Discussion Question

I've always been interested in the value of "pressure" on the basepaths.

How much can the runner change the pitcher’s performance? What is the value of forcing an error?

by Daniel Berlyn on Jul 22, 2025 4:43 PM EDT reply actions 0 recs

There is some on this in The Book

If I remember it right, it helps to be an OK stealer, but if you are someone that takes off quite often the batter’s performance drops.

Jeff Zimmerman - Protecting the world from RBI's and Wins from my mom's guest house.

by Jeff Zimmerman (TucsonRoyal) on Jul 22, 2025 6:24 PM EDT up reply actions 0 recs

I'm talking about forcing errors on defense.

Misses on pickoffs, throwing into the dugout on a chopper, airmailing a throw to the backstop. These plays are often the difference in wins and losses and aren’t accounted for much of the time and they add up a lot over the course of a season. I think this is partly why the Angels always seem to outclass their Pythag and whatnot. They make it an organizational philosophy to exploit chinks in armor. I think there are some newer designer stats that account for these things but we don’t use them much.

by Daniel Berlyn on Jul 22, 2025 11:07 PM EDT up reply actions 0 recs

This argument doesn't hold water

Teams that outclass their pythag do it because they score more runs at “clutch” times than at other times. Unless you’re arguing that the Angels “put pressure on” (whatever that means) at some times and not others (and if so— why are they not doing it ALL the time?), there’s no way it could have the impact you describe.

The impact (if any) of scoring runs through “smallball” would be seen in the difference between Pythagorean record and third-order record, which is based on offensive and defensive linear weights rather than actual runs scored/allowed.

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

by PaulThomas on Jul 23, 2025 12:00 AM EDT up reply actions 0 recs

I'm talking about efficient baserunning, which definitely has importance.

Putting pressure going from first to third and second to home, not giving the other team a lot of chances to throw you out in the process. Aggressive baserunning. Using speed to one’s advantage.

I don’t think we’re on the same page. One of my theories is that a Juan Pierre or Rajai Davis could have more than his perceived value if they are causing the opponents to throw the ball all over the yard.

by Daniel Berlyn on Jul 23, 2025 12:43 AM EDT up reply actions 0 recs

In your theory, does this "aggressive baserunning" lead to the Angels scoring more runs?

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

by PaulThomas on Jul 23, 2025 12:11 PM EDT up reply actions 0 recs

Little by little, yes.

And why wouldn’t it? If you want to look at what the market inefficiencies are, just look at how the Rays and A’s draft. Speed guys have been at the top of the list the last couple of years. Just that little edge that makes a difference and doesn’t cost as much.

by Daniel Berlyn on Jul 23, 2025 3:21 PM EDT up reply actions 0 recs

Anything that causes a team to score more runs

does not, ipso facto, cause that team to overperform its Pythagorean record. Runs scored is half of the calculation used to determine Pythagorean record.

You are, of course, correct to observe that the chance of an error should be evaluated in determining the value of a basestealing attempt (the “break-even” point is thus probably a little lower than the figures arrived at purely through SB/CS linear weights) and, i guess, in a player’s baserunning runs for other situations, but that’s not a mystery— we already know how to do that (it’s just a matter of compiling the data).

And it has nothing whatsoever to do with differential from Pythagorean record, which is not about how many runs you score/prevent but when you score/prevent them. The only “strategy” (I use the term loosely, since it’s not at all clear that it’s a good idea) which helps with that is extreme leveraging of your bullpen (so that you always have a good pitcher in when the game is close and always have an awful pitcher in when the game is not close).

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

by PaulThomas on Jul 24, 2025 1:55 AM EDT up reply actions 0 recs

Close the patent office, everything that can be invented has been? For all our stats we know basically nothing. We’ve only started recording baseball events half-decently in the last few years, and there’s still a long way to go. All of sabermetrics to this point is based on hash data and approximations. Come back in fifty years and the idea that all the significant work has been done will look even dumber than that patent clerk. We’ve barely started.

by timprov on Jul 22, 2025 5:07 PM EDT reply actions 0 recs

timprov beat me to the patents line...

We certainly are better at most things because of saber advances, so I guess you could say we don’t have as far to go. But there are still big projects that will take as much time to answer, if not more, as the projects that have been answered.

Breaking down the pitcher/hitter matchup has only just begun. Figuring out how things like pitch selection, speed, and break, are matched by the hitter to produce batted balls that are good and bad depending on where they’re hit and how hard will be huge.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Jul 22, 2025 5:14 PM EDT reply actions 0 recs

sportsvision

all sorts of new ways to evaluate talent are going to be available once the sportsvision project takes off. Physical data is going to supplant a lot of statistics as the most relevant part of sabermetrics. Pitch f/x was the beginning (and is still being improved). Hit f/x will provide more batted ball data than ever before. And the new sportsvision project will be even bigger.

