Navigation: Jump to content areas:


Pro Quality. Fan Perspective.
Login-facebook
Around SBN: Falcons Vs. Giants, NFL Playoffs 2012: G-Men Advance, 24-2
Sprint-network-bar2 01

Who else gets confused by all the new stats?

Hello fellow community members,

I must admit that I jumped on the statistical bandwagon many years ago. Even as kids my brothers and I would mention that Tony Phillips was underrated and would make a great leadoff hitter. When Moneyball came out it opened up my eyes to a new world of stats to analyze baseball players with.

That being said, I must admit that I am becoming increasingly confused by all the new stats out there. For example, when somebody says someone has a good wRC+ I really don't know what that is referring to, and what number is considered good.

Therefore in this thread I request that other community members can either:

A: Post a stat to get clarification of what it means and what is considered a normal and good number

B: Post why you think a certain stat is key to judge players on

All I would ask is to keep it civil as John is very busy and we don't really need people arguing in a rude manner about why one stat is better then another (besides it makes us look nerdier than we are which is actually hard to do).

KBR

Tweet Comment 111 comments  |  Add comment  |  5 recs  | 

Do you like this story?

Comments

Display:

To start it off if somebody could explain wRC+ to me?

What does it tell us? How is it calculated? What is a good number and an average number for a player?

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 4, 2026 10:51 AM EST reply actions  

I can't answer most of this, but maybe I can say something helpful

wRC+ is supposed to be an all-in-one offensive stat — it stands for “weighted runs created,” I think — normalized for park factors. And it’s also scaled so that 100 is league average. So an average wRC+ is always exactly 100. If you go 0-for-everything your wRC+ is 100, but in general I think a wRC+ of 150 is supposed to be 50% better than average - which is ridiculously good.

Pujols’s career wRC+ is 167. In 1992 Barry Bonds’s wRC+ was 203; in 2002 it was 245.

Here’s the Fangraphs primer on it.

Not actually affiliated with whygavs.

by WHYG Zane Smith on Jan 4, 2026 11:18 AM EST up reply actions   1 recs

bah, I hate auto-markup

The crossed-out part was supposed to start with “negative 100.”

Not actually affiliated with whygavs.

by WHYG Zane Smith on Jan 4, 2026 11:18 AM EST up reply actions  

THanks Zane!

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 4, 2026 11:36 AM EST up reply actions  

I don't believe in "all in one" stats. I want to know what the slash line is.

If somebody tells me "player X had a wRC+ of 117, I know that means he’s good, but it doesn’t give me a mental picture of the player.

Telling me that a guy hit .297/.363/.483 with 22 steals in 26 attempts gives me a picture of what his talents are.

by John Sickels on Jan 4, 2026 12:51 PM EST up reply actions   3 recs

REC

As much as one stat can tell you, it is always better to view a players full stat line to get a good picture of what kind of player he is.

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 4, 2026 2:09 PM EST up reply actions  

wRC+

wRC+ is pretty simple. It takes wOBA (which adjusts the fact OBP is worth more than SLG), and adjusts with park factors and league factors (ex. AL is stronger than NL).

There are things a slash line can’t tell you, like how much playing in NYS or Fenway boosts offense or playing in Petco depresses offense, or how someone should look in the NL vs the AL. wRC+ tries to capture those factors.

It’s hard to visualize wRC+ since it’s just a %. But if you look at 2 players with similar slash lines, but one’s wRC+ is 120 and the other is 130, you know the second player is better because he played in a tougher league/park and posted the same stats.

by valencia on Jan 4, 2026 2:51 PM EST up reply actions  

That makes sense for your purposes

When you’re trying to isolate a player’s individual skills so you can project what he might look like in the future, a stat like wRC+ doesn’t do a whole lot for you. Catch-all stats like that are much more economical in nature, rather than predictive. They tell you a value by putting many different facets of the game on the same scale, which is something a triple-slash line is far less precise with. But a triple-slash line is much more precise in terms of telling you what particular skills a player has, which is much more useful in terms of understanding where a player derives his value, and also tends to lend more to predicting future performance. Both schools of thinking are very useful, and there’s a lot of crossover between their usage, but but I certainly understand that for trying to project future performance of minor leaguers, why a stat like wRC+ wouldn’t have much appeal. I would be less interesting in a minor leaguer’s wRC+ than how his other skills project, so I might get a picture of what his wRC+ looks like down the road.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 4, 2026 2:57 PM EST up reply actions   2 recs

wRC+ could still be used to adjust for park/league at the MiLB level

like for example, High Desert is very offense friendly, so we just tend to ignore the stat lines there. But with wRC+, you can see if how he’s hitting relative to the league’s average, to get a sense of how good his numbers are even in a hitter friendly environment.

While wRC+ is the best “catch-all” stat, it’s best used in conjunction with wOBA and slash lines.

by valencia on Jan 4, 2026 3:03 PM EST up reply actions  

Agreed, good point

Especially since minor league park factors aren’t so readily available. My point was more that conceptually, its more designed to describe value rather than evaluate skills or make predictions, all of which are viable enterprises for statistical analysis, and a lot of times people seem to forget and try to do things with certain stats that they’re simply not designed for.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 4, 2026 3:38 PM EST up reply actions   1 recs

right

Right, that’s what I’m trying to get at. A catch-all stat doesn’t help me analyze or predict what a minor league player might do in the future. Slash lines give me a much better breakdown of what his skills and tools are.

I’ve used OPS for years just to get an idea about how the player is doing in comparison to his peers. I am well-aware of the weaknesses in OPS (OBP is underweighted). It has only been the last few years that wRC+ and wOBA have cropped up into public consciousness with the rise of Fangraphs.

The problem with integrating these measures into the book is that the audience that I’m writing for in the book is not as heavily-sabermetric as the many of our posters here. When I start throwing numbers around, it turns some readers off.

It was groundbreaking to use OPS and Secondary Average when I started writing books 15 years ago. I’m well-aware that it is “behind the times” now. But I’ve also seen a lot of whiz-bang-cool-stats come and go, so I tend to be cautious about pushing new metrics until they are well-established.

That doesn’t mean I don’t look at them. I was looking at ground ball rates and BABIP and things of that nature for a couple of years before I started incorporating them into the book.

by John Sickels on Jan 4, 2026 3:51 PM EST up reply actions  

Good points John

Personally I like to read your anecdotal comments more then your statistical comments about a prospect, as the anecdotal stuff is something I do not personally have access to.

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 4, 2026 5:06 PM EST up reply actions   2 recs

Absolutely, another good point

I could definitely see citing SIERA and wRC+ as a bit alienating to your audience.

Even more, linear weights in particular are tricky for minor leaguers since they’re dependent on run environment. The formula for major league wOBA (and thus wRC+) changes slightly every year, because the run environment in the majors changes slightly every year (thus the value of a single or a home run is slightly different). Translating stats like this to the minors properly would involve calculating those numbers for those run environments as well. I’m sure there are some people out there who try and do this, but at best its unclear and not as well organized as it is in the majors (and even if the majors the standardization for stats like this isn’t always perfect). Either way, you can still look at these stats within leagues comparatively, but it means you using them to compare players from different leagues is a bit tenuous.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 4, 2026 4:55 PM EST up reply actions  

I know I am going to be made fun of for this

But can you explain SIERA to me?

