by the TL coaches poll over my methods? I’d say by all means.
This is an exercise in the wisdom of crowds, but there were worries about stufing and other bias, plus idiots are part of the crowd. And even with the stuffing, that allegdly came from fans of some pretty stocked systems.
Will say
I have always loved MILB’s prospect list because it is from actual scouts..
Just like point made on Mazzaro being voted as the pitcher of the year.
Huh?
MILB.com’s prospect list is from Mayo, unless I’m completely missing something.
by RedSoxFaithful on Jun 21, 2009 8:53 PM EDT up reply actions
an interesting system
Have you ever heard of the statistical term ‘design of experiments’?
I appreciate the considerate response
I doubt Mayo’s uses any methodology to determine if his list is actually based on data that is statistically significant. We probably would have heard of it.
In any system where there are many shades of gray (in developing medicines or predicting prospect performance for example), one would need to set up a carefully controlled system to determine the effectiveness of ones efforts.
If you are interested in the topic of design of experiments, there is a nice write-up of it in wikipedia.
A poll is notscientific
As compared with 1 ,2 or other peoples opinion?
the answer is
A poll can be highly scientific or not scientific or anything in between.
It depends on the purpose and the design of experiment.
Well it is the same system
As any college ranking poll… is that scientific?
by novaoakland on Jun 22, 2009 11:21 AM EDT up reply actions
some are, some aren't
Obviously coaches polls are not, but if you look at the bracketology of the NCAA basketball placements, those are probably pretty scientific.
The record of success for BBall bracketology is pretty amazing and I would guess they come complete with highly scientific ratings systems and monte carlo simulations.
I appreciate the considerate response
I doubt Mayo’s uses any methodology to determine if his list is actually based on data that is statistically significant. We probably would have heard of it.
In any system where there are many shades of gray (in developing medicines or predicting prospect performance for example), one would need to set up a carefully controlled system to determine the effectiveness of ones efforts.
If you are interested in the topic of design of experiments, there is a nice write-up of it in wikipedia.
Gammons was big on Toe Nash, too.
I like Mazzaro, but the Carpenter comp reeks of hyperbole, and this spam level of posting is obnoxious. You’re not Vin’s mother. Give it a rest.
David.Aaron
I will give you credit, since you were saying he should be in the community poll since around #33:
http://www.minorleagueball.com/search?btn=Go&order=date&q=david.aaron&type=Comment
He started getting some serious consideration in the second half of the top 100. Didn’t make it on until 97, as you know. And some folks felt that was too high.
Looking ath poll posts, I started saying he should be on around 90. In fact, by the time 97 came aound i voted Ivan Dejesus, but there were some posts where I ws saying he belonged on before Adenhart (may he rest in peace).
Someone else posted their topp 300 and had mazzaro 199, but said he had real potential to shoot way up the poll.
I also thought he could shoot up my personal list. Is he better than Feliz? maybe for now he is, but Neftali really could be something special (not saying mazzaro can’t be eiither. Jhoulys Chacin also profiles a s a groundballer. I don’t know that he is better than Chacin based on. or cahill, although he seems more polished out the gate, but cahill is improving as the season goes on.
These aren’t my top 3, just guys I am throwing out there who have been raised in comparison for one reason or another.
Will Mazzaro be bettter than Aaron poreda? Brad Holt? Chris Tillman? Nick barnese? Kris medlen (a guy who didn’t make the top 100 and is struggling now but lit up the IL this year)? Who the hell knows yet? Maybe you do and he’ll be better than all of them.
But I wouldn’t beat this horse as much as you have. So far, you look ok in having championed him early on.
Do you still think Daric Barton is a better prospect than Justin Smoak? Because you posted that as well.
"Do you still think Daric Barton is a better prospect than Justin Smoak? Because you posted that as well."

by PissedMick on Jun 22, 2009 1:22 PM EDT reply actions 1 recs
ha!
Nice pic.
http://www.minorleagueball.com/search?btn=Go&order=date&q=daaron+barton+smoak&type=Comment
Not trying to bust on daaron. All in good fun. We’ll see how barton and Smoak turn out.
That sums it up
I am so tired of reading this crap. Everyone KNEW who Mazzaro was entering the season, a good prospect. We all STILL know who he is, a good prospect. A GOOD prospect is anywhere in the top 250 if you’re talking on a overall scale of the minors. OK, now I sound like Denny Green. I only stopped in to see why this post got another 90 comments, and was rewarded with a gem from PissedMick.
Maybe I've Missed Something...
…because I’m not going to spend three hours reading through every reply in this threat. But I’ve read some, and maybe this is too simple a question to pose at this point in the discussion..
But —
It’s been 26 innings, folks. No one has "hit’ or “missed” with their prognostications for Mazzaro yet. The jury’s not just still out on him, but the only crime that’s been committed is all this huffing and puffing and bickering and ego-stroking. No one’s right or wrong about Mazzaro yet because he’s still learning his way around the A’s clubhouse. He hasn’t done anything to make anyone “hurt” for underrating him. And in no way can you glean enough information from 26 innings to say someone was a “huge” miss. The Mazzaro hit-or-miss conversation shouldn’t be had until the end of the season.
If the real point was to investigate the current paradigm of prospect evaluation, then there are literally thousands of better prospects with which to start that discussion. You know, guys who have definitively over- or underperformed expectations over the past 20 or so years, guys who “are who we thought they were” or were nothing of the sort.
For all the preaching going on about signals and noise and the null hypothesis, you’d think the original poster would understand the only thing to reject here is the premise that there’s enough data to justify holding up Mazzaro as evidence of anything conclusive.
by Jaeti on Jun 22, 2009 4:38 PM EDT reply actions 1 recs
you got it
so no need to read the other hundred or so posts. But there is a good picture of Elvis.