The name of this new game is going to be locations, velocities, and accelerations. Magnus effects on balls, ball-turf interactions, fielder placement and motion will all be measured (even if only indirectly).

A lot of questions are going to be answered with this.

I know this really isn’t true sabermetrics, but when it comes to assigning value to players these physical data will all be integral to that process.

There are movements to do this now: measuring line drive rates, ground ball rates, etc. But these are all in bin form. When does a line drive cease to be a line drive and become a fly ball? When does a hard hit ground ball become a line drive? To have a continuum instead of cut-offs will be nice. I know hit f/x will do this, but it’s not widely used yet (to the best of my knowledge).

As an engineer this is going to make baseball very interesting for me.

by gorilla_baller on Jul 22, 2025 5:16 PM EDT reply actions 0 recs

You beat me to the sportsvision stuff. That's going to flip statistical analysis on its ear once enough data is collected.

http://www.fangraphs.com/blogs/index.php/pitchfx-summit-recap
http://www.nytimes.com/2009/07/10/sports/baseball/10cameras.html?_r=1

by JonBBT on Jul 22, 2025 5:18 PM EDT up reply actions 0 recs

And, honestly, I wouldn't count out the effort made by Trackman, the company that does ball tracking for golf.

Beyond the Boxscore Not a member? Sign up.

by Sky Kalkman on Jul 22, 2025 5:20 PM EDT reply actions 0 recs

i had never

seen the trackman stuff… if it can do some of the things they say then it’s pretty awesome… i’d love to see some data that they’ve collected

by gorilla_baller on Jul 22, 2025 5:38 PM EDT up reply actions 0 recs

Hit F/x should completely revolutionize hitter analysis

There’s no more important variable to getting a base hit than the speed of the ball off the bat. Right now, we know utterly nothing about that, and only a very small amount about the launch angle (enough to tell us whether a ball is a grounder, liner, fly or popup).

As of today, batting average is basically a black box. We know some people are good at it, but have very little idea why. Five years from now, it’ll be far better understood.

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

by PaulThomas on Jul 22, 2025 7:50 PM EDT reply actions 0 recs

MLB is going in the right direction

PitchFX and HitFX are revolutions, and hopefully someday soon, FieldFX (positions of fielders when the pitch is thrown, positions when the play is made, etc) That will really expose who has the best positioning, throwing, and so forth. Right now, fielding sabermetrics are pretty fuzzy since there is no hard data to go on, just comparisons between peers. That’s good, but not great.

Catchers are still a ways off from being completely figured out. Catcher Fielding is a crap shoot, and most metrics don’t even bother trying. The dynamic between pitcher and catcher may either be too simple (little effect, but lots of random noise) or too complex (too many variables) to ever get a good handle on.

A lot of smart people are tying dollars to wins, and even making that team-specific, but there’s a long way to go in evaluating trades/signings in that team-specific manner and getting that into the mainstream.

It’s funny how some 8-10 years later we now have even stodgy old broadcasters quoting OPS. It’s only a matter of time to get to WAR.

by AnotherAaron on Jul 22, 2025 8:23 PM EDT reply actions 0 recs

On that subject

Capturing the value of catching defense is really the least understood aspect of the game right now, I thin. Between figuring out pitchcalling, blocking bad pitches, blocking the plate for runners, arm accuracy, arm strength, etc., there’s a lot to be done.

Honestly, I think capturing manager value is another necessary step to a complete understanding of what’s going on in the game. It’s tough, though.

And more completely, we need to figure out how to link player “tools” or “skills” to performance in a more substantial way. Prospect evaluation is still something of a crapshoot; why do some guys (Andy Marte) seem to have it all but have no success at the major league level? How can teams draft better - what in college/HS stats or physical tools correlate best to major league performance?

Injuries, as a component of player value, are also somewhat poorly understood. There’s certainly work being done there to figure out mechanics, but why do some players seem to be “injury prone” when those injuries aren’t actually connected? Is that just chance?

by OccamsRazor on Jul 22, 2025 9:15 PM EDT up reply actions 0 recs

Keep it Simple Stupid

:-)

You are obsessed with bodily wastes A turd being held under my nose, this team being a steaming pile, Jonny Donuts having a stinky diaper, the front office pissing themselves, the team crapping the bed…

I think you should seek counselling.