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 4, 2026 5:07 PM EST up reply actions  

you should not be made fun of for that

SIERA is one of the most complicated stats, basically the principle is a little while ago, a man named Voros McCracken found that there was little to no correlation from year to year for BABIP. He found that the important things were strikeouts, walks and homeruns, Dave Studenmund then found that a pitcher has little to no control of his Home Run to fly ball rate, and created xFIP. SIERA builds on this, it uses strikeouts, walks, groundballs, and fly balls, Matt Swartz, the creator of SIERA has found all sorts of important things. He found that having a lot of strikeouts will give you a lower BABIP, and found that pitchers will be hurt a lot more if they walk a lot of people, than if they walk a middle amount, he also found that pitchers who gave up more groundballs had easier to field groundballs. It’s usually not available for minor leaguers because of batted ball data, but minorleaguecentral.com has calculated it. For more info read Matt’s five part series about SIERA

by Bososx13 on Jan 4, 2026 8:14 PM EST up reply actions  

also SIERA is scaled to ERA

so you should know what’s good and what’s not.

by Bososx13 on Jan 4, 2026 8:17 PM EST up reply actions  

THanks man

I am a big believer in FIP but not so much xFIP. SIERA seems like it takes those to the next level so I will have some fun going over different pitchers to see how they did.

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 4, 2026 8:32 PM EST up reply actions  

Why not xFIP?

It is a better predictor of future performance than FIP. If you are talking about just what happened in one season, I could see you not wanting to use xFIP, since it does add or subtract HR allowed to what actually happened. But if you are trying to figure out a pitcher’s skill level, xFIP is a really good way to do so.

by cookiedabookie on Jan 5, 2026 8:53 AM EST up reply actions  

I don't agree in normalized homerun rates

Certain guys are prone to give up more homeruns.

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 5, 2026 9:55 AM EST up reply actions  

Well FWIW, its not normalized HR/PA

Its normalized HR/FB, so it does credit flyball heavy pitchers as being more homer prone, and most evidence suggests that the majority of pitchers only have marginal control over HR/FB. This isn’t exclusively true though, and this is one of the things SIERA is attempting to correct for (along with FIP/xFIP’s assumption that pitchers have no control over BABIP). One of the axioms of SIERA is that strikeout pitchers tend to give up weaker contact, and thus have lower BABIPs and HR/FBs. There’s also some selection bias with fly ball pitchers. Fly ball pitchers who are particularly homer prone tend not to make it to the majors—the fly ball pitchers who do make it to the majors tend to have some skill in generating popups in particular, but also generally in HR/FB reduction. So fly ball heavy pitchers tend to have lower HR/FB than ground ball pitchers. On the other hand, for ground ball pitchers, since they can keep their HR rate low in other ways, we often see slightly higher HR/FB.

While xFIP doesn’t correct for these issues, it still gets pretty close to predicting future performance—almost as close as SIERA, both of which are substantially better than other popular DIPS metrics (FIP, tRA, QERA). The flipside of the coin with a complicated stat like SIERA, however, is that its quite rich in terms of inputs, and therefore also more flexible. The only in inputs for xFIP are strilkeout rate, walk rate, fly ball rate, and league average HR/FB, so you can only improve it so much by toying with the values associated with those inputs. Having more inputs with SIERA makes it a bit harder to understand/calculate, but it also gives you a lot more to tweak and adjust not only to improve it, but also to keep it in line with changing run environments.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 5, 2026 10:22 AM EST up reply actions   1 recs

I know it is a normalized HR/FB

It just seems to me that some pitchers flyballs SHOULD travel further. I would expect a flyball hit off Jamie Moyer to go a hell of a lot further then one hit off Pedro Martinez for instance. SIERA does seem to be on the right track though and I look forward to analyzing it further.

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 5, 2026 10:25 AM EST up reply actions  

Matt Swartz

who originally created it at BP and refined it for fangraphs did a great five part series on it when fangraphs introduced it. Part 1, Part 2, Part 3, Part 4, Part 5

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 5, 2026 10:50 AM EST up reply actions  

SIERA does make adjustments for that

KBR. Read Matt’s five part series that Mark and I both linked for the whole thing, but if I remember correctly, pitchers who have more Ks not only have lower BABIPs, but HR/FB rates too. There was some other stuff too so read Matt’s 5 part series.

by Bososx13 on Jan 5, 2026 4:55 PM EST up reply actions  

Thanks Bososx13

I will check it out.

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 5, 2026 5:00 PM EST up reply actions  

Oh my bad

Missed that you linked it above. Really the series is so good its worth mentioning twice.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 5, 2026 10:59 PM EST via Android app up reply actions  

Also, Re: lower HR/FB

Right I do think that’s in there, but my point was its a bit less clear if that’s because there’s a casual link between K% and HR/FB or if it’s simply selection bias, since guys who have high HR/FBs and are also FB prone would very rarely reach the majors—they would have to have excellent command AND miss a lot of bats. It may have been in that series but it would be interesting to see if there’s a difference in HR/FB rates between heavy FB and heavy GB pitchers. My guess is there is, but that low-k FB pitchers might actually have similar or even lower HR/FB than high-k FB pitchers. Could certainly be wrong about that though.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 5, 2026 11:08 PM EST via Android app up reply actions  

Why?

I’m not a physics major (which is why I’m setting myself up for possible face-palm) but why would a pitch from a slow-pitcher go further than a pitch from a fast-pitcher? The velocity at impact should create a bigger result, no? It’s just a matter of whether or not you can catch up to it.

follow @casetines

by Kenneth Arthur on Jan 5, 2026 5:42 PM EST up reply actions  

Reasoning

I just figured it would be harder to make solid contact by lining up the bat with a hard tosser, as opposed to someone who throws in the low to mid 80s.

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 5, 2026 5:57 PM EST up reply actions  

Did I say something funny?

Instead of just saying LOL how about you give me some constructive feedback. John just reminded posters to treat each other with respect so I would expect you would do the same on his birthday.

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 5, 2026 5:56 PM EST up reply actions  

You sure I was talking to you?

And laughing isn’t being disrespectful.

Lighten up.

by Kelsdad on Jan 5, 2026 6:31 PM EST up reply actions  

Pretty sure since it was in response to my comment

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 5, 2026 6:32 PM EST up reply actions  

that's true, but in a sample size of a season or two as a starter, hr/fb variation from the norm

is much more likely to be a result of luck than skill.

in a single season sample size, xFIP is more predictive of the next season’s ERA than FIP or ERA and roughly equal to SIERA.

if you are looking at a huge sample size of data, like numerous seasons or career numbers, FIP probably captures what’s happening better.

i used to be disgusted, but now i try to be amused . . . - macmanus

by tom s. on Jan 5, 2026 5:33 PM EST up reply actions  

Right, Mark.

Comparing player between the California and Florida State Leagues, for example.

I try to use the KISS principle. I know it isn"t the most advanced sabermetrics, but if I write that Player X had an OPS of +25 percent compared to his league, my father-in-law understands what that means.

If i started talking about his wOBA, my father in law wouldn’t get that yet. And the difference between OPS and wOBA, for what I am trying to do, isn’t big enough to justify confusing people yet.

Now, if I was doing advanced sabermetric analysis of major league players, then that would be different. But that’s not my brief.

by John Sickels on Jan 4, 2026 5:28 PM EST up reply actions  

"Comparing player between the California and Florida State Leagues, for example"

Seems like I’ve seen that comparison somewhere before…..

by Kelsdad on Jan 4, 2026 5:38 PM EST up reply actions  

John, you should check out this story by danmerqury over at Athletics Nation.

Here it is. The part I think you’ll find interesting is how closely OPS correlates to wOBA.

"The Lord has blessed us with birthday cake!"

by ozzman99 on Jan 5, 2026 1:25 AM EST up reply actions  

good stuff

Good stuff….so if OPS correlates to xOBA at .97, it is nearly as good, and for my purposes in the book OPS is really just fine.

by John Sickels on Jan 5, 2026 9:05 AM EST up reply actions  

Exactly.