All Mazzaro has done
Was dominate AAA as a 22 year and quite frankly, dominate in the majors also. Will he regress to the mean? of course.
However the key question that you missed is this: do you think that this performance is any surprise to the AA coaches who said he was the best pitcher in the Texas league last year?
What exactly
did the scouts think they were being asked? Was his performance the best in the Texas League. Best stuff? Best projected success over his career in majors? Most likely to succeed off the bat? I doubt they are surprised. And his ranking suggests to me people in the community poll thought he was good and could even improve. No real issue in asking him to do it again. Brett Tomko also had a fairly “dominant” first half season in the majors, ya know.
In the last start against the Dodgers
Average FB sat at 93.5 MPH (56 total) going up to 96 MPH with good command and good movement.
Are you saying Tomko has this stuff?
Do you think
That the Texas League coaches might have seen a few games with a 93.5 MPH FB with good command?
Again, Mazzaro is still only a metaphor for the lack of statistically significant information that is routinely used as the basis for performance estimations.
uy
Arthur Rhodes went 7-5 with a 3.60 e.r.a., 7ks per 9, threw 95+ and scouts loved him when he was 22 in 1992. Look what happened.
I am not saying Mazzaro isn’t good or wasn’t perhaps underrated coming into this season but you are making a huge deal out of small sample and are about to break your arm patting yourself on the back. Coaches loved Chad Hermanson, but it didn’t mean squat in the long run. Sometimes they are right, sometimes wrong. You aren’t advancing the ball.
If you think this thread is about Mazzaro
and I am patting myself on the back, then you have not read what I have written. You have read what others assumed that I thought, but I have simply have not read what I have written.
The words maybe, but never comprehended their meaning.
The problem here
Is that you aspire to advance the discipline of prospect evaluation beyond its current capacity, but are oddly ignoring an aspect of the process that will always be relevant and integral to assessing players: sample size. Or, put in obvious terms, players who are still prospects cannot be used to form conclusions about the nature of prospect evaluation.
Mazzaro may, years from now, be someone you could rightfully point to when trying to hold a discussion about what’s right and wrong with prospect evaluation. But right now he offers absolutely nothing of value with regard to that larger discussion because, truly, he’s still a blank slate.
Judging the strengths and weaknesses of prospecting can only be done using players who have moved beyond the point where they can be described as prospects.
by Jaeti on Jun 23, 2009 12:17 AM EDT up reply actions 1 recs
What "performance?"
Let’s say Mazzaro is useful enough to accumulate at least 1800 career innings. His first 26 innings would make up a whopping 1.4% of his career total. For an analogy, compare that to a 120-minute film. If you watched 1.4% of that film, you would have sat through all of 100 seconds of it. The opening credits, if you will. Just enough time for half of an opening montage and the words “Vin Mazzaro” and “Produced by: Oakland Athletics” to flash across the screen.
Let’s really split some semantic hairs here.
Has Mazzaro “performed” in the Major Leagues? Yes.
Has he done so to a degree that his statistics thus far should be classified as a representative “performance?” No. Much as the actors and director would be puzzled by anyone who watched 100 seconds of their work and tried to declare it a success or failure.
So it doesn’t matter if opposing coaches from last season would be surprised or not by these 26 innings. If they thought they could draw anything remotely resembling a meaningful conclusion based on them, I would tell them they are misguided.
I believe in America
America made my fortune.
Dude (or gal, or whatever)--
Mazzaro has pitched 25 innings at the big league level. He’s made four starts. He has 14 strikeouts. His ground-ball rate is basically average.
I’m an A’s fan, but I firmly believe that anyone who is not kidding themselves should be able to see that his performance so far is a combination of small sample size and luck. The scout’s quotes are more or less ridiculous— it’s results-based analysis at its very worst, and it’s going to look really dumb in about a month.
Side note: we’d better hope he isn’t Chris Carpenter, who was basically somewhere between mediocre and terrible for six seasons with Toronto while often being injured. It’s not often that a #4 starter gets hurt, misses a season, then miraculously comes back with vastly improved control and better stuff— and more to the point, in Carpenter’s case, he did it for another team.
Linda's in the cold ground, won't see her anymore
Somewhere out on the highway tonight, the drunken engines roar
It's just one of those things, one of those things
-- Al Stewart, "Accident on 3rd St."
In memory of Nick Adenhart and all victims of drunk driving
The injuries were the reason why Carpenter was mediocre to terrible.
It wasn’t until he was relatively healthy that his control improved and his stuff was better. He was a top prospect and had previously shown good control and stuff.
As for Mazzaro, his stuff isn’t as good as that of Carpenter’s. He’ll be a solid pitcher, probably more of an innings eater; however, I’d much rather have Porcello, Tillman, Holland, Feliz, Cahill, Anderson, among many of the other pitchers who were ranked ahead of him. They are rated that highly for a reason. They all have better stuff, most have better velocity or higher ceilings. This is not to say Mazzaro is a worse player, it’s to say that their talent levels are better based on the opinions of scouts and observers.
"When Justin Upton faces Lincecum, I think Christ might appear in the heavens, and the world will end." -JakeFree
Velocity
I don’t think the velocity comment is really true.
Of the ones with MLB experience, here are their avg fastball velocities (according to Fangraphs):
Porcello – 91.0
Holland – 93.3 (9 bullpen app and 5 starts)
Cahill – 89.1
Anderson – 91.6
Mazzaro – 92.7
Clearly Holland stand out a bit (especially since he’s left handed) – but you do have to take into account that a lot of his pitching has been from the bullpen. Aside from Holland, Mazzaro’s velocity has been pretty much as good if not better than all the others.
would these be the same scouts
that provided the information that Price was the best pitching prospect last year?
