A Visual Look at wOBA

If you’re any sort of saberist, you should already know that Weighted On-Base Average (wOBA) is vastly superior to On-Base Plus Slugging (OPS) at measuring offensive value. While OPS is a mishmash statistic, throwing together OBP and SLG for kicks and giggles, wOBA was created based on research on the historical run values of events. It weighs all the different aspects of hitting in proportion to their actual, real-life value to a team’s offense.

But how exactly do these two statistics differ in assigning value to events? See for yourself:

What you see in that chart is a representation of how much wOBA and OPS weigh each individual outcome. The wOBA coefficients are very easy to find and straightforward, but I had to take some shortcuts to come up with coefficient values for OPS. While straightforward in theory – the sum of OBP and SLG – OPS is actually a rather convoluted statistic. You want to try adding these two stats together?

Instead of tangling with all that, I took the shortcut of just assuming both statistics had the same denominator and calculated the coefficients that way. It’s good enough for an estimate, and it gets the point across in the visual. So this is another area that wOBA trumps OPS: simplicity.

As you can see from the visual, wOBA puts more stress on walks, hit by pitches, and singles, while OPS attaches a huge value to homeruns and triples.* Since OPS is calculated by adding OBP and SLG, many people believe it treats both power and on-base skills as equally important, but that’s simply not true. When you dig down into the actual values OPS attaches to each outcome, it still favors power hitters by a wide margin.

*OPS also ignores Reached On Errors (ROE), but these happen so infrequently it isn’t a huge concern.

Also, when you look at OPS like this, doesn’t it seem slightly ridiculous? How can we treat it as a serious statistic when its coefficients look like they were created by a third grader? I’ll stick with the one backed by research and history, thank you very much.

For more on wOBA, see its page in the FanGraphs Saber Library.




Print This Post



Steve is the editor-in-chief of DRaysBay and the keeper of the FanGraphs Library. You can follow him on Twitter at @steveslow.


111 Responses to “A Visual Look at wOBA”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. thomas says:

    I thought wOBA included stolen bases… am I incorrect?

    Vote -1 Vote +1

    • Cardinals says:

      Probably because the most common formula is :

      ((0.72 x NIBB) + (0.75 x HBP) + (0.90 x 1B) + (0.92 x RBOE) + (1.24 x 2B) + (1.56 x 3B) + (1.95 x HR) / PA

      Vote -1 Vote +1

    • CSJ says:

      The Fangraphs version of wOBA does include stolen bases, but I haven’t been able to find what coefficient they use. Also, these coefficients change yearly. I would really like to see these numbers.

      Vote -1 Vote +1

    • James says:

      In my opinion, an ideal overall offensive value metric should somehow factor in the difference (SB-CS). Otherwise, it seems wOBA still favors power hitters, albeit to a lesser extent than OPS. If you believe RBOE should be included, then it would also make sense to somehow credit base advancement due to erroneous pickoffs/throws.

      I am curious about the rationale for crediting RBOE and weighting RBOE and HBP greater than 1B and NIBB, respectively. Perhaps those are the empirical values based on the historical research, but logically, from a player evaluation standpoint, it doesn’t seem to make much sense, since the differences in value between 1B/RBOE and NIBB/HBP would seem to be attributable to events unrelated to the offensive player’s skill.

      Vote -1 Vote +1

      • It is indeed based on historical data. Not that it matters much — the league average effect of wOBA is somewhere <.005.

        Vote -1 Vote +1

      • Al Dimond says:

        As for stolen bases, I’m quite sure that any wOBA including SB also includes CS. In effect, it’s closer to (SB – 2*CS) than (SB – CS) — getting caught stealing seriously hurts offense.

        My guess for why RBOE is on average more valuable than 1B is that occasionally you reach second on an error. My guess for why HBP is on average more valuable than BB is that HBP occur pretty much randomly, whereas BB might be slightly more likely to result from “unintentional intentional walks”, in base/out states where pitchers rightly care more about limiting hits than walks. These are just pure guesses, though.

        Vote -1 Vote +1

      • JT Jordan says:

        “Perhaps those are the empirical values based on the historical research”

        Yup. NIBB are around .31 runs, HBP around .34, 1B .47 and ROE .51. Some LW equations track total walks and HBP; the overall run value is about .33.

        Vote -1 Vote +1

      • Yirmiyahu says:

        I don’t really understand why RoE is included at all. I understand that it has a quantifiable run value, but is it something that the hitter is responsible for to any significant degree? If we’re excluding IBB’s, why not exclude RoE’s?

        Also: if IBB’s are excluded from the numerator, but PA’s is still used for the denominator– aren’t we assuming that, but for the IBB, the player would’ve been out? Shouldn’t the denominator be (PA-IBB) ?

        Vote -1 Vote +1

      • Tangotiger says:

        IBB are excluded from the numerator AND denominator.

        As for ROE, asked and answered. Do a search for my name in this thread.

        Vote -1 Vote +1

      • Yirmiyahu says:

        Tango, you are a Golden God.

        It’s hugely appreciated when you answer questions in these threads.

        Vote -1 Vote +1

  2. Xeifrank says:

    Would be interesting to see how the visual changed if instead of using OBP+SLG you used (1.7 x OBP)+SLG or w/e the proper scale factor for OBP is. Nice post.
    vr, Xei

    Vote -1 Vote +1

    • True, that’s a shortcut for calculating wOBA…I’m assuming that it’d be pretty similar to wOBA’s values.