It wouldn’t be worth the trouble of trying to get your readers to learn a new stat when OPS is very nearly as good.

"The Lord has blessed us with birthday cake!"

by ozzman99 on Jan 5, 2026 10:49 AM EST up reply actions  

I still have a problem with it.

It normalizes over the course of many players (their correlation), but it still undervalues walks and overvalues slugging.

You could have a walk machine and a slugger and have their average combined OPS be the same as their average combined wOBA. That doesn’t make the individual one correct, if you see what I mean.

by mr. maniac on Jan 5, 2026 12:15 PM EST up reply actions  

Avg?

I’ve thought for quite awhile that batting average doesn’t mean much at all. Just out of curiosity, John, do you give much credence at all to batting average, or merely as a reflection of one’s OBP and Slugging?

The wind is in the buffalo.

by journeymen on Jan 5, 2026 4:35 PM EST up reply actions  

I still like batting average

Walks are great but you can’t drive in a guy from second by walking to first. Average isn’t as important as OBP but it still has a lot of value.

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 5, 2026 4:46 PM EST up reply actions  

average

It is something of a legacy stat. As long as baseball exists, people will talk about batting average and ERA.

It is in the language….“Hey, at least I’m hitting .500 when i ask hot girls for dates”. etc.

Do I put credence in it? OBP is obviously more important…SLG is more important…but there is something just aesthetically pleasing about seeing .300+ as part of someone’s stat line. Calls up images from childhood…Rod Carew, Wade Boggs, etc.

by John Sickels on Jan 5, 2026 6:54 PM EST up reply actions  

It could.

Batting average as a reflection of BABIP or vice versa is useful in figuring out certain things.

Fans are typically idiots.

by The Typical Idiot Fan on Jan 6, 2026 1:53 PM EST up reply actions  

Essentially correct

wRC+ is basically what you said it is, but to clarify a bit further, it is based on wOBA. It is wOBA adjusted for park factors and scaled to the league average.

For those who don’t know, wOBA is a linear weights method of assigning offensive value. That means each individual outcome of a plate appearance is assigned a discrete value, which is scaled to OBA to make the scale more familiar (basically so it looks like OBP instead of having to learn a whole new scaling system, similar to how FIP and xFIP are scaled to ERA). For example, the weight for a double in wOBA is about 1.24, and a home run about 1.95. So if a player hits 20 HR you take 20 × 1.95 = 39, and that’s the HR component of wOBA. Every outcome has a similar type of multiplier, and once you do that for each type of outcome of a PA (BB,1B,2B,3B,HR,RBOE, etc., FWIW, the multiplier of an out = 0 so outs are simply ignored, because anything x 0=0), you add up all the values and divide by the total number of PAs for the player, and you have wOBA.

For wRC+, you first have to adjust for park factors, and then multiply by a factor you find by equating the league average wOBA (usually around .325 or .330, depends on the year) with 100, so essentially .325x = 100. Then whatever you get for that “x” value you can multiply any wOBA by and you get the equivalent wRC+.

So lets say league average wOBA is .325 and you want to find out the wRC+ for a player with a park adjusted .345 wOBA.

.325/100=307.69
307.69 × .345 = 106.15

So, a player with a park adjusted .345 wOBA has a wRC+ of approximately 106.

wOBA is also the basis for the offensive component of fWAR (fangraphs WAR), which is wRAA, aka weighted runs above average, aka “batting runs”. To calculate wRAA, you take a players wOBA-lgwOBA (league average wOBA)/wOBAscale (the number in wOBA that converts the scale to look like OBA, changes slightly from year to year but typically 1.15) and then multiply that number by PAs*. That gives you the batting component of WAR. To complete the WAR calculation, you need to adjust for baserunning and defense, and then adjust for position (see positional adjustments on fangraphs) and since all of these things are scaled to average rather than replacement level, also for replacement level (20 runs below average per 600 PAs). Add that all up, divide by 10 (since we’re dealing in runs here, and 10 runs=approximately 1 win) and you have fWAR.

*Technically, you also need to adjust for park factors here, but since those aren’t as readily available as all this other stuff, this step often gets skipped in doing off-hand WAR calculations, which is okay but should be noted in any discussions.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 4, 2026 2:52 PM EST up reply actions  

the thing that confuses me about this is the zero point

Most of the explanations I see suggest that an out is worth zero, as you did. And that works for wOBA — if you have an OBP of 0, you have a wOBA of .000. If you were just scaling so the league average was 100, that seems like it should translate into a wRC+ of 0. But a wOBA of .000 actually seems to translate into a wRC+ of -100. Why is that?

Not actually affiliated with whygavs.

by WHYG Zane Smith on Jan 4, 2026 10:01 PM EST up reply actions  

I was mistaken actually

That was just a formula I was trying to reverse engineer, because the actual wRC+ formula is a bit tricky to find on the web, but I think I found it now (on a forum from mlb.com no less!). It is (wRAA/PA)/(lgRun/PA)1*100*ballpark adjustment. Part of the reason for this that I foolishly omitted was that it circumvents the wOBA scale. The reason for the "1" is to set league average = to 100 (instead of 0), and the reason you get negative numbers is that in in extreme cases (wRAA/PA)/(lgRun/lgPA) will be lower than -1.

For example, take Chin Lung Hu in 2011. He had a wRAA of -4.3 in 23 PAs. The league scored 20,808 runs in 185,245 PAs. So you have (-4.3/23)/(20,808/185,245) = -1.66. -1.66+1=-0.66*100= -66. There also should be a park adjustment for CitiField, but in 23 PAs it doesn’t have a huge effect. Fangraphs lists his wRC+ as -60, which makes sense given the positive adjustment you’d expect from CitiField (if anything it seems like its too large for just a handful of PAs like that).

If you take the example of a player with a .000 wOBA, you can also get the first component of wRC+ by figuring out their wRAA/PA. 0-.316 (lgwOBA) = -.316/1.15 (approximate wOBA scale)= -.274. -.274/(lgRun/PA)=-2.44+1=-1.44*100=-144. So the lowest possible wRC+ (wOBA of .000) in 2011 would have been -144.

Really, techincally, wRC+ isn’t strictly based on wOBA, they’re just both based on the same linear weights. The difference is that wRAA is a straight counting stat while wOBA is a rate, and the issue with direct conversion being that wRAA/PA is not the same as wOBA, because there’s an extra scaling added to wOBA to fit it to the OBA scale, and also that wOBA is strictly based on theoretical the theoretical run creation of the linear weights, while wRC+ takes into account the actual run scoring environment of the league.

The one thing that still confuses me here is why the lgwOBA (.316)/wOBA scale (1.15), which is about .274, isn’t closer to lgRun/lgPA, about .112. Shouldn’t those be almost identical? The former is theoretical runs created per PA, while the second is actual runs per PA. I could understand if they were a bit closer and things just hadn’t been adjusted for the declining run environment enough, but that doesn’t seem to be the case here. I’m going to check my math a bit better later and if I still can’t figure it out ask someone who probably knows better than me and get back to you guys if you’re interested….