Because outside of 14 innings with TB, Mazzaro looks like a close comparison.
simple question for Daaron....
If Tommy Hanson > David Price
& David Price > Trevor Cahill
& Trevor Cahill > Brett Anderson
but Jarrod Parker, Matt Latos, Brian Matusz, Chris Tillman, Rick Porcello, > Vin Mazzaro and Brett Anderson
then is Vin Mazzaro > Aaron Poreda ?
by SteveHoffmanSlowey on Jun 22, 2009 5:36 PM EDT reply actions
ha
that’s awesome considering the ridiculousness of this thread.
This is actually pretty simple
there are a couple main reasons that Mazzaro is ranked where he is and a pitcher like Derek Holland is ranked much higher after both began the year in relative obscurity. I’m going to keep it pretty simple here, I don’t really feel like putting too much into something that is this easy to comprehend.
The first difference is pitcher type/stuff.
Mazzaro is a right handed pitcher who relies mainly on a sinking 2 seam fastball and pitching to contact,.
Holland is a power left handed pitcher who can throw his 4 seamer 95+, getting a lot of swings and misses and K’s.
Given that last year both showed comparable control and lets just say comparable secondary pitches, most people are going to view Holland as the more valuable prospect.
Second difference is previous results.
Mazzaro had two previous seasons that I will call disappointing. He was hittable and walked a few too many batters. Coupled this with an average K rate and you get the results he had, unimpressive. Compared to the other prospects ranked ahead of Mazzaro, I don’t think there is one who has two seasons as poor as these of Mazzaro’s are.
Holland had a solid first year in 2007 3.95 K/BB rate with good H/9 and K/9 rates. Maybe a few too many HR but a very good debut. He showed signs that he could be a decent prospect, much more so than Mazzaro had to that point.
So Mazzaro had two seasons of relatively average at best pitching coming into last year and Holland had one season of above average performance. I personally think that Mazzaro’s performance in AAA last year put some tarnish on his season and left some doubt in peoples eyes as to which Mazzaro was for real. Was it the one who dominated AA or the one from the previous two seasons that was pretty average. If he had continued his level of performance from AA to AAA I think he would have been thought about a little higher. But again, look at the performances of the prospects ahead of him, find me one that has had two seasons like Mazzaro had in 06 and 07. Put that together with his AAA performance and you have why people were willing to rank Mazzaro ahead of guys like Porcello, Parker, Matusz.
Let me reply to the 'stufff' question
I went to pitch fx and compared the last game the Holland started against the Yankees and the last game that Mazzro started against the Dodgers and what did I find?
The numbers say Mazzaro has better stuff than Holland. Yes, Mazzaro’s average FB was 93.5 MPH with excellent command and movement.
Holland
Pitch Ave Speed Max Speed H-Break V-Break Number Thrown Strike Perc Nibbleness
(FourSeam Fastball) 91.59 94.4 8.56 9.58 62 62.90 5.91
(Changeup) 83.43 87.7 6.93 9.68 12 66.67 3.74
(Slider) 78.80 83.9 -1.27 2.01 12 50.00 6.72
Curveball) 73.75 74.4 -3.40 -0.87 2 100.00 3.32
Mazzaro
Pitch Type Ave Speed Max Speed H-Break V-Break Number Thrown Strike Percentage Nibbleness
FF (FourSeam Fastball) 93.50 95.8 -8.85 9.51 56 67.86 4.78
CH (Changeup) 85.37 86.4 -7.09 6.12 3 66.67 3.58
SL (Slider) 84.11 87.2 3.02 2.01 28 67.86 6.37 0.449
CU (Curveball) 80.60 80.6 4.69 -2.19 1 0.00 7.10
FT (TwoSeam Fastball) 90.30 90.3 -9.02 7.60 1 0.00 4.15 (only 1 2 seemer thrown)
and Holland
was at 97 mph last year near the end of the year, and dominated in his last few starts and the playoffs. Were coaches asked to list jst starters wo played most of the year in the league? Were feliz or Holland even really considered? Do pitcher’s fast ball speeds or break change sometimes from year to year?
So people may reevaluate where they see Holland versus Mazzaro after another year of data. hardly shocking.
do you have objective evidence
That Holland was consistently at 97 last year?
What was his average FB and what kind of movement?
no
and I don’t give a rat’s ass about it. I read in BA most likely that he was hitting 97, not sitting, but fairly regularly by the end of the year.
It would be wonderful if all of that information waas out there, and maybe it is. I didn’t know it, and even if it was out there I don’t know whether to trust the source unless I built and held the speed gun. Waspretty confident he was mid 9s and he absolutely smoked AA in his last few regular season starts and then 3 playoff starts or so that weren’t listed in his minor’s record and milb.com but I toted them up and he certainly seemed like the real deal. In the few innings they have pitched so far this year Mazzaro’s results have been better. Holland has thrown a little harder, but as pointed out above some bullpen.
And Randy Wells has been awesome too. Doesn’t throw as hard, didn’t in a coaches poll, but look at that k rate and k to bb and he is lefty. Tom Glavine redux I tells ya!
This whole flippin’ thread should be about how we ALL missed Randy Wells. And how well his first 8 starts would go.
2 questions
If you had known that for any given start (the last one for example) that Mazzaro’s FB is a full 2 mph faster than Holland, do you think that would have influenced your estimate of their relative places as a prospect?
Second, do you think that it was possible to obtain a better read of comparible stuff?
Holland average FB velocity this year is 93.3
Mazzaro’s is 92.7. Considering one is a lefty and one is right handed the relative rankings are fairly accurate.
93.3 puts Holland ties at 13th with Lester, Oswalt and Guthrie.
Here are a few questions for you:
Do you expect Mazzaro to keep a BABIP in the .230’s?