      Vote -1 Vote +1

    • craigtyle says:

      If you assume that about 10% of all PA’s are walks, you get:

      Walk, HBP: 1.7
      Single: 2.8 (i.e., 1.7+1.1)
      Double: 3.9
      Triple: 5.0
      Homerun: 6.1

      Note that the ratio of a walk to a single in wOBA is about .80; for 1.7OBP+SLG, it’s much lower –.61.

      Vote -1 Vote +1

  3. The Green Lantern says:

    Why is a HBP weighted slightly higher than a NIBB?

    Vote -1 Vote +1

    • Because if a pitcher hits a batter, in some cases that’s a sign that he’s wild and more likely to let up runs going forward. These values were calculated using historical records or games, so wild pitchers are simply ever-so-slightly more likely to lead to scoring.

      That’s the same reason Reached On Error (ROE) is slightly higher than a single. If someone is throwing a ball away, sometimes batters advance an extra base.

      Vote -1 Vote +1

    • Lee Panas says:

      Because walks are sometimes issued in situations that benefit the defensive team. For example, you often see unintentional “intentional walks” with a runner on second and a strong hitter up at the plate. HBP’s typically happen more randomly and are rarely issued to benefit the defensive team. Thus, the average walk is worth slightly more than the average HBP.

      Vote -1 Vote +1

  4. Rick says:

    I’m curious, what’s the correlation between OPS and wOBA?

    OPS may be quite contrived and not even close to what you’d come up with if you started from scratch. But considering that we inherited a world where batting average was king, getting to the point where we could just combine a few stats people knew and were comfortable with to come up with something much more accurate was a pretty major leap.

    Conceptually, it meets people where they’re at — counting both getting on base and hitting for power without really introducing anything new. So long as OBP and SLG are still in wide use, OPS is a simple way to get a good way towards a good estimate of overall offensive production.

    No doubt, wOBA is more accurate and OPS is biased towards SLG. But how much so? For the purposes of facilitating a more informed conversation among the non-sabermetric types, OPS is a great compromise. Perhaps 10 or 15 years from now, it can be retired (hopefully next to the save) and we can all stick with wOBA. But for now, let’s not let the perfect be the enemy of the good.

    Vote -1 Vote +1

    • I disagree…I think wOBA is incredibly simple to explain to even the most basic saber-newbie. Considering how simple it is conceptually and how much better it is, I’m not content just leaving things at OPS. It’s a good first step and early compromise, but it’s not like wOBA is all that more confusing.

      Vote -1 Vote +1

      • I’m way with Steve on this one.

        If we have a fan who knows nothing but batting average (which includes many fans), them wOBA is an easy next step. Proof: I explained it to my wife in less than 5 minutes. She respond, “That makes a lot of sense.”

        Q.E.D.

        Vote -1 Vote +1

      • test says:

        wOBA is way more confusing for anyone brought up on BA, SLG%, and sometimes OBP%. Yes, the actual formulae used to get to the pieces are odd for OPS, but it is easier to understand for those who follow baseball already. Any stat you can calculate in your head as a game goes on is going to be easier to use. ie. going 1/3, BB, 2B gives OBP = .5, SLG= .667, OPS 1.167. Easy (relatively), and took longer to type than to figure out. What’s the wOBA for this day? Anyone who says wOBA is easier to figure out is kidding themselves.

        I was struck on how similar the two are – wOBA is clearly NOT “vastly” superior. It’s incrementally better. That’s important, and worth worrying about when discussing finer points, but seriously, if regular fans use OPS to value players instead of BA, HR, RBI, that’s “vastly” superior.

        Going from 50% to 90% is more important thatn making it to 95%.

        Vote -1 Vote +1

      • cavebird says:

        I think you are analyzing too deeply to come up with the conclusion that OPS is confusing. It is a combination of two basic stats, and therefore a bit of a mishmash. But wOBA is much more complex in that there are coefficients (not just effective coeffecients) of .72 and the like, making it impossible to calculate simply without a computer or calculator. With OBP and SLG being easy to find everywhere nowadays, OPS is incredibly easy to calculate in one’s head.

        Obviously, wOBA is a more accurate statistical measurement of a hitter’s performance, but I don’t think we can say it remotely resembles perfect. Giving such a high weight to BROE, above NIBB and singles, bothers me. Yes, BROE may contribute more to scoring; but did the batter actually contribute to the result of the play? Do some players repeatedly reach more often on errors than others? Maybe so, but I don’t know of any data to support that. Perhaps, if, as I suspect, a batter contributes, but only partially, to BROE, it should remain, but have a lower coefficient, because while it contributes more to run scoring, the batter does not contribute as much to it. The inclusion of BROE also begs this question: why are intentional walks excluded? If those are considered out of the batters control (I assume they lead to scoring more runs), why aren’t BROE?

        Vote -1 Vote +1

      • For calculating in your head, sure OBP and OPS are superior (though SLG is a bit harder to calculate). But being able to calculate and being able to understand something are two wholly divergent issues.

        My wife knows nothing of OBP or SLG, so my first task would be to explain how these two metrics use different weights for the same events, thereby requiring to complete and separate definitions for the terms (i.e. a double equals 1 and it also equals 2).