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 5, 2026 9:47 AM EST up reply actions  

I'd like to learn

about ISO. I see it used on this site a lot and don’t know what it is.

by HarleyMila on Jan 4, 2026 1:27 PM EST reply actions   1 recs

ISO stands for Isolated Slugging Percentage

And it is essentially a way of separating regular slugging percentage from batting average, giving a clearer picture of power than SLG (FWIW, SLG=TB/AB, and the league average last year was around .400. Above .450 is very good and above .500 is outstanding, while .350 would be considered poor for most positions). You can think of ISO as “isolating” the effect of singles out of SLG. ISO it is calculated as TB/AB - H/AB. You can essentially fudge it by just taking SLG and subtracting AVG, with the only issue here being that you’re dealing with numbers already rounded to the thousandth (three decimal places), so you might be a thousandth or so off in your final result. The major league average in 2011 was roughly .145. Anything below .100 is sub-standard, and anything above .200 is outstanding.

For example, Lance Berkman slugged .547 in 2011 while Mark Teixiera slugged .494, yet their ISOs were identical at .246 because Berkman hit .301 while Texieria hit only .248. So essentially their rate of extra bases per AB was similar, but Berkman hit quite a few more singles, thus we conclude they had similar power even though Berkman did a much better job hitting for AVG.

On the other extreme, a slap singles hitter like Juan Pierre routinely posts ISOs below .075, which is a pretty large outlier amoung major leaguers.

ISO also tends to correlate very strongly with HR/FB, which is another less common measure of raw power for hitters. HR/FB is what it sounds like, its the rate of Fly Balls a hitter hits that turn into Home Runs. The major league average is typically around 9-10%, with 5% being poor and 14% being very good. HR/FB tends to be more stable in smaller sample sizes because it doesn’t include extra base hits on balls in play (doubles and triples), but at the same time, leaving doubles and triples out of the measurement entirely can mean it doesn’t accurately capture gap hitters, though this is less common than one might think intuitively. Also, HR/FB is essentially useless for minor leaguers, since there is no baseline definition for FB, so instead you may sometimes see HR/BIA used (BIA=Ball In Air), which thought a bit noisier thanks to the inclusion of line drives, is also a bit more objective in nature, since its much harder to mix up a ground ball from a non-ground ball. I’m not sure what the HR/BIA league averages tend to be, but educated guess would be around 6% for major leaguers.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 4, 2026 2:27 PM EST up reply actions   2 recs

Thank you to you and Valencia for these responses

I always was a little confused about what is a good ISO number.

Big Sexy

Follow KBR and Dewey on Twitter! @KBRandDewey

by King Billy Royal on Jan 4, 2026 3:41 PM EST up reply actions  

My pleasure

ISO more than a lot of other stats can vary from position to position. You won’t see too many MIs who have above average ISOs, so a .120 ISO for a second baseman is respectable even though its below league average. On the other hand, a .150 ISO is pretty weak for a first baseman. You’d really rather see a first baseman up in the .180+ range. Of course, because power tends to develop late, you have to scale all this back in the minors (except the PCL, lol, an extreme hitter friendly Triple-A league will tend to have high ISOs.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 5, 2026 10:03 AM EST up reply actions  

if you want a short and dirty explanation

ISO is basically the number that stands for power. SLG-BA is how it’s measured if you’re trying to eyeball it on B-Ref.

Anything about .200 ISO is great, above .160 is good, and below that it starts getting bad.

by valencia on Jan 4, 2026 2:59 PM EST up reply actions  

Fangraphs is our friend

http://www.fangraphs.com/library/index.php/offense/

They have a glossary on their site that explains all those new fangled numbers. Google is also a powerful search engine.

by Boxkutter on Jan 4, 2026 1:39 PM EST reply actions  

Pretty weak snark.

Even after adjusting for park effects. Weak.

(For Reply Fail, see below).

by None Taken on Jan 4, 2026 7:39 PM EST up reply actions   4 recs

Pretty weak snark.

Even after adjusting for park effects. Weak.

by None Taken on Jan 4, 2026 7:38 PM EST reply actions  

For people who say I don't know what is good, look at the percentile charts at fangraphs in the library

first look at a stat you’re farmilar with like ERA, since you know what’s good you know what 75%, 90%, whatever are like, then look at the stat you want to know’s percentile chart.

by Bososx13 on Jan 4, 2026 8:19 PM EST reply actions  

I don't like the fact that WAR uses UZR 150 (which has its own flaws) so much in its formula

I mean a great Catcher for example maybe say Buster Posey or Briann McCann is always going to be a tad/ smidge undervalued in WAR because catchers UZR is a bit oFF, and i think its clear.

I also don’t care for the Fangraphs $ figures…. of saying so and so was worth 16 Million last season and so and so was worth 24 Million last season…

To me, they’re taking that whole entire product and basing it on as if all 30 Teams had an even $100 Million payroll WHICH just simply isn’t true for teams like the Marlins, Rays, Athletics and others.

So for a team like the Red Sox or Yankees in 2002 Johnny Damon probably was worth at least $ 7.5 - 7.75 MM; BUT to the A’s at that time he couldn’t have been TRUELY worht more than 6 or 7 Million per year.

Although, sometimes players are currently underpaid and its obvious like Prince Fielder in 2011, I just don’t think it means a team, any team for that matter, should go out and spend 95% of what Fangraphs says he was worth in 2011 on him per year on a Long-term contract. Just like in 2011, Prince Fielder had about a 5.5 WAR THUS Fangraphs says he’s worth 25 Million dollars a year. (even just that year is ridculous to me, because thats for the Yankees, Red Sox, Mets, Angels, Tigers teams payrolls for my factor)

Also Fangraphs believed Fielders’ Value was almost $ 29 Million in 2009, and while that was his career year to date, i think its a bit luny.

With this said, my favorite stats or probably WPA+ , ISO% , BB % , K % for hitters.

and K per 9 , K per BB , BB per 9 and GB % , Hrs allowed for pitchers are a few i still like

Yoenis Cespedes
http://www.youtube.com/watch?v=aW9ge8l3jY8

by SteveHoffmanSlowey on Jan 5, 2026 1:26 AM EST reply actions  

i meant xFIP for pitchers sorry , also SIERRA and WRC + are ok...

Yoenis Cespedes
http://www.youtube.com/watch?v=aW9ge8l3jY8

by SteveHoffmanSlowey on Jan 5, 2026 1:32 AM EST up reply actions  

Couple of clarifications.

Fangraphs WAR doesn’t use UZR150 in the calculations, just the actual UZR. UZR150 is the UZR rating normalized (expanded or contracted) to what a player’s UZR would be over a 150 game period. So, if the UZR is 25 or so and the player participated in 162 games, their UZR150 would be something like 22.5. If the UZR was 0.1 for 50 games, it’d be about 0.3 for UZR150.

But, again, fWAR just uses the actual UZR number, which is appropriate for the other stats, which are a reflection of play time. wRAA is based on wOBA but also games played.

I mean a great Catcher for example maybe say Buster Posey or Briann McCann is always going to be a tad/ smidge undervalued in WAR because catchers UZR is a bit oFF, and i think its clear.

Currently, there is no catcher based UZR at all. Catchers are actually graded based on a stat they call rSB. Here’s Dave Appelman to explain it:

http://www.fangraphs.com/blogs/index.php/catcher-defense-in-war/

Posey did get a knock on him for UZR because he played first base a few times. You’ll note that McCann’s “fielding” line in the WAR section of his Fangraphs stat page is the exact same number as his rSB.

I also don’t care for the Fangraphs $ figures…. of saying so and so was worth 16 Million last season and so and so was worth 24 Million last season… To me, they’re taking that whole entire product and basing it on as if all 30 Teams had an even $100 Million payroll WHICH just simply isn’t true for teams like the Marlins, Rays, Athletics and others.