Do you expect him to maintain a WHIP of 0.97?
Do you expect him to keep a BB/9 at 2.10, which is below any of his career marks?
Do you expect him to maintain a LOB% of 84.8%, which is well above league average?
All of those will have a profound effect on his ERA. Right now his peripherals do not support your argument and your claims that Mazzaro has a fastball 2 MPH faster than Holland’s is a blatant lie.
Get over your love affair for Mazzaro. He is a nice mid tier pitching prospect that has had an excellent start to the season pitching 3 of his 1st 4 games in pitchers park and going against substandard offenses. Regression is a nightmare.
Hopefully when he gets roughed up people will be a little more kind than what you deserve with this pointless chest beating.
JD’s like, "you want some fucking pitching? Here’s all the pitching you can stand. Now choke on it, bitches!"- RCCook
by laxtonto on Jun 23, 2009 1:18 PM EDT up reply actions 1 recs
I was as accurate as the data
that was provided. I provided the context. It is a fact that in Mazzaro’s last start, he threw 56 FBs that averaged 93.5 mph. It is a fact that in Holland’s last start, he threw 62 FBs that averaged 91.5 MPH.
This is according to pitch fx.
What I don’t appreciate is being called a liar when I provided the complete context of the data. It appears that your data also includes relief appearances which, in my mind is not comparing apples to apples.
However, with that said, it seems to me that when a thread is reduced to honest people calling other honest people liars, it is time to stop.
I won’t be posting in this thread again. (after a few clean up posts)
Well, just for the record,
the notion that Mazzaro’s AAA performance last year “tarnished” his status was as moronic then as proclaiming him the next Carpenter is now. Both based on small sample sizes and completely ignorant of the role of luck.
Linda's in the cold ground, won't see her anymore
Somewhere out on the highway tonight, the drunken engines roar
It's just one of those things, one of those things
-- Al Stewart, "Accident on 3rd St."
In memory of Nick Adenhart and all victims of drunk driving
eh
John admitted it influienced him slightly. It wsn’t just those numbers, but fact he also had not been great in 2 prior seasons, AND it apparently wasn’t all just luck, as reports were his fastball speed diminished and his other offerings weren’t as crisp. It wasn’t simply those numbers alone but fact, admittedly tiny sample, they reinforced slightly some other info.
Yes
I’m sure that his getting hit in AAA was all “bad luck” and small sample size and had nothing to do with his actual pitching not getting the job done
I don't know why people insist on thinking that pitchers' BABIP has something to do with "getting the job done"
It
is
fucking
meaningless.
Linda's in the cold ground, won't see her anymore
Somewhere out on the highway tonight, the drunken engines roar
It's just one of those things, one of those things
-- Al Stewart, "Accident on 3rd St."
In memory of Nick Adenhart and all victims of drunk driving
Question
Do flyball pitchers tend to have a lower BABIP then groundball pitchers?
"When a true genius appears in the world, you may know him by this sign, that the dunces are all in confederacy against him."
-Jonathan Swift
by King Billy Royal on Jun 23, 2009 3:21 PM EDT up reply actions
49 hits in 33 2/3 inn.
I love how he started getting unlucky upon his promotion to AAA
overheard in AAA mangers office
Manager: What happened out there today Vince
Mazzaro: What do ya mean skip
Manager: There were gettin to ya pretty good
Mazzaro: No way, I was dealin, those balls they were hittin off me was all luck,
Manger: Luck, is that what you call that
Mazzaro: Yeah you see, there’s this really awesome guy called PaulThomas and he clued me in on BABIP,
Manger: BABIP what??
Mazzaro: yeah skip, it means that it doesn’t matter how I really pitch, eventually, those balls are gonna be outs.
Manager: Oh is that how it is, well I’ll tell ya one thing, you keep layin those babies in there you had better be one lucky SOB
Mazzaro: You see skip, that doesn’t really matter either, this rad guy PaulThomas also let me know about small sample size.
Manager: I gotta here this
Mazzaro: Yeah, since I’m only gonna be making a few starts here before the season ends, what I do here doesn’t even really matter. It has no bearing on me as a pitcher at all.
Manager: Well it matters to me and the rest of these guys. Let me tell you one thing, we’re evaluating every start and I don’t give a damn about no BABIP or small sample size or any of the rest of that. What matters to me is when you are out on that mound, your job is to get outs. I don’t care how you do it, but it’s what the coaching staff expects of you, the guys playing behind you expect, and the fans sittin in those seats.
thats how fuckin meaningless it is chief
So, your response is an appeal to the authority of minor league managers...
Not only is that a logical fallacy, it’s a particularly comical one, given that baseball managers are a group which is selected for characteristics other than (the perverse side of me wants to say “selected intentionally for lack of”) intelligence.
Linda's in the cold ground, won't see her anymore
Somewhere out on the highway tonight, the drunken engines roar
It's just one of those things, one of those things
-- Al Stewart, "Accident on 3rd St."
In memory of Nick Adenhart and all victims of drunk driving
babip
It isn’t completely meaningless. Some pitchers have shown some modicum of effect. Carlos Zambrano has tended to suppress. Chris Young.
For example, BABIP rank among starters:
Zambrano 5th in 2005, 2nd in 2006, 6 in 2007, 13th in 2008.
Young 1st in 2006, 1st in 2007, 9th in 2008 (among those with 100 innings pitched, as he got hurt), 16th in 2009, and he has been pretty “bad” this year.
That would be flukey random, for Young to lead in that stat 2 years running, if it is all just random. of course, he is a flyball guy in PETCO. Wakefield tends to suppress.
Mike Mussina, otoh, was around .320 4 of the last 5 years, including his 20 win season last year.