        I’ll leave Tango (below) to respond to the RBOE argument, adding only: Yes, there is evidence supporting RBOE is a skill for some players. (Nonetheless, detractors of wOBA often over-predict the otherwise marginal effect of RBOE on a player’s wOBA).

        Vote -1 Vote +1

      • B N says:

        @Cavebird: Some batters definitely reach more than others on errors. Non-true outcome types are typically going to put a lot more balls in play. Since # errors are going to be proportional to # balls in play, you’d definitely expect contact hitters to get a lot more of them. So you won’t see Cust picking up lots of reached-on-errors. But you probably will see a guy like Pierre doing so.

        Likewise, speedy guys are going to induce more errors- since defensive players have to work faster to throw a guy out.

        It won’t necessarily be a huge impact, but it’s significant.

        Vote -1 Vote +1

      • Rick says:

        I agree, wOBA is not confusion on its own merits. But even a But I’m talking about the person who is still skeptical of sabermetrics altogether. They don’t have to use OPS for long, but I think you’ll find you get much better traction when OPS is the first (or second) step away from batting average and RBI.

        Step 1: Move from AVG & RBI to OBP and SLG (better stats for the same basic idea)
        Step 2: Move from OBP and SLG to OPS (combine the better stats to get a holistic view)
        Step 3: Move from OPS to wOBA (simply the craziness of what goes in to OPS by moving a similar metric that’s more accurate and similarly easy to understand)

        Sure, some people can jump straight to wOBA. And some people are already using OPS and ready to to take the leap. But a lot of people are still in AVG & RBI world and I think part of the reason they resist is because the sabermetric community has gotten too many steps ahead of them for them to make the logical leaps on their own.

        When I was started getting in to sabermetrics, I had the benefit of the next generation of stats being a clear evolution from where my understanding was, allowing me to test it out with one foot instead of having to jump in with both.

        We shouldn’t confuse objective simplicity with the reality that many people have a firmly rooted belief and, frankly, are not going to be purely objective on their own and won’t likely have a sabermetrician there to guide them.

        I’m glad that ESPN is starting to use the more advanced metrics, but they’re still struggling to find the right way to build the bridge from here to there. It’s often just a “here’s a new stat with funny letters” without any real explanation of why it makes more sense or how it relates to what people already know.

        Vote -1 Vote +1

      • cavebird says:

        @B.N. Okay, if we take as a given that batters have some control of BROE, I think it is still obvious that they don’t have total control. Would it not improve the stat if a coefficient was placed on the BROE portion to make only part of BROE credit go to the player–that portion of BROE that is within a player’s control?

        Vote -1 Vote +1

      • Lee Panas says:

        I think wOBA is easy to understand conceptually. If someone is interested only in the concept and not the calculation, then wOBA is as easy to explain as OPS. If you are dealing with someone who doesn’t trust stats and wants to know the math, then wOBA becomes complicated. I’m talking about the group of people that have some math skills but are skeptical of sabermetrics and not interested in learning linear weights theory. They can be a pain in the butt! :-)

        Then there’s the problem that a whole generation of fans have finally bought into OPS. And now we want to push wOBA on them. They want to know why can’t we make up our minds.

        So, I continue to use both depending on my audience.

        Vote -1 Vote +1

    • B N says:

      @cavebird: That would probably make sense if one was trying to have wOBA work more like FIP. However, from my understanding and reading these comments, it appears that the intention is more to measure the value of what happened (as opposed to the credit for what happened). In that way, it’s more like AVG. And clearly, for something like AVG, we don’t discount players for getting lucky BABIP or anything.

      While there’s probably merit to a stat that applies additional damping factors to each term in order to account for the ability of a player to control them, that would be a different stat entirely I think.

      Vote -1 Vote +1

  5. Tangotiger says:

    1. The coefficients for SB and CS are roughly 0.25 and -0.50. You use them if you need them.

    2. The data for all the events through 2008 is here:
    http://tangotiger.net/bdb/lwts_woba_for_bdb.txt

    The blog post that describes the details of wOBA, including SQL, is here:
    http://www.insidethebook.com/ee/index.php/site/article/woba_year_by_year_calculations/

    3. Hit batters are mostly random, regular walks are somewhat random, and intentional walks are not random. Hence, HBP are more likely to happen with a runner on 1B, a regular walk might happen with a runner on 1B, and IBB almost never occurs with a runner on 1B.

    4. When you reach base on error, you can reach 2B, 3B, or even home plate. A single where you stretch to 2B is actually a double (d’uh). And we don’t care about who was responsible. We are just counting “what happened”. Otherwise, we’ll be talking about a batter being very responsible for a K, mostly responsible for a BB, alot responsible for a HR, somewhat responsible for a 2B, and flip of the coin on 1B.

    That’s not the discussion wOBA is having, any more than OBP or SLG is having that discussion either. wOBA simply represents WHAT happened, without asking WHY.

    5. 1.7*OBP + SLG is going to be very close to wOBA.

    6. Correlation? Well, if most players are around average, then all correlation is going to do is tell you that, yup, a good wOBA means a good OPS. Again, d’uh. But, what does it mean for players at the extreme, with a lopsided OBP/SLG view? Nothing, because you can’t extrapolate.

    OPS is fine, if that’s all you have. But, don’t be too serious about it. Think of OPS as your f-buddy.