You’re misunderstanding the purpose behind that. The idea of providing dollar value doesn’t mean player X should get paid Y amount based on their WAR, it simply means that this is how many dollars worth of WAR they produced. It is a way to show how some players have produced either better or worse than their actual paycheck gave them. So if, say, Adrian Beltre is being paid $14 million a year, but puts up 5 WAR, he’s putting up the same value as what would cost a team $20 million normally. It is based on what teams pay, based on that year’s transactions, for every win above replacement. Thus, it fluctuates year in and year out as teams get more spendy or more frugal. Thus, the real purpose is more to show ‘excess’ value or loss based on the player vs. the league. It really doesn’t serve any other purpose and does not at all say that soandso should get paid what they produced in value.

“Also Fangraphs believed Fielders’ Value was almost $ 29 Million in 2009, and while that was his career year to date, i think its a bit luny.”

It was based on his WAR x what teams payed for 1 WAR’s worth of value. Fangraphs has nothing to do with that. The numbers are reflections of the league’s spending practices.

“and K per 9 , K per BB , BB per 9 and GB % , Hrs allowed for pitchers are a few i still like”

Starting to find out that K% (strikeouts as a percentage of batters faced) and BB% (same) is becoming more revealing as a measure of a player’s skill than K/9 or BB/9. But the difference is only noticeable for a few players.

Fans are typically idiots.

by The Typical Idiot Fan on Jan 5, 2026 3:25 AM EST up reply actions  

This
Starting to find out that K% (strikeouts as a percentage of batters faced) and BB% (same) is becoming more revealing as a measure of a player’s skill than K/9 or BB/9. But the difference is only noticeable for a few players.

As you say, in most cases the difference isn’t a huge deal, but K% and BB% will give you a much more accurate representation. There are two different problems with the /9 numbers, and in both cases its a denominator issue, but a slightly different problems in both cases.

With K/9, the problem is that the denominator of “per 27 outs” essentially breaks down to simply strikeouts per out. It tells you what rate of outs a pitcher gets via the strikeout, not how many batters he strikes out as a function of the number of batters he faces. This is important, because BABIP can have a large influence on on K/9. A larger percentage of a pitchers outs will come via the strikeout when more of the balls batters are putting into play are falling for hits, so a pitcher with a high BABIP may have an artificially high K/9, and the opposite is true as well—a pitcher with a low BABIP will have a low K/9. This isn’t true with K%, since K% is simply the number of strikeouts per batter faced.

With BB/9, the problem is that we’re essentially talking about walks per out. This isn’t an inherently meaningless stat, but you do a) have the above problem in this case as well, but also b) the numerator isn’t a subset of the denominator, and while this somewhat mitigates the BABIP effect, it also means this really isn’t a true rate as much as a ratio of two separate sets (More like K/BB). With BB%, the numerator IS a subset of the denominator, so again, you get a true rate of how many instances of a subset occur within a superset.

The problem with BB/9 is the same as HR/9, though since HR happen much less frequently, the problem isn’t as significant.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 5, 2026 11:58 AM EST up reply actions  

Taylor Jungmann

Sorry if you don’t want to talk abouit specific examples but, I was thinking of this when people were comparing taylor Jungmann and Sonny Gray here on our prospect ratings. Jungmann did evertything better than Gray except strike out people (per/9 innuings). I don’t have that data but perhaps Jungmann DID k more as well, (per PA). He must have faced quuite a few less batters.

"Does it make your life easier to just throw a quick, racist term at somebody? A man who has seen the things I’ve seen… experienced the loss and pain that I’ve experienced… I transcend race, hombre." - Kenny Powers

by casejud on Jan 5, 2026 12:09 PM EST up reply actions  

Unlikely he struck out MORE

I can’t find TBF figures for their college numbers, but the BABIP effect would have to be extreme to reconcile a K/9 difference between Jungmann’s 8.04 and Gray’s 9.43. But the fact that Jungmann had a considerably lower H/9 despite the lower K/9 does suggest there’s a good deal of BABIP difference here, so at the very least its probably quite a bit closer than the K/9 suggests.

Josh Beckett is a good example of the effect over the last few years.

2009 .290 BABIP, 8.43 K/9, 22.5 K%, 8.4 H/9
2010 .338 BABIP, 8.18 K/9, 20.1 K%, 10.6 H/9
2011 .245 BABIP, 8.16 K/9, 22.8 K%, 6.8 H/9

So of the last three years, Beckett had the lowest K/9 in 2011 despite the highest K%, because his BABIP was so low. His lowest K% came in 2010, but because his BABIP was so high, his K/9 was actually slightly higher than 2011. His K% were pretty close in 2009 and 2011, but his K/9 was quite a bit higher in 2009 because of the BABIP difference. The BABIP was most nominal in 2009, so since the K% were close, we would expect with a more typical BABIP, Beckett’s K/9 would probably have been closer to 8.5 in 2011, and closer to 7.9 or 8.0 in 2010.

The H/9 can also give you a frame of reference with Jungmann and Gray. Just based on their K/9 and H/9 figures, my best guess is that there’s maybe about a 2.5% difference in K%, while if I was just looking at the K/9 I’d imagine it’d be more like 5%, but that’s just a guess.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 5, 2026 1:36 PM EST up reply actions  

"K% and BB% will give you a much more accurate representation."

Yes. Started using this a couple of years ago when I created my First Inning account and found it more indicative.

by charles wallace on Jan 5, 2026 5:03 PM EST up reply actions  

Typical Id Fan

Thanks for this, you definately clear a thing or two up for me.

However please understand me when i say so many people out their in the “non-mainstream media” Bloggers, Fan Chats, and the ‘new world’ age baseball writers or even just the more than casual fan takes it that way.

They think if Fangraphs has player x at a value of y . Then said player x is worth close to or at least y money.
I see it written all the time, people tend to take what a Coco Crisp contract 14 MM for 2 years and compare it to the Fangraphs Value menchant, To see for themselves if it was a reasonable deal.

You see it all the time in the media I and I think a lot of us follow.

Its frustrating

Yoenis Cespedes
http://www.youtube.com/watch?v=aW9ge8l3jY8

by SteveHoffmanSlowey on Jan 5, 2026 10:55 PM EST up reply actions  

what those people think isn't entirely incorrect though

i’m repeating somewhat what TTIF said above

the $ values are attempting to measure what the free market would dictate as the player’s value. so for instance, say in 2011, teams paid for $4.6 million per WAR in free agency, and in 2011, Coco Crisp was worth 2.2 WAR. Thus, in 2011, Crisp’s on-field production was worth $10.1 million (if you believe the 2.2 WAR and $4.6 million per WAR to be correct).

another way of thinking about it, is in a theoretical world where all the GM’s in baseball knew that Coco Crisp was going to produce the way he did during the 2011 season, then each and everyone of those GM’s (in need of an outfielder) in January 2011 prior to the season, would have offered Coco Crisp a one-year $10.1 million contract.

there are of course three key questions:

1) Is the WAR value of Crisp’s production accurate? There is a lot of debate about this obviously (especially in how defense is measured) but would probably not be useful to go over extensively again here.

2) Is the $4.6 million per WAR accurate? The biggest issue I have with the WAR pricing, is that it doesn’t really take into account increasing marginal cost. Presumably, in terms of market price, the difference between a 5 WAR player and 6 WAR player should not be the same as the difference between a 2 WAR player and 3 WAR player. There’s value in concentrating more WAR in one single player, than spreading them out equally amongst different players, and I don’t think $4.6 million per WAR accurately reflects that value.