Anyway, i didn’t raise the hit level at AAA, John did. Moreover, mazzaro’s FIP did rise to 4.00. He gave up 3 homers in 33 innings compared to 3 in about 4x more innings in AA. The FIP certainly smooths his line, but he did go 3 to 4. His groundball rate fell. His babip had been kinda low for a groundballer in AA. So maybe the feeling was that was the lucky portion.
It wasn’t a huge sample, agreed. And most pitchers the babip is random and reverts to “normal”. His tRA* was 5.68, putting him 89th for pitchers with s many outs recorded, right ahead of Josh Tower. His WAR was -.1. His tRA* in AA was 4.40, his WAR 3.3. Per stat corner.
But there can be some slight effect. Albeit with mazzaro it was a very small sample, as conceded. John added the flavor of reports his velocity was down at year end and other offerings didn’t seem as crisp.
If you have like 6 years of major league data, you only have to regress a pitcher's BABIP halfway to league average
Or something like that. Don’t eat me if I’m slightly off, but that’s the kind of effect we’re talking about. Ask Tom Tango if you want a more exact number.
For the purposes of this site, minorleagueball, in which virtually all statistical evaluation references sample sizes against wildly fluctuating competition in increments of 400 AB or less, you cannot draw any conclusions from pitchers’ BABIP.
Linda's in the cold ground, won't see her anymore
Somewhere out on the highway tonight, the drunken engines roar
It's just one of those things, one of those things
-- Al Stewart, "Accident on 3rd St."
In memory of Nick Adenhart and all victims of drunk driving
well
I will look at Tango, and I understand it isn’t huge, but Young, for example, is so consistently low. Leading the league 2 years in a row is a pretty odd feat for a supposed random non-repeatable stat. Zambrano too. There are outliers of course.
and over a few starts, sure, random as all get out. of course, it wasn’t just his babip. still small sample.
Chris Young is a flyball pitcher in Petco Park
The Padres have a 47 point spread between their BABIP in Petco and on the road this year; opposing teams have a 43 point spread.
It’s not a statistical proof or something, but I’d say that’s pretty damning for a ten-second internet search.
Zambrano probably really does have some ability in that area. So does Barry Zito, actually, of all people. Greg Maddux had it. The thing is, it takes a long time for a pitcher to amass enough evidence to present a convincing case that he’s doing something repeatable. In the minors, where defenses are iffy, players come and go, the quality of competition varies wildly and the sample sizes are invariably small-ish… just forget about it.
Linda's in the cold ground, won't see her anymore
Somewhere out on the highway tonight, the drunken engines roar
It's just one of those things, one of those things
-- Al Stewart, "Accident on 3rd St."
In memory of Nick Adenhart and all victims of drunk driving
yeah
I mentioned that in my first post. But he was FIRST 2 years in a row. And Petco may help him especially. And I noticed that about Zito, too. But yes, tough to compile sufficient evidence for minor leaguers. I just wanted to counter the idea that it is completely meaningless in general, although certainly for 49 hits in 33 AAA innings it is. Like chicken soup in reverse, it don’t help.
Just a thought
May it wasn’t bad luck and maybe it wasn’t that his AA perfomance was a fluke. Maybe after 137 innings in the Texas heat, he was just tired.
yes
That’s exactly what people said it might be. John said that. The problem isn’t that AA was a fluke maybe, but how repeatable was it. I actually think his 4 e.r. game last night was pretty good, bouncing back from 4 e.r. early, 6ks 1 bb. Against the giants of course, so discount.
What this kid is talking about
is exactly what JParks did here two (?) years ago…
I’m not sure what all the comments about. Obviously he is a troll. But nothing that he is saying isn’t as earth shattering as he is claiming.
What exactly am I missing here?
Remember: baseball guys... baseball...
apologies
JParks does Texas stuff
JPahk did this stuff
http://www.minorleagueball.com/2007/11/8/04955/2157
Remember: baseball guys... baseball...
I appreciate the link
Unfortunately, no, this is nothing like I am discussiong.
But your effort to find the link is appreciated.
what are you saying then?
a collective think tank on rating prospects on both statical and anecdotical perspectives (i.e. scouting reports?)
Tango does something like this for defense.
Remember: baseball guys... baseball...
Yes in part
However, I think that the missing element is that there needs to be a method to make anecdotal perspectives ststistically significant in much the same way that the medical field does with medication (how do you feel?).
Again, I am not confident of the solution, I am still asking questions.
How do you feel
The problem with how do you feel is that its objective. There isn’t a basis for it at all. While my pain is a 6 yours could be a 1, or vice versa. Not to mention other than a preliminary chart, that data isn’t used as raw data to draw conclusions. In a way, those “statistics” are an anecdote. A story, or brief description one gives ones self.
What you suggest is the 20-80 scale scouts use. They take objective data and turn it into a raw score that an be correlated to other observations. It probably should only be used against their own ratings, because the are the ones giving the score, but its a standard in the industry.
That is as “statical” one can get with anecdotes. And its not like those numbers are processed into a large sample size over every game. We (the community) hears those numbers from 10 sources all year, about a very small handful of players. And it is usually only used in the extreme cases.
Montero has 80 power. Porcello has a 70 fastball, Max Ramirez is a 20 behind the dish etc. etc.
I don’t see much of a future in what you are arguing unless all scouting reports were made public (obviously never going to happen). Or there were private unbias third parties (BA, BP, JS, KLAW) releasing this for all their scouting reports. But these guys get their reports from primary or secondary sources (most don’t see that many games a year expect mabye Law).
Remember: baseball guys... baseball...
Now you are thinking
Although your definition of “objective” is wrong.
I think you have presented to problem pretty well. You are also correct in the there may be no future in my questions.
Do you think that there is a way to leverage the people who have to play against and coach against a guy?