    Vote -1 Vote +1

    • You’re the bestest, Tango…thanks. That hits on everything.

      Vote -1 Vote +1

    • Telo says:

      Did Tango just say f— buddy? Awesome.

      Vote -1 Vote +1

    • cavebird says:

      Well, apparently, my question was answered before I posted (because I took so long writing mine). The key to my question was Point 4—it doesn’t ask why, just what. Is there a good stat for hitters limited more closely to things that the hitter controls? Also, if it is just what not why, why aren’t intentional walks considered?

      Vote -1 Vote +1

    • Azmanz says:

      Thanks for the clarification on some of this.

      Can you explain why HBP’s are more likely to happen with a man on first than walks?

      Vote -1 Vote +1

      • Tangotiger says:

        Walks are controlled by the pitcher far more than they control hit batters. And so, when they allow walks, those walks won’t happen as often with a runner on 1B, or with less than 2 outs.

        You can just look at the league splits, and look at how many walks are issued in the various base/out splits.

        Vote -1 Vote +1

        • Free Bryan LaHair says:

          then a HBP should be heavier with a man on third than on second and so on. separate coefficients based on what the situation the HBP takes place in would be necessary for saber-accuracy.

          beyond that, though, if the ultimate measure is “what” rather than “why” then none of it matters if a runner is on second when the pitcher hits the batter, and then sits the rest of the side down. you’re then measuring the possibility of offensive production that ultimately didn’t happen. just like if a leadoff batter hits a double or even a triple but still ends up getting left on base, those extra bases are worthless and didn’t contribute anything to offensive production. only the possibility of offensive production, and that certainly isn’t a “what” factor if it didn’t actually happen. unless, instead, we’re measuring the possibility of offense and not actual offense. possible and actual are too intertwined in these scenarios (and are identical in many cases) to separate them and measure only a portion while dismissing the other.

          Vote -1 Vote +1

      • Azmanz says:

        Ok, this seems intuitive.

        Vote -1 Vote +1

    • Rick says:

      If you’re a sabermetrician, you care about the tails. If you’re not, you’re probably going to be just fine with the stat you already know and which works pretty well for 90% of the league.

      I understand your point, Tango and I recognize that you’re not one who particularly cares about shepherding the masses. If this post is really just about getting sabermetric-friendly folk to move on from OPS, I’m completely on board.

      I just wouldn’t want people to lose sight that OPS represents an important logical step in saying “you need to look at both avoiding outs and acquiring bases” without further complicating things by introducing linear weights. But if people are willing to make the leap in to a committed relationship without hopping the sack first, more power to them.

      Vote -1 Vote +1

      • Tangotiger says:

        Right, no need to marry OPS just because it was your first. OPS does a good job, no doubt about it. It doesn’t do a great job.

        Vote -1 Vote +1

    • GTW says:

      I vote to change the title of this article to “OPS is your f-buddy.”

      Vote -1 Vote +1

    • intricatenick says:

      The two correlations that are of utmost importance here are

      Team OPS vs Team Runs

      Team wOBA vs Team Runs

      I think we know how that will work out.

      Vote -1 Vote +1

      • KC_Baseball says:

        You might be surprised…I did this a while ago for all AL teams between 2005-2010, and got these correlations:

        Team OPS to Team Runs: .888
        Team wOBA – Team Runs: .859

        I can’t explain why that would be, but it does make me think OPS is a great stat in spite of seeming like a bit of a hodge-podge mess.

        This article finds similar correlations, but argues that on the individual level wOBA is much more telling: http://www.thehardballtimes.com/main/article/the-great-run-estimator-shootout-part-2/

        I will take wOBA over OPS, but they are both great measures. Cramming together two stats with different denominators may seem messy, but if it correlates so well to runs scored, it is hard to argue with.

        Vote -1 Vote +1

  6. AJS says:

    I think part of what turns people off is the name: wOBA. How do you even say that? Why is the w small? Etc, etc.

    OPS (On-base Plus Slugging) is a very simple concept.

    Perhaps new branding for wOBA could help.

    Vote -1 Vote +1

    • Ben D says:

      I completely agree. I’ve stated this a number of times as well. Crappy names hurt the ability for a metric to become more mainstream.

      Vote -1 Vote +1

    • Lee Panas says:

      If you explain to people that it’s like OBP with appropriate weights (Thus, weighted on base average), then it becomes more intuitive. For example, you might say:

      “The wOBA statistic is like an on-base-percentage (OBP), except that it gives appropriate weights to different events. As you know, the OBP calculation counts every event where a batter reaches base (walk single, double, etc) the same. In contrast, wOBA gives a hitter more credit for a hit than a walk and more credit for doubles, triples and home runs than singles. The result is a rate statistic which measures a players total batting contribution.”

      Vote -1 Vote +1

  7. ccomp says:

    What did you use for the denominator of OBP and SLG in this analysis? It makes sense that OPS would favor power hitting and have depressed coefficients for BB and HBP because OBP has a bigger than denominator than SLG.

    In my thinking, the only real problem with OPS is that it is a simple sum of two raw numbers. It’s confounding, and if OBP and SLG were standardized, they could reasonably be summed and probably produce coefficients that look a lot like what you have here for wOBA.