3) Assuming 1 and 2 are accurate, how much does knowing the market value of Crisp’s production in 2011 inform you of the reasonableness of a deal he signs in 2012? This then ties back into how well any metric measuring the past can predict future performance.

by blue bulldog on Jan 6, 2026 1:44 AM EST up reply actions  

I think point 2 is the most important

You could sign a 25-man team of one WAR players for probably $25-35 million, and that team would end up about a 66-70 win team, but according to Fangraphs, the actual value of those 25 WsAR would be $115 million at $4.6 million per WAR, which is crazy.

by cookiedabookie on Jan 6, 2026 9:50 AM EST up reply actions  

Well the issue is that no team is built strictly on the open market

If all players began their careers as free agents, the average cost of a market value win would be considerably less than it is now.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 6, 2026 12:18 PM EST up reply actions  

I understand that

It’s just the theoretical aspect of the fWAR to $ doesn’t completely hold up, and to me it has more to do with treating each WAR created equally and not giving them any marginal value.

by cookiedabookie on Jan 6, 2026 12:24 PM EST up reply actions  

I mean, its a measurement of market estimates

Not a strict translation of dollars-to-wins. Perfect projections would be required for such a conversion. As for the issue of increasing marginal value, I’m not sure the theoretical example of a 100% market-bought team really says much about that, but either way my thoughts on the idea are below.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 6, 2026 12:33 PM EST up reply actions  

There's been a lot of debate about the increasing marginal value of a win

I get both sides of the argument, but I’m more on the side that its not particularly significant. The way its often put is that one 6 WAR player is more valuable than two 3 WAR players because of roster limitations. The thing is, this is balanced by budgetary limitations. If all teams had infinite financial resources, this would be true, but they don’t, so if two 3 WAR players cost the same as one 6 WAR player, what you’re really saying is that two 3 WAR players cost the same as one 6 WAR player AND one 0 WAR (or replacement level) player. Are two 3 WAR players less valuable than one 6 WAR and one 0 WAR? Probably a bit, because teams can typically pay someone in their system as a 0 WAR player when that player has the potential to be better. But this brings up the flipside of the issue—risk. Investing the cost of 6 WAR into one player is also investing more risk into fewer assets—there is much more potential for that 6 WAR player to be worth significantly less than 6 WAR than there is potential for two 3 WAR players to be worth significantly less than 6 WAR. So in the end, I think it roughly evens out, and the rather than one method being strictly more efficient than the other, its more a question of risk tolerance. It makes sense for teams on extreme ends of the resource spectrums—teams that can either afford to have a risk blow up in their face and make up for it after, or teams that have so few resources that they rely on high-risk payoffs for success—to concentrate their resources into fewer roster spots. On the other hand, for teams in the middle, where big risk blowups can drastically set them back and there’s the possibility for success without high risk payoffs, it makes more sense to diversify their resources and spread their risk out as much as possible.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 6, 2026 12:29 PM EST up reply actions  

i think these are valid points

and i definitely didn’t consider the effects of varying risk profiles enough, as well as the possibility of saturated financial resources

the question i have though, is that while i think your argument does address the demand side of the issue, what about the supply side? shouldn’t decreasing supply of high-end WAR players skew the marginal cost of WAR higher? does it depend on supply decreasing linearly or exponentiallY?

my impression has been that the WAR values were calculated such that essentially in a given year say $65 million dollars were spent in free agency in total, and 15 WAR was generated out of that money. then the marginal cost of WAR is $4.33 million per WAR. it still seems logical to me that if this is the calculation method, more likely you have a situation something like guys who are worth 1 WAR are worth $3 million, guys who are worth 2 WAR are worth $7 million, guys who are worth 3 WAR are worth $12 million, guys who are worth 4 WAR are worth $18 million, guys who are worth 5 WAR are worth $25 million. this is still a total of 15 WAR and $65 million, but with increasing marginal costs.

the other thing is that i think the market value of free agents is affected by the fact that baseball operates not in a free competitive market, but more like an auction

by blue bulldog on Jan 6, 2026 3:27 PM EST up reply actions  

I'm certainly not an econ expert, but I believe I understand most of your points

And they make sense.

To address the supply issue in particular, I think I can partially answer that question by expanding a bit on its relationship to the demand issue: The thing about high value assets is that, yes, the higher the value of the asset, the lower the supply, but there is also an decrease in demand at that level as well. In a given offseason, not every team is going to be looking to add a 5-6 WAR player at the market rate—there have to be certain conditions: adequate financial resources and competitive potential of the team in the short term (or lack thereof) are the two big ones that come to mind On the other hand, virtually every team will consider adding a 1-2 WAR player at market value. It won’t necessarily limit a team that has a grim short-term competitive projection, and many more teams will have the financial resources to get involved, while teams that do have financial resources and/or look to be competitive in the short term will still be involved on players like that. So in a weird, counter-intuitive sense, there’s actually a higher volume of demand for players at that level. Because of the lack of supply, demand at the higher value levels will be more competitive, but there will be less actual volume, so the two issues offset to a degree. I think this all speaks to your last sentence there about how MLB free agency functions more like an auction than a competitive market.

The other piece of this that I think gets overlooked sometimes is that there’s a somewhat loose but very significant linear relationship between the AAV of player contracts and the length of those contracts. The increasing marginal cost does manifest, but rather than in increasing AAV dollar figures, it manifests in the length of contracts teams are willing to commit to high value assets. Players paid as 1-2 WAR players don’t get six year deals, and players paid as 6 WAR players don’t get one or two year deals. This difficult to convert into strict economic figures as fangraphs attempts to do, but it is nonetheless probably a result of the very observations you’re making and a fairly unique and highly significant aspect of this type of market.

"All energy flows according to the whims of the great magnet

What a fool I was to defy him"

-HST

by Mark Himmelstein on Jan 6, 2026 4:33 PM EST up reply actions  

If I may then...

Hate the players, not the game. If those people are using stats wrong, then they’re the ones who are the problem, are they not?

Stats are easily abused by people who want them to mean what they want them to mean.

Fans are typically idiots.

by The Typical Idiot Fan on Jan 6, 2026 1:56 PM EST up reply actions  

UZR being flowed into any metric is a near fatal flaw in my opinion

Even the data beyond one year samples is unreliable at best & beyond polluted. In my opinion which was mostly shaped by reading the more fleshed out opinions of others btw.

I think the rationale behind wanting to rely on WAR is probably well meaning, but there are just so many common sense dicrepancies over even the past couple of years that are hard to look past even from an ‘offense only’ perspective. I agree with many of the sentiments above & in the past which seem to forward the idea that offensive WAR can be viewed as one piece when attempting to evaluate guys, but it’s far from an all encompassing &/or complete tool.

by Matt0330 on Jan 5, 2026 11:01 AM EST up reply actions  

I agree with all of this

. . epecuially the UZR part and, even the less popular "even from an “offense only” perspective" part.

"Does it make your life easier to just throw a quick, racist term at somebody? A man who has seen the things I’ve seen… experienced the loss and pain that I’ve experienced… I transcend race, hombre." - Kenny Powers

by casejud on Jan 5, 2026 12:13 PM EST up reply actions  

This.

Seth Smith is a good example.

by mr. maniac on Jan 5, 2026 12:18 PM EST up reply actions  

Why?

I’m too lazy to look him up right now.

"Does it make your life easier to just throw a quick, racist term at somebody? A man who has seen the things I’ve seen… experienced the loss and pain that I’ve experienced… I transcend race, hombre." - Kenny Powers

by casejud on Jan 5, 2026 12:20 PM EST up reply actions  

Here's the thing...

The offense numbers, positional adjustment, and replacement level are all still valid components. You can quibble about the baserunning component if you want, because it’s relatively new and needs scrutiny, but when it comes to the defensive portion, substitute your own number or a number reflected by your preferred defensive metric.