(AJM to Josey Donuts)

by Michael Cave on Jul 23, 2025 12:50 PM EDT up reply actions 0 recs

I see what you did there

by thejd44 on Jul 23, 2025 1:28 PM EDT up reply actions 0 recs

8-10 years later

than what?

When did Bill james baseball Abstract first get wide distribution? 1981 or so? I know I have a 1982 edition somewhere, I think. I even had the Eddie Epstein Minor League handbook. 1994.

by wobatus on Jul 23, 2025 1:16 AM EDT up reply actions 0 recs

OPS is nice...

OPS is nice, but let’s get them working on wOBA. There’s a stat I can get behind.

by nobodyinparticular on Jul 23, 2025 1:10 PM EDT up reply actions 0 recs

Probably setting up WAR better

like I read this thing on the nats website about haveing somthing like 37 WAR but 26 wins

by The_Fan on Jul 22, 2025 10:09 PM EDT reply actions 0 recs

The tricky part...

about all of this new data and new ways of looking at information is the level of credibility. More detailed cuts of data lead to less significance in the numbers and therefore more assumptions need to be made.

That being said, there is a lot more to learn! I particularly love the Pitch F/X stuff and think that some correlation between quality, command and predictability of pitches will be developed (ie, great stuff allows you to be more predictable and less accurate with pitches whereas lesser stuff requires more command and less predictability).

"God, I'm from Cleveland. When is it going to be our time?"

by BStal11 on Jul 22, 2025 10:36 PM EDT reply actions 0 recs

The next step

I think the next thing we can do is try to combine biometrics and sabermetrics. Right now we’re making developments in the output of physical actions (pitch flight, hitting charts, etc). I think the next step would be to observe how efficient and safe players are in achieving these outputs and using the two in concert.

by sagecoll on Jul 22, 2025 10:55 PM EDT reply actions 0 recs

On certain issues

I’d say yes. Minority opinion apparently. For example, once you get past strikeouts, walks and groundball/flyball tendencies of a pitcher you are gilding the lily. You can certainly learn more about how a pitcher does what he does, this much break, velocity, this many swings and misses, etc., but beyond the basics you are looking at minute differences relatively.

by wobatus on Jul 23, 2025 12:32 AM EDT reply actions 0 recs

Exploiting market inefficiencies can be the difference between making the playoffs and not.

These things add up over a full season and when multi-millions are at stake, every potential edge is worth looking at.

by Daniel Berlyn on Jul 23, 2025 12:44 AM EDT up reply actions 0 recs

I am not saying

not worth looking at. I just think beyond the basics in some areas you are going to find you are looking at the trees for the forest. Sure, every half a WAR may count, luck is a residue of design, etc. I am not too sure how much you are going to improve your odds of getting that half a WAR (much less whether it translates into actual wins) beyond the basic stats I mentioned for pitchers (toss in durabilty, etc.). And I just meant in certain areas you aren’t going to advance very far beyond that.

Fielding is an area where vast improvements in data collection and analysis have been made and will still. And there will always be something.

But I was laughing the other day at someone on a fangraphs comment who belittled someone because he was relying on “antiquated” and poor stats like batting average and home runs. And i am thinking that’s going too far. Sure, batting average isn’t everything, but it is still a fairly useful component of a players value over time, as long as you are aware of the contexts and other variables. And home runs antiquated and a poor stat? Yeah, I guess Jim Rice sucked because all he did was hit home runs with a high batting average and didn’t walk (and no, I am not getting into a Jim Rice for the hall argument).

by wobatus on Jul 23, 2025 1:07 AM EDT up reply actions 0 recs

On a purely numbers level, there are nooks and crannies left when it comes to offense.

People are just starting to be able to quantify things not available to the public in years past. Velocity of a ball hit, for example. There is still a lot of progress to be made.

by Daniel Berlyn on Jul 23, 2025 1:24 AM EDT up reply actions 0 recs

yes

I am aware of that. But I stick by my belief that beyond the essential verities you are not getting as much return. You get something. Just not as big of a leap as when folks discovered, hey, strikeouts to walks is more consistent than batting average on balls in play against a pitcher. Or ya know, signing Wayne Garland to a huge contract based on one good e.r.a. year ain’t such a hot idea.

by wobatus on Jul 23, 2025 9:46 AM EDT up reply actions 0 recs

Dave Cameron

I am reminded that he wrote a couple of articles this off-season justifying the Padres holding onto Brian Giles and slamming the Twins for signing Kubel (using Hinske as an example of someone they could have signed instead), based on their WAR’s last year or so and salary versus value. There were debates about each of these opinions. John for one wrote a piece here suggesting it was worth the risk for the Twins, Kubel might break out. I think Cameron had way over-analyzed the situation, or actually, underanalyzed, putting a huge amount of emphasis on, in Giles case, the bounce back year of an older player otherwise in recognizable decline pattern over a long stretch, and in Kubel’s case ignoring minor league track, injury history and some evidence of improvement. He ignored historical patterns of these players in particular and age patterns in general and over-relied upon his own pet stats, and fairly small samples at that in the scheme of things.