    The pie charts are very interesting because they show that the relative areas of 1b, 2b, and HR are very similar in wOBA and OPS. However, wOBA takes a big chunk of area out of 3b and sticks it into BB and HBP. This is a very compelling difference and it intuitively makes a lot of sense – OPS is biased because SLG has a smaller denominator and therefore inflates the value of offensive production relative to OBP.

    In quantitative terms, this article doesn’t really make a case for how wOBA is better… it’s just a visualization of how wOBA and OPS are different.

    Vote -1 Vote +1

  8. Tangotiger says:

    Pronunciation is here:

    http://www.youtube.com/watch?v=maYnqbdo2jw&feature=related

    ***

    IBB: Actually it IS included, behind the scenes. The IBB coefficient is equal to THAT player’s wOBA. So, if Bonds is a .500 wOBA, the coefficient for IBB is .500. If Neifi Perez is a .250 wOBA, the coefficient for IBB is .250.

    Vote -1 Vote +1

    • Argh! I don’t like wuh-buh! I want more whoa-buh!

      Vote -1 Vote +1

      • sleepingcobra says:

        Seeing as how we’re talking about On Base Percentage – or Ahn Base Percentage – shouldn’t it be pronounced Wah-buh, and not Wuh-buh?

        By the way: WORST. PRONUNCIATION. FOR. A. STAT. EVER.

        Vote -1 Vote +1

      • NWS says:

        I support Woah-buh!

        Although “Weighted oh bee ay” is nice too.

        Vote -1 Vote +1

    • B N says:

      Ahhh, that makes sense now that you put it that way. I was kind of wondering how the stat would dance around the issue that your IBB is almost entirely determined by your threat level due to other aspects. I’m still a bit unclear as to how that is added in, however. Are we talking about something along the lines of:

      wOBA(with IBB) = wOBA(non IBB)(1 + IBB/PA)

      What is the denominator used for IBB, considering that they’ve been completely removed from the initial wOBA calculation from what I can see?

      Vote -1 Vote +1

      • Tangotiger says:

        You are thinking to hard about this. Let’s say Neifi has a .250 wOBA. That’s based on 100 in the numerator and 400 non-IBB PA in the denominator.

        Then I tell you he has 10 IBB. You put 2.5 in the numerator and 10 in the denominator. Guess what:

        100 + 2.5 divided by 400+10 = 0.250!

        See, that’s the beauty of it. You can set the coefficient of IBB to the production level of the player, and you don’t even need to worry about the formula!

        Vote -1 Vote +1

      • B N says:

        Hmm… in that case I was definitely thinking about it wrong. With that said, the system as you describe it now just seems weird, and appears to be completely equivalent to ignoring IBB entirely (except in the case where all you have is n IBB, in which case it seems to be flat-out undefined).

        Considering that you’ve described wOBA as an outcomes-based stat, and getting any sort of walk would seem to increase your run expectancy… shouldn’t IBB actually change the final wOBA in some way? Otherwise, what was the point of all the lovely discussions over if teams were helping or hurting themselves by walking Bonds all the time?

        Vote -1 Vote +1

    • Telo says:

      Wubba, Woh bah, double you oh bee ay – they all sound terribleawful.

      “Weighted On Base”. Try it, you’ll like it.

      Vote -1 Vote +1

  9. Barkey Walker says:

    This article actually convinced OPS is a really good stat. It is almost wOBA!

    If I think wOBA is king, the graph shows me that, for OPS, triples are overweighted (who cares, nobody hits a lot of them), and free passes to first are underweighted. But your coefficient is also too small for them. They do not count against the SLG, so they add one to the denominator, not two (like hits do). So, you should multiply their size by some factor close to two to blow them back up. Once you’ve done that, the graphs look pretty much the same.

    OPS is a great proxy for wOBA.

    Vote -1 Vote +1

    • Colin Wyers says:

      The trouble with the example above is that it holds if you calculate SLG as TB/PA rather than AB. If you calculate OPS that way, the relative weights of all the values hold constant as a player’s walk rate changes.

      Of course, in actual OPS, the relative weights of the event change with the player’s walk rate. That’s where the problems start to occur.

      Vote -1 Vote +1

    • Barkey Walker says:

      The right way to do this calculation is to calculate the partial derivative of each term evaluated at the league average. The weights change for the wOBA too, even though the denominator terms are constant because the partial of f(a,b) = a / (a+b) with respect to a changes with b.

      Because of this, wOBA has the same “problem” of weights changing as OPS.

      Vote -1 Vote +1

  10. Tangotiger says:

    “wOBA is way more confusing for anyone brought up on BA, SLG%, and sometimes OBP%.”

    And Celsius is confusing for those brought up on Fahrenheit. Teach Celsius in school, and Fahrenheit disappears in time. My kid is being taugh Fahrenheit this year, and I have to work with him every day about Celsius. Fahrenheit is a cult, plain and simple. So is batting average, to a large degree (no pun intended).

    ***

    “making it impossible to calculate simply without a computer or calculator. With OBP and SLG being easy to find everywhere nowadays, OPS is incredibly easy to calculate in one’s head.”

    You mean after someone ELSE uses his computer to calculate OBP and SLG for you (and that person remembering how to handle the SF… counts for OBP but not SLG), then you are telling me it’s easy for YOU to add those two numbers up?

    If that’s the test, then of course wOBA is going to fail.