So, say you like +/- or DRS instead of UZR, remove the UZR, plunk one of those numbers in, and recalculate. So if, say, Derek Jeter goes from a 10 UZR defender to a 0 +/ or even a 2 DRS, he moves up nearly a full win by avoiding a defensive crackdown by a single UZR flub.

I find that UZR works most of the time. It’s just those oddities that spring up every now and again that make it look bad. Someone mentioned Seth Smith. Not sure that’s a good example, but other Rockies like Dexter Fowler get nailed, probably because (and MGL admitted this) there might be a flaw with the way that certain parks’ effects skew the numbers for certain positions. In that case, either ignore it entirely and add the negative back in, or just replace it with what you think they’re worth. If you think they’re a +5 defender, remove whatever UZR says and replace it with 5 runs. It’s not that hard.

You shouldn’t kill the rest of WAR just because of one component. The rest of it works really well and corresponds really well to league results.

Fans are typically idiots.

by The Typical Idiot Fan on Jan 5, 2026 3:44 PM EST up reply actions  

God damn, that came out ugly.

“That strike-through should be a negative 10 UZR defender to a 0 defender according to plus / minus or a positive 2 defender based on DRS.” Or something to that effect.

Fans are typically idiots.

by The Typical Idiot Fan on Jan 5, 2026 3:46 PM EST up reply actions  

this is my biggest beef

with people who criticize WAR because of the defensive component as well

if you really don’t like the defensive component, then just substitute your own measure as an estimate. for instance, i basically just box players into -10 guys, or -5 guys, or 0 guys, etc.

if the conceptual framework is fine, then don’t discard the framework just because one component doesn’t work the way you want it to. the whole purpose of the framework is merely to create a stronger confidence level or narrower range of what a baseball player is actually doing.

by blue bulldog on Jan 6, 2026 1:48 AM EST up reply actions  

David Appelman found this wasn't true

and that it fluctuated just as much as wOBA. Just search google for fangraphs UZR 2008-2009

by Bososx13 on Jan 5, 2026 4:58 PM EST up reply actions  

Not a typical idiot fan

response, thanks for the clarifications!! i really appreciated the catcher breakdowns related to UZR or the lack of it, I was under the misguided conception that they used that to compile WAR for them as well…

by jerzbravesboy24 on Jan 5, 2026 9:26 AM EST reply actions  

There are some extremely raw defensive stats out there now.

Throwing out would-be basestealers is relatively easy to figure out, but the rest of a catcher’s defense is really hard to decipher. We can also calculate passed balls, though the league sometimes mislabeled a passed ball as a wild pitch and vice versa. Other intangible aspects, such as pitch framing, pitch calling, and whatnot are harder to figure out, but Mike Fast did a heck of a study on it (here: http://www.baseballprospectus.com/article.php?articleid=15093) and progress is being made, so… maybe one day.

In the meantime, only thing we have to go on is the easy to empirically calculate stuff, so that’s what we do. Incidentally, catchers received one of the higher positional adjustments because of their importance, so they already benefit in WAR calculations that way. This is why a bad hitting catcher can still be worth a lot in WAR calculations and why the best hitting catchers get really high marks. Thus, maybe defensive calculations just aren’t necessary for a catcher. They’re important, we can just leave it at that.

Fans are typically idiots.

by The Typical Idiot Fan on Jan 5, 2026 3:53 PM EST up reply actions  

I've never really understood how the defensive metrics are calculated

Specifically, how are they different.

Take Troy Tulowitzki for example. Why is his career number for Rtot (Total Zone) a positive 62, his career number for DRS (Defensive Runs Saved) a positive 65, and his career number for UZR (Ultimate Zone Rating) only a positive 27.7? Looking at the year to year data, it seems as though Rtot and DRS consistently think he’s an outstanding fielder while UZR thinks he’s just above average. Is UZR underrating him or are both Rtot and DRS overrating him? What is causing this difference?

One of the reasons I’d like to know is because this can have a significant impact on WAR scores. For instance, in 2009 Rtot (which baseball reference uses for its WAR score) rated Tulowitzki at a +15, but UZR (which Fangraphs uses for it’s WAR score) rated him at only a +2.4. Looking at his final fWAR score for that season, I see it was 5.7 while using UZR for the defensive component. However, had Rtot been used, his WAR score for that season would have been 7.0, which seems like an enormous difference to me. Not saying that either one is correct, but I’d like to know where the truth lies.

Thanks for any help on this.

by RoxRule on Jan 5, 2026 10:20 AM EST reply actions  

This kind of thing is where most negativity towards UZR comes from.

UZR seems more unforgiving when it comes to defense, while it seems to overrate other guys incredibly high. What we’re finding out is that UZR’s biggest, and probably only, problem seems to lie with outfield defense. This is why guys like Gardner, Gutierrez, and others are posting ridiculously high UZRs. This isn’t completely whacked, since by just watching them you know that they’re really good at their jobs. The question is, though, is a CF that good at his job worth THAT MUCH MORE than a SS good at his job?

To expound, in 2011, according to Fangraphs, the highest UZR rating for a SS was 11.9 posted by Alexei Ramirez. Troy Tulowitzki was 7th on the list with his 7.3. So while you can say that Tulowitzki looks bad compared to DRS and TZ rating, compared to other SS, he’s right up there with the top of his peers. UZR isn’t saying he’s a bad defender at all, just that the numbers aren’t as gaudy as you’re going to get from some other metrics. One of the reasons that short stop defense is graded so highly against a bigger curve is the importance of the position. A good short stop, in conventional wisdom, has to be the best defender on the diamond. Thus, in theory, a really good short stop would make an amazing second or third baseman, and so on.

Meanwhile, the highest outfielder was Brett Gardner with a whopping 25.2 UZR rating, playing primarily in left field. The next highest OUTFIELDER is Ellsbury with 15.6 UZR playing primarily in center field. In fact, the next highest position player, PERIOD, behind Gardner is Dustin Pedroia at second base with a 17.9 WAR. I’m sure Gardner is good at his job, like I said above, but is he that good that he was worth over two-and-a-half wins with just his glove? Other defense metrics love him, too, so maybe there’s something to it. But that seems like a highly ridiculous number no matter how you slice it. And it certainly overrates a player who is, otherwise, a speed demon with a modest bat.

I think, inevitably what is going to happen is that MGL is going to redo the calculations to balance the outfield defense more in line with the infield defense. That might sound like punishing some players for their position, but there is something to be said for playing an “easier” spot on the diamond.