Also, if it were possible to do so, is there a way to do this in a way that is statistically significant?
No.
Please, before you use a term like statistically significant, learn the limits of statistics. Statistics can do very many things very very well – they cannot, unfortunately, predict the future
TheSouthWing.com - A Magazine of essays, prose and poems
by OldProspects on Jun 23, 2009 12:17 AM EDT up reply actions
do you think
that statistics are used in predicting the outcome of presidential elections?
By stupid people?
Yes.
Political polls told us that Barack Obama had 52% support amongst likely voters on November 4th. It told us that John McCain had 45%. It told us nothing about how many of those voters would change their minds, or how the 3% undecided voters would break. In other words, it could tell us what was, but not what would be.
TheSouthWing.com - A Magazine of essays, prose and poems
by OldProspects on Jun 23, 2009 12:56 AM EDT up reply actions
Nate Silver would like a word with you.
Well, actually he’s probably a bit too busy these days for that . . .but he’d certainly disagree.
OldProspects - google is our friend
simply does not understand that under the statsitics umbrella, there are different branches called descriptive and predictive (inferential). He is correct that descriptive statistics describe a history of sorts.
However, predictive statistics are used all the time in the estimation process. It is how I make a living in part. It is how the insurance industry makes a living. It is how most big government contractors make a living.
I suggest googling ‘prediction interval’
That refers to large quantites of cases
Again, not baseball where we’re dealing with dozens of prospects who are in way similar to each other. At the major league level, where players tend to behave in a more consistent fashion than in the minor league level (while still developing), systems like PECOTA have had some success. But not at the minor league level. And certainly not at the individual minor leaguer level.
In other words, insurance companies have no problem betting on the health of a million people, but they would not bet their company on the health of one individual person. They assume that there is a solid likelihood that they will err on the single cases, but don’t care because they are trying to manage the risk. That being said, the recent efforts by some to manage their risk using methods like this, did not exactly serve companies like Lehman Brothers especially well.
TheSouthWing.com - A Magazine of essays, prose and poems
by OldProspects on Jun 23, 2009 10:40 AM EDT up reply actions
this is progress
At least you have admitted the existance of inferential statistics.
However, inferential statistics are applicable in many fields that do not have the luxury of millios or even thousands or sometimes even hundreds of data points. If fact, my calculated are aften based on dosens of data points.
Does this affect accuracy and confidence? Sure, but at least there is a basis for prediction other than guess, which is often difficult to defend.
What?
Just because you can show your work doesn’t make it any more correct. I’m not denying the role of statistics, nor do I imagine it having anything outside of its role. Statistics are invaluable for telling you what was and what is. The past and present are good indicators of what will be. But they don’t tell you the future.
TheSouthWing.com - A Magazine of essays, prose and poems
by OldProspects on Jun 23, 2009 2:08 PM EDT up reply actions
You know, I don't know why you're exasperating me so much
I think it’s because of your self-congratulation and your suggestion that there might be a unified prospect theory, but I’m not sure.
But now I’ll try to be nice: I think a formula could be very helpful, and I endorse your efforts to make one. Just realize that at best, it will only tell you part of the story. If you don’t understand why, then read Asimov
TheSouthWing.com - A Magazine of essays, prose and poems
by OldProspects on Jun 23, 2009 2:14 PM EDT up reply actions
Honestly
Nate Silver needs to read up on the mistakes pollsters made in the 1948 Presidential election. It’s a little remarkable how similar some of his methods are to what they did
TheSouthWing.com - A Magazine of essays, prose and poems
by OldProspects on Jun 23, 2009 10:36 AM EDT up reply actions
Comparison Fail
We’re talking about athletes who provide an ever-increasing of data about themselves through which we must sift. Running for president and accumulating “stats” in the national general election is by and large a once-in-a-lifetime event.
That doesn’t mean that statistics can’t be used to make certain predictions about a particular election, but those predictions will never be based on a candidate’s prior performance in a presidential election.
I think what Nate Silver has proven, though I don't know if it was his intention
Is that it’s far less about the people running for POTUS than it is about the people voting.
by thejd44 on Jun 24, 2009 12:49 PM EDT up reply actions
What People Say
How many people said Mazzaro was the best pitcher in his league last year? How many said someone else was the best? The specific answers to these questions aren’t particularly important, because unless everyone who speaks to and assess minor league baseball players collectively lost their minds or conspired against him, I’ll go out on a limb and say that more people said other players were better than Mazzaro than the converse case.
This isn’t hard to reverse engineer. If he was generally ranked behind X number of other pitchers, I think that means “what people say” was “leveraged” into establishing that hierarchy of pitching prospects. And they are all still prospects, and in reality will largely still be so even when they pass 50, 100, 150 innings pitched.
Seriously, let’s talk in four years.
What People Say, cont.
“What people say” isn’t synonymous with “who gets to vote for an award.” Nor is “who votes for an award” synonymous with “most capable of voting for the most deserving.”
It’s fine to say some people thought he was the best, but it seems fairly obvious that very few people thought he was one of the twenty best pitchers in the minor leagues.
now I am intrigued
Exactly what was he talking about and what did he do?
Please be specific if possible.
Not that it will be useful to wade in so late...
but I guess these threads seem kinda silly, and I feel like I missing something…
First, it seems kinda obvious that Mazzaro is reason why Daaron is posting, I mean the subject and content of the OP makes it clear it’s gloating. But it is funny how he trolls the community with the OP, then spends the rest of his time, first denying the obvious, that’s he’s gloating about Mazzaro, and then spends the rest of the time confusing a simple topic with a bunch of useless pseudo-academic BS.