And not to be harsh, but he seems like the kind of guy in your fantasy league who knows all the stat-head predictive minutiae and still gets his clock cleaned every year (and not even in some standard roto or 5×5 league but ones that count all the stats and try to mimic reality as closely as possible). I wonder if he held onto Giles for half a season and let someone else draft Kubel. Eh, he likely doesn’t even play.

by wobatus on Jul 23, 2025 12:46 AM EDT up reply actions 0 recs

WAR doesn't have predictive value.

I don’t why people think so much otherwise.

by Daniel Berlyn on Jul 23, 2025 1:27 AM EDT up reply actions 0 recs

I think

Cameron was just guessing that things would be close to what they ahd been recently.

by wobatus on Jul 23, 2025 9:40 AM EDT up reply actions 0 recs

Of course it does

A player who is 40 runs above average this year is a hell of a lot more likely to be 40 runs above average next year than, say, I am.

There’s no bright-line dichotomy between “stats with predictive value” and “stats with no predictive value.” It’s a continuum. Every stat has SOME predictive value, outside of the sort of bizarre trivia you only find in the dark corners of ESPN.com (day/night splits or something stupid like that).

Even pitcher “wins” and batting average have predictive value, just not as much of it as more refined statistics.

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

by PaulThomas on Jul 23, 2025 12:19 PM EDT up reply actions 0 recs

day/night splits/

Hey, I have nightvisison, and I party so hard I can’t get up for day games. So i suck at all hours.

by wobatus on Jul 23, 2025 12:51 PM EDT up reply actions 0 recs

Funny thing: The Cubs are trying to only use Harden in away games that take place at night

Because apparently this year he’s been better in those.

by thejd44 on Jul 23, 2025 1:53 PM EDT up reply actions 0 recs

I should add that I think, at times, there is something to guys who are terrible during the day

But it’s hard to know (especially through statistical analysis) who the big-time party guys are. At any rate, the manager should know and should probably not play these guys in too many day games.

by thejd44 on Jul 23, 2025 1:54 PM EDT up reply actions 0 recs

I think we still have a long way to get on catcher defense (primarily the non-throwing aspect)

True impact of managers/coaches (I know we are pretty sure about some of this, but I think we need to really study these things), a proper definition and then study of “clutch” (the worst part of The Book, IMO), and of course how to get the average joe to realize that numbers aren’t evil.

by thejd44 on Jul 23, 2025 3:35 AM EDT reply actions 0 recs

I agree wholesomely.

Evaluating catcher defense properly has been limited to advanced scouts and skewed statistical accounts so far.

by Daniel Berlyn on Jul 23, 2025 3:23 PM EDT up reply actions 0 recs

yes

Yes I definitely agree with this.

by John Sickels on Jul 23, 2025 3:32 PM EDT up reply actions 0 recs

opinion

I come at this from the perspective of a prospect analyst. Until such time as pitch/fx and such data are available at all minor league parks, it doesn’t have much relevance for my work. For major league stuff it’s fascinating, but i’m not exactly sure what it tells us yet. We are so early into the process.

I think the big breakthroughs are on the fielding side, and a lot of progress has been made there in the last five years.

But I also think we are in danger of overthinking things to some extent. Perhaps this is just because the cutting edge stuff is getting to the point where my math breaks and I just don’t understand all of the discussion any more.

by John Sickels on Jul 23, 2025 10:09 AM EDT reply actions 0 recs

I don't understand much of a the higher math now

But I can read about string theory, grasp the concepts, and not understand ANY of the math. When I read some of the more technical discussions on The Book’s blog, I get fuzzy during the math, but as long as you can get the gist of it, I think you can still be pretty well-versed in the newer ideas.

by thejd44 on Jul 23, 2025 1:31 PM EDT up reply actions 0 recs

I am reminded

for some reason of Elaine’s father’s comment in Seinfeld:

“I don’t need a weatherman to tell me it’s raining. I just stick my head out the winda.”

It’s the same thing to a degree. Not with fielding, but give me the basic stats over a period and I think I pretty much have most of the info I need. Although the concept of basic stats is getting pretty expanded these days. Don’t get me wrong, I love to know pitch movement, line drive rates, swinging strike percentage, etc.

by wobatus on Jul 23, 2025 12:10 PM EDT reply actions 0 recs