Fans are typically idiots.

by The Typical Idiot Fan on Jan 5, 2026 4:05 PM EST up reply actions  

i actually kinda disagree with this

UZR is still a measurement of runs saved

you would think that a middle infielder preventing singles is naturally going to generate fewer runs saved than an outfielder preventing doubles

then there’s the matter of chances. UZR is a counting stat, and outfielders should naturally have more chances than infielders. aside from there being four infielders (five if you include the pitcher) there are only three outfielders. and very few hitters hit over 50% GB (though of course some amount of line drives are in play for infielders as well).

finally, infield just seems like a harder area to generate value for runs saved compared to the outfield. closer distance to plate means that there’s less time for reaction. if reaction times between outfielders and infielders are even close to comparable, then an outfielder has a lot more time to do something with his athleticism than an infielder.

these three reasons are why, intuitively (for me anyway), an elite outfielder saves a lot more runs than an elite infielder.

by blue bulldog on Jan 6, 2026 1:54 AM EST up reply actions  

People don't understand total zone.

they say stuff like baseball-reference’s defensive stat is better than fangraphs’ because it X is above y. Total zone estimates where the ball is by telling if the pitcher is right handed or left handed, UZR gets where the ball is from BIS video scouts who watch every play more than once. Peter Jensen did a study that the most BIS data is off by is about 4 feet. DRS, does use the same data as UZR, but interprets it a different way. To know more about how they are similar and different read post 3 by MGL here after reading that, it seems that DRS is going to have some serious sample size issues.

by Bososx13 on Jan 5, 2026 5:02 PM EST up reply actions  

I don't understand,

FieldFX. I’ve heard that the public will likely never see it, but what i want to know is, do we know if teams use/trust it?

by AckAttack on Jan 6, 2026 12:54 AM EST reply actions  

i thought

HitFX and FieldFX weren’t fully developed yet?

by blue bulldog on Jan 6, 2026 1:55 AM EST up reply actions  

I must admit I am a little skeptical and/or biased with many of these stats

With regards to defensive metrics Im skeptical they have any value whatsoever. There are questions with the quality of data, and there are questions with the methodology. The methodology separates the batted ball data into types that combine both trajectory and velocity but these should not always be combined. For example a softly hit ball has low value in these systems but in reality a low trajectory softly hit ball is one of the hardest to catch. The system really needs to catch time as it’s the crucial data point, FieldFX should provide this. Balls in play in the IF especially needs this as the range is very small for an IF and so the speed of the ball makes a huge difference. There is also the major issue on what to do with a fielding shift but this issue will still be around when FieldFX arrives.

With regards to the offensive stats I don’t have the same skepticism. I think many of them are great for getting an idea of players that you have not seen an awful lot. If I have seen a player a lot then I prefer component stats. Each of the composite stats makes some assumptions and are not always correct. For example we all know what Petco does to a hitter and so a factor is applied to get the wRC. But some hitters defy the standards. Gonzalez was supposed to be affected as a LH hitter and he probably was a little. But the factors don’t account for he is an opposite field power hitter and ghat park mostly supresses pull power.
For babip people look at really high or low numbers and say the number will revert to the mean. But it’s kind of obvious things like popups and speed are not a factor of luck.

I think there is a lot of value in these stats. But when possible I prefer the components and the scouting combined to form an opinion.

by pedrophile on Jan 6, 2026 7:58 PM EST reply actions  

I would definitely be skeptical of the baseball-reference defensive stats or Sean Smith, but that's where people see the stats mostly,

total zone estimates where the ball goes by if the pitcher is left or right handed. UZR, however uses baseball info solutions hit location data, some people did some studies that it’s usually within 4 feet of the actual distance. David Appelman showed it to be as reliable as wOBA year to year.

by Bososx13 on Jan 6, 2026 9:18 PM EST up reply actions  

and what do they do for velocity and/or trajectory? That is the biggest problem

since without velocity & trajectory we don’t have a time component.

by pedrophile on Jan 6, 2026 9:45 PM EST up reply actions  

BIS supplies MGL with that

so UZR and DRS adjust for that, TZ dosen’t

by Bososx13 on Jan 7, 2026 2:09 PM EST up reply actions  

I think they have angle

you can also get free angle data from Jeff Zimmerman’s site

by Bososx13 on Jan 7, 2026 10:26 PM EST up reply actions  

they don't

we can’t get that data, if we get field fx we will be able to, but unfortunately the teams haven’t released it to the public.

by Bososx13 on Jan 8, 2026 7:32 AM EST up reply actions  

Almost every play is a shift at some point. There is no such thing as a defined defensive placement.

A CF playing more towards RF for a lefty hitter is a shift.

If the hitter hits a normally catchable ball to left center that isn’t caught, UZR penalizes the outfielder because he was in a “shift”.

UZR is a really, really bad stat.

by Kelsdad on Jan 8, 2026 5:57 PM EST up reply actions  

you're assuming

that 1) shifts don’t normalize in the long run and 2) some fielders aren’t better at figuring out when/how much to shift than others

by blue bulldog on Jan 8, 2026 8:32 PM EST up reply actions  

yes, fielder positioning

is a skill for the fielders.

by Bososx13 on Jan 8, 2026 9:29 PM EST up reply actions  

I think total zone uses more detailed information when it's available

From Sean Smith’s explanation:

For most games, I have information on which fielder makes each out, and the batted ball type. Without information on the hits, I have to make an estimate. I look at each batter’s career rates of outs by position. For example, if 30% of a batter’s outs are hit to shortstop, then every time that batter gets a hit the shortstop is charged 0.3 hits. Repeat for every position. I look at batting against righthanded and lefthanded pitching separately, as switch hitters will have very different ball in play distributions depending on which side of the plate they hit from….

The second method is used when hits are coded with a batted ball type and we know who fielded each. The responsibility for ground ball singles hit to left field is split between the third baseman and shortstop, for center field it is between the shortstop and second baseman, and for right field, the second and first basemen. Groundball extra base hits are charged to the first or third baseman. Outfielders are charged with line drive and fly ball hits that they field.

Not actually affiliated with whygavs.

by WHYG Zane Smith on Jan 7, 2026 1:08 PM EST up reply actions  

that's TZL

fangraphs carries that, baseball reference just uses the basic TZ with estimates in their WAR.

by Bososx13 on Jan 7, 2026 2:08 PM EST up reply actions  

ah thanks

Not actually affiliated with whygavs.

by WHYG Zane Smith on Jan 7, 2026 10:11 PM EST up reply actions  


User Tools

Minor League Ball: Where the Future of Baseball is Discussed

FanPosts

Community blog posts and discussion.

Recommended FanPosts

Small
Post your fantasy baseball want ads here (please rec)!
Diesel1_small
Happy Birthday John Sickels!
Rich_tuning_small_small
Rich Wilson's Top 100 Prospects (#26-#50)
Diesel1_small
Who else gets confused by all the new stats?
March2111_084_small
Reminder: Treat other Posters with Respect

Recent FanPosts

Dsc00490_-_version_2_small
Fantasy Baseball Chat Over at Fake Teams Monday at 9 PM EST
Small
Community Positional Prospect #45
Small
Overall Community Prospect #61
Small
Community Pitching Prospect #42
Small
Dynasty Trade - Top Prospects
Gorilla_small
BP Oakland Athletics top 11
Small
Deep League, High-Impact Prospects: Friends, I love this Site and would love the community's advice!
Small
John Sickels Ebook - Customer Service?
Img00006-20101226-1702_small
FakeTeams Fantasy Baseball All Questions Answered Thread
Small
Fantasy ?- who to draft

+ New FanPost All FanPosts >

Baseball Nation Recent Stories

Photo

It's The Second Wild Card, Stupid

He shouldn't say that about Belgium. (AP Photo/Paul Beaty, File)

ProGS - The Pronunciation Guide For Sabermetricians

OAKLAND, CA - Howie Kendrick #47 of the Los Angeles Angels of Anaheim watches the flight of his ball hitting his second two-run home run of the game against the Oakland Athletics in the ninth inning at O.co Coliseum.  (Photo by Thearon W. Henderson/Getty Images)

Report: Angels, Howie Kendrick Agree To Extension

More from Baseball Nation >


Managers

March2111_084_small John Sickels

Jeri_avatar_small mssickels

Authors

Small SethSpeaks

Osnation2_small Jordan Tuwiner

Img00006-20101226-1702_small Ray Guilfoyle

Headshot_small dougdirt

Lax-xl_small Marisa Ingemi

Mlbbonusbaby-xl_small Matt Garrioch

Small Marc Hulet

Moderators

Small mrkupe


Site Meter