So, I guess I don’t understand the running conversation, and probably because it’s troll junk, or maybe it’s because it’s what I already do as part of prospecting, which I’ve been doing for 15 years. I weigh scouting reports of stuff, current stats, trending stats, and then I use prospect lists in this order: BA’s league top 20s (because its the people who saw them play), BA’s Team Top 10s (because its how the team sees their prospects), and then a big gap and all the various top 100s or whatever done by what I consider to be armchair scouts.
So anyway, it’s always going to be inexact because prospect development, especially with pitchers, is too complex, too many variables in terms of mental make-up, and genetic make-up, not to mention how complex the throwing motion can be. One mechanical change, however subtle, can mean the difference between boom and bust. A kid could be a late bloomer, gaining 2-3 MPH and exploding onto the scene. Or has late mental maturation, then he makes a developmental jump, that may be what’s happened to Bailey.
Anyway, an entertaining thread, I salute daaron for being able to troll the community every couple of weeks, but this thread he’s elevated his game, and almost has a 200 post thread! Kudos to you sir!
Poster formerly known as artie
and I salute you
for at least having a method in estimating the predicted performance of a prospect.
Establishing a method is half the battle.
This question is: do you think that it is possible to devise a methon that would make a predicted estimation that is statistically significant? Especially in the area of non-numerical data (top 10 lists, etc).
this is life however
I would rather have a much smaller thread with a much higher signal to noise ratio.
All the posts say that this is stupid and that Daaron is a moron etc reduces the signal to noise ratio quite a bit and makes it less likely for poeple to actually try and understand the conversation.
I knew it would happen and actually inserted that into my calculation of what I wanted to accomplish, but I did not think that the thread would grow this large.
one more thing
would you say that your method is repeatable?
is yours?
Ya know, John Sickels has actually been pretty forthright about how he came up with his opinion and that he may have been wrong. Albeit he said number 3 after his debut and you are now saying upside higher.
I am not certain that you have really quantified a method here any more than anyone else has.
Texas league coaches poll seems to weigh heavily.
Last 9 winners of that Texas league pitcher of the year (I don’t think they considered Holland or feliz as those guys were not there for very long):
2000 Bud Smith
2001 Tim redding
2002 Kirk saarloos
2003 Travis Blackely
2004 Jeff Francis
2005 Jason Hirsh
2006 matt Albers
2007 Josh Geer
2008 Vin mazzaro
A lot of those guys were soft tossers.
You also emphasize mph. John says that Vin did up his velocity last year. he had ben 88-92. last year up to 90-93. 94-95 early in the season he says before he tired out some.
The reports from AAA per John were he was tiring out and less crisp.
John also had concern that 2 1/3rd seasons weren’t great, 2/3rd AA season great, so he wanted some repeat. No harm in that.
Cameron says his fastball in his debut was higher at first then slowed some. Did he get tired, or was he trying to command it better? Was there high adrenaline at first? In any event, fangraphs has him averaging 92.9. Very nice.
Will he tire this year as well? He seems to be fine now, and it looks like he has sustained last year’s mph bump up. As a fellow Jersey boy, I do like Vin.
Let’s go back to the Texas league winners.
Albers is a good comp. 92.2 mph back in small ‘96 stint. he is a groundballer. he was 21 on John’s list of top 50 pitchers going into 2007. Hirsh was 15.
The community in 2007 listed Hirsh 25 overall. Albers 31. BA had Albers 42. Albers was off the radar in 2006 really.
None of those Texas League winners coaches has lit the world on fire. Hirsh had ben good prior to that. Albers had one so so year prior to that and a good one. Injuries too.
BA and BP ignored mazzaro on their top 100s. John B-. Community here had him at 97 and he was seriously considered prior to that.
Actaully, the community didn’t do so badly, based on the small sample thus far, although the AAA and major line thuis year together, on top of his AA line last year, may finally lend him more legitimacy.
I know the sample was small, but the AAA line did kinda harm him somewhat as far as his rankings here. And you did cavil as early as 33 on that poll ‘wither Vin mazzaro", although you didn’t actually name him. :)
IF, big if, things continue, Mazzaro may end up being the best of those Texas league poll winners. The added fastball oomph may push him ahead of the other hardish throwers on the list, especially the biggest comp, who I gues is Albers. His numbers weren’t quite as good as Vin’s. Hirsh’s numbers in TL quite comparable, but he is more a flyballer, kinda death in Coors.
Since Joh said his speed also diminished late in year, perhaps observers put too much emphasis on his AAA, but of course that ws on top of poorish low A performances.
I dunno, what am I missing from your methodolgy, how much of it is repeatable?
I will say, SO FAR, he looks underrated, even though the community can be proud it ranked him ahead of a lot of the experts.
And again, SO FAR, randy Wells is the real case of someone who got completely underrated and ignored. MAzzaro, well, he SO FAR has done what a lot of people suspected he MIGHT do: if his mph held up, he sustained and built on his success, he could be pretty damn good. And it looks like SO FAR he has been doing those things, and we may not expect a sub-2.00 era rest of year, haha, it seems like no one is shocked he has done ok in a few starts. far from it.
But what should we have done? said, withour collective brain “gee, he is Texas league pitcher of the year-I know the last 8 of those guys have done nothing yet, but he DOES throw harder than all of them, at least this year, at least until he tired out at year end, and I guess he has made the big leap forward, I will ignore his low A seasons AND his AAA line, and we should rank him ahead of Holland, etc., even though all the other sources don’t rank him top 100 at all.”?
Nope, 97 ws a damn good community rank and better than anyone else thus far, and you can be proud if somehow in some small way you got him onto the poll sooner than otherwise.
Now, back to daric barton versus Justin Smoak. :) Teasing ya dude.
my bad
I think Albers didn’t make the community list in 2007. Did he? i may have gotten some of my notes wrong. :) Oh well. In any event, Albers and Hirsh got more play from BA and john than did Vin after his AA season, likely due to Vin’s poor prior seasons and small blow up in AAA.