    Vote -1 Vote +1

    • Nate says:

      While Celsius is perhaps more scientifically based that Fahrenheit, it’s only moderately less arbitrary. And Fahrenheit makes much more sense for its original purpose — 0-100 being the typically temperature range for mid-latitude regions.

      Vote -1 Vote +1

      • Nate says:

        If you want to really be minimal arbitrariness for temperature, use Kelvin.

        Vote -1 Vote +1

      • Tangotiger says:

        Sounds like the 0-100 Fahrenheit thing is completely made up.

        I agree about Kelvin being the correct scale. Celsius is like “Kelvin above replacement level”.

        +8 Vote -1 Vote +1

      • B N says:

        At least Celsius and Fahrenheit are base 10. What kind of an idiot decided to measure time in seconds/minutes/hours when standardization of units happened with the metric system? Whoever that is, I need to go kick some dirt on their grave.

        Vote -1 Vote +1

      • Tangotiger says:

        Babylonian astronomers seemed to have a good reason for it.

        At some point in the future, base-2, base-8, or base-16 would make more sense. When that happens, Skynet will have won.

        Vote -1 Vote +1

      • fredsbank says:

        i mean, obviously -273 should be baseline low for all measurements from here on out

        Vote -1 Vote +1

      • siggian says:

        “While Celsius is perhaps more scientifically based that Fahrenheit, it’s only moderately less arbitrary.”

        I’d say it’s a lot less arbitrary. 0 Celsius is the temperature at which water freezes and 100 Celsius is the temperature at which it boils. (Yes, I know there’s a little more to it than that, but there’s an actual definition of the scale and it’s tied to something concrete.)

        I’m a hybrid person. For some things, I use Imperial (feet, inches, pounds) and others I use metric (km). I learnt Imperial for things that a child would understand and then the conversion happen and I learned metric for what an adult understands (km for driving)

        Vote -1 Vote +1

    • craigtyle says:

      I wrote to Tango years ago imploring him to base wOBA on a batting average scale, since that’s what people are used to. I used my kids as an example.

      Fast forward a couple years — my kids are both in a fantasy baseball league, and in their discussions, they’re continually bringing up a batter’s OBP and wOBA.

      So, I guess it is possible to convert to Celsius.

      Vote -1 Vote +1

  11. Greg says:

    I blame OPS on the decline of corporal punishment in schools. A math teacher should be allowed if not obligated to beat the crap out of you, if you try adding two measurements that have different units.

    Vote -1 Vote +1

  12. craigtyle says:

    For those who find wOBA’s coefficients too complex or difficult to work with, David Smyth posted this simplified version on The Book blog:

    .36*[(2*(NIBB+HBP))+(1.5*H)+TB)]/PA

    It’s pretty close and very straightforward.

    Vote -1 Vote +1

  13. JT Jordan says:

    The marginal run values for OPS (set so that a single = .47):

    BB = .24
    1B = .47
    2B = .87
    3B = 1.27
    HR = 1.67

    versus the empirical ones:

    BB = .31
    1B = .47
    2B = .76
    3B = 1.02
    HR = 1.42

    OPS works on the team level because it’s a crude approximation of run scoring. On the individual level, however, it’s going to severely underrate players that walk and overrate those that rely on extra base hits.

    Vote -1 Vote +1

  14. Chris says:

    Wow, it’s holier than thou commentary like this that turns casual fans away from stats. Yes, wOBA is a better metric than OPS but seeing how simple OPS is to calculate, it’s pretty damn good for what it is: a summary metric. Both are nice, quick ways to look at the relative value of player.

    Vote -1 Vote +1

    • Norm says:

      How do you calculate it?
      “Add OBP and SLG”
      How do you calculate SLG and OBP?
      “I don’t have to, it’s already given to me”

      Well you don’t have to calculate wOBA either, because its given to you for free, right here on this wonderful site.

      +6 Vote -1 Vote +1

  15. Tangotiger says:

    One of the career leaders in reaching base on error, per PA, is Derek Jeter.

    Vote -1 Vote +1

  16. Ben D says:

    For the future, I hope triples is weighted less in a stat that attempts to be more of a projection stat than results-based.

    Vote -1 Vote +1

  17. Ben D says:

    In a perfect world, there would be a way to perfectly analyze who a better base runner is, and a positive weight would be added if a hitter hit into a fielder’s choice but replaced a worse base runner.

    Vote -1 Vote +1

  18. TK says:

    I don’t understand why IBB are ignored???

    Vote -1 Vote +1

    • T-Car says:

      People hate Barry Bonds

      Vote -1 Vote +1

    • JT Jordan says:

      Tango mentioned this above:

      “IBB: Actually it IS included, behind the scenes. The IBB coefficient is equal to THAT player’s wOBA. So, if Bonds is a .500 wOBA, the coefficient for IBB is .500. If Neifi Perez is a .250 wOBA, the coefficient for IBB is .250.”

      Vote -1 Vote +1

      • TK says:

        Yeah, I read that. I still don’t understand. I understand how that would in a semantical sense include IBB, but I don’t understand why IBB aren’t treated the same as other outcomes, with their coefficient being based on their anecdotal run creation value.

        Vote -1 Vote +1

      • Tangotiger says:

        IBB:

        wOBA was created for use in The Book. And if you read The Book, you would realize and accept that you have to remove IBB from consideration. The Book is about batter-pitcher matchups, and IBB don’t help us there.