Doing my part to push to 200!
Ok, I’m going to try and translate your questions into something more easily understood…
Your first question seemed to ask do I feel a method could be created that significantly impacts the success rate of predicting prospect future performance? Of course, I believe my method increases my success rate when prospecting. I believe my method better identifies false positives and sleepers than just relying only on websites and publications.
Are you asking if believe this system could be standardized? There’s no way to standardize the scouting reports or the top 10 lists because their methods aren’t standardized enough for me. The scouting system tries, but it still relies on observational data captured in one point in time with that player by a person. Granted that person is trained and all that, but it’s still suspect because it’s not purely objective. Even radar gun reports are dubious.
Do I think my method is repeatable? Of course, or else it wouldn’t really be a method I used over time. I have a spreadsheet that assigns a number to a player based on stats, trends, and scouting reports. But that’s just a starting point because then I have to research the top 10/20 lists, and any googled news articles about the player before I slot him. This whole method is especially error prone because I’m making a judgment on subjective observational reports. I trust my judgment, but it doesn’t make it any more valid than any other armchair scout/prospector.
Everything has to be taken with a grain of salt, and you can try to clear the noise as best you can, but there’s too much noise to avoid relying on gut on some level. I don’t feel you could construct a reliable formula if any part of that formula relied on word of mouth reports.
Poster formerly known as artie
The very best post in the thread
and I appreciate you posting it.
Maybe someday we can discuss the concept of adding statistical significance to your very key terms of standardization and repeatability.
This thread however has devolved into honest people calling honest people liars and I think it is time to quit.
Maybe later in the season we can discuss this again?
Don't forget
that nearly every post of yours on this thread (as well as all of your posts on this subject) are dripping with condescension toward someone, whether you talk about signal to noise, get all high-minded about your statistical knowledge and put down others for lack of same, or compliment someone on their thoughtful post (implying that others’ posts are not). You are not blameless, sir.
+1
nothing else to say … if it wasn’t for the way Daaron presented his stuff, I think a lot more people would be engaged positively. As it stands, people are only engaged because of his flamebaiting (and yes, I consider “tone” when pondering if a post is flamebaiting).
Interesting
Lets not be surprised Daaron posted about this..we all knew it was coming. I KNEW I would come home from vacation and see a post of his about Mazzaro’s start.
Daaron, Mazarro might turn out to be a stud, but do you realize there was not a good reason to rank him as a top 50 prospect?? He had poor seasons and one breakout season. We all know you want to crown the kid, but come on. He was a nice prospect coming into the year, but the fact that was won the stupid AA pitcher of the year award means NOTHING. His stuff is
just
not
THAT
good.
Again, you are trying to make him into something he is not. To use what Peter Gammons writes just makes you look like a bigger idiot.
Since Mazzaro has given up 4 runs in 2 innings to the Giants, does this mean we don't get another one of these threads later tonight?
by thejd44 on Jun 23, 2009 10:46 PM EDT reply actions
I'm sure the umps were biased
They probably threw off his game just to piss Daaron off.
final line
4 runs in 6 innings. That’s a 6.00 ERA and he had a 1.5 WHIP. I wouldn’t want that from my number 5 pitcher. Obviously, Mazzaro is garbage, and I have no idea what all the fuss was about
Dumbass
Dont you know this thread isnt about Mazzaro?
Solace: Law says he's a fourth OFer
PaulThomas: I think Keith Law is only a fourth analyst
It' about how
each year the 97th best prospect will always be #1 in Daaron’s heart.
Adoptive parent of Kyle Nicholson
Comparisons
Personal issues in this thread that sort of deflect from the discussion aside, the case of Vin Mazzaro has gotten me interested enough to join this forum after several years of just reading.
I do have a special interest in Mazzaro after drafting him in a strat keeper league this winter. Back then, I liked the reports of how his overall game had progressed – which isn´t all that unusual for a young pitcher to finally harness or improve his stuff and especially his command in the strikezone.
What I do find interesting is that Mazzaro´s prospect profile is that it is somewhat similar to that of Brandon Webb about 6 years ago. Both were drafted as promising pitchers with a nice sinker but questions regarding their ceilings and ways to go. Both spent a couple of seasons posting mediocre stats in A-ball before bursting onto the scene with a breakout year in Double A. Both received some recognition with prospect sources after that strong Double A season but failed to get mentioned among the top P prospects in Baseball and weren´t even on most top 100 lists.
Both – along the way – saw their stuff & command improve to a point where a lot of people are / were surprised about their initial major league results.
Now, of course, Mazzaro has a very, very long way to go to even get close to Brandon Webb. I just find it interesting that both – up until now – have had very similar paths to the majors and in terms of how their were regarded as prospects.
Mazzaro – from what I´ve seen – does have a plus fastball – in terms of velocity, movement and command, plus – at least so far – seems willing to challenge hitters with it which is a very good starting point. The big question going forward for him is whether his breaking pitch and changeup will be good enough to be effective even on days when his fastball isn´t in such a great shape and to be able to keep hitters off balance too. That will probably decide whether he´ll be a solid middle of the rotation starter like a Jake Westbrook or a Jon Lieber or whether he can become a frontline type along the lines of a Brandon Webb or a Tim Hudson eventually. And – of course – assuming he remains reasonably healthy.
Finally, Mazzaro was somewhere in the # 90 to # 120 overall range in just about every prospect ranking prior to this season, so the # 98 spot on this list was in the upper range compared to other sources.
by Doob on Jun 25, 2009 3:04 AM EDT reply actions 1 recs

by 