        Now, you can argue that for what Fangraphs is doing that IBB should, by virtue of philosophy, be included in a more explicit and consistent fashion. That the IBB coefficient should be say 0.3333 for everyone, rather than .25 for some hitters, .32 for others, and .41 for others.

        What does it practically mean? Setting aside Bonds, let’s look at an extreme example. You have a hitter who has a wOBA of .400 (meaning say 240 in the numerator and 600 non-IBB PA in the denominator). This guy also has 30 IBB. If we give him a standard .3333 coefficient, he gets an extra 10 in the numerator and an extra 30 in the denominator. Now, we have 250 in the numerator (240+10) and 630 in the denominator (630 PA). And 250/630 is 0.397.

        Now, do we REALLY want to argue that an extreme hitter (.400 wOBA) with an extreme IBB profile (30 IBB) should be philosophically shown as a 0.397 wOBA instead of the 0.400 wOBA that I would show?

        I know I don’t want to be part of that argument.

        However, if this argument is solely about Bonds, then you will likely lose that one too, because Bonds’ IBB were not issued like everyone else. If they were to create two categories of IBB: normal IBB and scared-sh!tless IBB, then the former IBB would get say a coefficient of 0.3 and the latter would get a coefficient of 0.5. And so, a Bonds IBB would come in far higher than a regular IBB.

        Vote -1 Vote +1

      • TK says:

        I think wOBA gets a little too far into the weeds in this regard. I’m not a fan of this picking and choosing. Conceptually, I understand how IBB don’t really factor into a players ability, i.e. there is no skill involved in drawing an IBB. I also understand that a better player will be IBB in situations where a lesser player wouldn’t and the walk is more meaningful, so maybe it should be worth more, but you can make the same argument regarding a players speed and any hit he gets. A Jose Reyes single is a lot more valuable than a Brian McCann single. The holes in wOBA as a perfect indicator of offensive production are already present, so I’d rather it stick strictly to being outcome-based, even if players are “punished” for being IBB.

        Thanks, though, for your reply. It was helpful.

        Vote -1 Vote +1

      • Tangotiger says:

        “The holes in wOBA as a perfect indicator”

        You can’t indict wOBA for not being what it didn’t set out to be. wOBA was created for a specific purpose in mind, and it does that exactly the way it needed it.

        If you read The Book, and I had included IBB, your first reaction would have been: “Uh, why are you including IBB in there?”

        You evaluate a metric based on its intended purpose, and not on some “global” purpose, of which it can’t possibly hope to make sense in every case.

        OBP gives a value of “1″ for walks and HR. How does that make any sense? Well, it’s intended purpose is to count the percentage of times a player reaches base safely. So, it makes sense there.

        But, as I said, even if you include IBB with a coefficient of 0.333 for every player, to quench your philosophical thirst, the needle barely moves. Nothing is stopping anyone from creating wOBA that includes a coefficient of 0.333 for IBB for all players.

        And nothing is stopping anyone from counting “legged out doubles” and “sliding doubles” either, if they wanted to separate them.

        You start with the question, and you set up your constraints, and THEN you answer the question.

        You don’t look at the answer, and then say “well, it won’t answer this question, so the answer must be wrong”!

        Paris Hilton is as dumb as rocks. Well, she’s not supposed to be anything else. She does what she does well. Measure her on that basis.

        Vote -1 Vote +1

      • TK says:

        I haven’t read The Book, perhaps I should. But I was basing my comment on the author’s post, specifically this line:

        “wOBA was created based on research on the historical run values of events. It weighs all the different aspects of hitting in proportion to their actual, real-life value to a team’s offense.”

        To me, a hypothetical player that would hit a home run every single at bat is not worth a home run every at bat, he is worth a walk every at bat, because he’d be walked every at bat.

        You say “The Book is about batter-pitcher matchups, and IBB don’t help us there.” I’m not sure what you mean by this, nor am I sure how IBB don’t factor in. I’m not saying you’re not right, I’m just saying I really don’t understand.

        So what exactly is wOBA supposed to be measuring?

        Vote -1 Vote +1

  19. Kyle says:

    Just out of curiosity does wOBA take into account when a player hits a single but is thrown out trying to stretch it into a double, or a double into a triple? They both count as hits despite the fact an out is produced on the play.

    Vote -1 Vote +1

  20. intricatenick says:

    It doesn’t make much sense to use a pie chart to visualize this. A bar chart – with the coefficients side by side for the various metrics makes much more sense.

    Using pie’s insinuates that a full 20% of a player’s wOBA comes from triples. I realize that people may know this, but a a more informative pie chart would scale the events by what % of wOBA that they accounted for.

    Vote -1 Vote +1

  21. T-Car says:

    I’m in love with this comment section. It has nearly everything I’m looking for.

    The one thing I’d love to see is how many runs a team is expected to score per game based on their wOBA. I like to describe players talent levels based on how many runs they’d score per game if they were the exclusive hitter. Albert Pujols is a 9 run player. Derek Jeter is a 5 run player etc. Basically, what I want is RC/27 except with wOBA as the measurement.

    Any place I can be sent to for that?

    Vote -1 Vote +1

    • JT Jordan says:

      I imagine you would use something along the lines of wRC/(AB – H)*27.

      Vote -1 Vote +1

      • T-Car says:

        They’ve already got RC/27 using the RC forumla. What I’m looking for is RC/27 using the wOBA formula.

        If a team hits X during the season how many runs would they score per game.

        wOBA
        .300 = X runs per game
        .310
        .320
        .330
        .340
        .350
        .360
        .370
        .380
        .390
        .400

        That way if a player is a .380 wOBA hitter I can say he’s a 7 run player, or whatever it ends up equating to.

        Vote -1 Vote +1

    • tangotiger says:

      wRC is based on wOBA (which itself is based on Linear Weights).

      wRC is indexed to league average. So, if you have 150, and the league average is 4 runs per game, then this player is worth 6 runs per game. I’m not sure if its entirely equivalent to the “all guys on the team like this guy”, but, it gets you most of what you want.

      Vote -1 Vote +1

  22. Eric says:

    the most striking thing to me when looking at the pie charts is how similar OPS and wOBA are. They aren’t identical, but they are pretty close in the grand scheme of baseball statistics. In the vast majority of situations, which one you use is irrelevant. They are merely different shades of the same color.

    Vote -1 Vote +1

  23. andy says:

    Why are strikeouts not included in wOBA? i would imagine that they have more of a negative impact than other outs(as a whole). I understand grounders could lead to double plays, flies to sac runs. Also are there different weights for different types of outs?

    Vote -1 Vote +1

    • tangotiger says:

      Because the run value of a strikeout, overall, is similar to other outs. K with runner on 3B and less than 2 outs are deadly. Horrible. But, a K with a runner on 1B and less than 2 outs are preferably to groundball outs.

      Basically, they cancel out mostly.

      Vote -1 Vote +1

    • Colin Wyers says:

      The one-line wOBA formula presented above is a bit misleading in this regard. There are really two parts to wOBA – the linear weights values, and the translation from R/PA to wOBA. You can use practically anything you want for the linear weights portion, so long as you get R/PA as a result. So there’s no reason you couldn’t include SO in wOBA.

      Vote -1 Vote +1

    • Tangotiger says:

      When I do my base/out version of wOBA, I do include the strikeout. That’s because it’s markedly different from a regular out.

      In an overall sense, if I were to include strikeout, the coefficient for that would be -0.02 in the numerator (and the other coefficients go up a tiny tic to compensate).

      So, a guy who strikes out 100 more times than an average hitter will end up with -2 extra in the numerator, changing his wOBA from say 0.400 to 0.397.

      I’m not opposed to ignoring the strikeout. I can include it if it bothers enough people.

      Vote -1 Vote +1

  24. Columbo says:

    Where can you find RBOE on the FanGraphs site? I can’t find it under the player
    stats.

    If FanGraphs promotes wOBA and that’s part of the calculation, should that not be included in the player stats on this site?

    Good discussion though.

    Vote -1 Vote +1

    • Tangotiger says:

      I don’t know if Fangraphs include ROE in its calculations. But, yes, if it’s included, then we should see it on the player pages.

      Vote -1 Vote +1

  25. Columbo says:

    Can anyone from FanGraphs clarify whether ROE is included in their wOBA calculation?

    If yes, then why is this stat not included on the player pages?

    Vote -1 Vote +1

  26. tangotiger says:

    The Book is free to read at Amazon, using Amazon’s Look Inside feature. It would probably be easier if you read it first. That’s not to say that you have to read it, but I’ve answered these questions so many times in the past, and links have been provided above as well.

    Vote -1 Vote +1

  27. strewth78 says:

    hate to ask questions I know everyone is bound to have obvious answers to, but as much as most people here seem to think wOBA is easier to understand than OPS, I disagree. That aside, I completely agree that OPS is a silly stat; simply adding 2 stats together doesnt achieve anything, especially when both OBP & SLG include singles.
    Anyways, my 2 questions are these:

    Does anyone consider OBP + ISO a better stat than OPS (I know everyone loves wOBA – Im not asking this question to suggest its better or even comparable to wOBA, only incomparison to OPS)?

    And second, if the coefficients included in wOBA set 2B’s at 1.24 and NIBB at .72, does that mean that a double and a walk is considered more productive than a home run (and then a pop-up)? Wouldnt the guarantee of a run scored from the home run be better than a walk and a double, despite the strikeout in the second at bat?

    Vote -1 Vote +1

  28. veshio says:

    Someone mentioned that if we do not include IBB in the numerator, and PA includes IBB, then it should be PA-IBB.

    Then the response was that IBB is not included in both the numerator and denominator… does this mean that the PA used does not take into account IBB? Since PA included IBB, essentially it is unreasonably taking away a chance for the hitter to do something.

    I agree with the first comment, and am unsure why he was so easily convinced by a statement with absolutely no justification.

    Can someone clear this up for me?

    Thank you :)

    Vote -1 Vote +1

  29. FRED says:

    good articlegreat comment the one bad thing about ops you cant figure it from season to the next or cummivatlyits interesting yearto year but the bad thing is it gears to powers who walk a lot than next powers who dont walk alot than ones one who walk a lot but not enough power and finaly it hurts the above hitter no walks no ppower or higher power higher ba but no walk like micky rivers or maty alou please eplain wobs to me and there should be no differnce between ibb or regular walksawalk is a walk and the bestmethod is still bill james won loss pct found in baseball ref leaders board

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current ye@r *