wOBA & Base Running Changes
To better isolate batting and base running skills, SB and CS have been removed from wOBA and given their own category wSB.
What’s Changed?
wOBA, wRAA, wRC, wRC+ – These all no longer include SB and CS. SB and CS have been removed entirely from the wOBA calculation include the weights calculation.
Batting – Batting in the Value section no longer includes SB and CS.
Base Running – Base Running in the Value section now includes SB and CS.
FanGraphs Guts! – Our Guts! section now has updated weights and the SB and CS weights for use in wOBA have been replaced with the raw run values.
What’s New?
wSB – how many runs above or below average a player is at stealing bases.
UBR – (miscellaneous baserunning / advancement) is now broken out into a separate category which is in the Advanced section. This contains all the information that was formerly just in Base Running.
David Appelman is the creator of FanGraphs.
And Trout’s still as good as Cabrera.
Is it possible to add a new stat like the old wRC+ so that we have both?
Yup, I’m going to miss my super-awesome all-in-one offensive measurement of yumminess.
Isn’t wOBA supposed to be a calculation of a players total offensive contribution? In that case, doesn’t removing SB and CS entirely change the point of wOBA? Is the way you calculate WAR unchanged?
WAR is unchanged because its SB/CS runs were just moved from one category to another.
wOBA is a calculation of a player’s total BATTING contribution. Once they get on base, it’s a different story.
Well yeah now that’s true. That isn’t an accurate description of what wOBA used to be though. Before the change it was meant to “measure a hitter’s overall offensive value”. Now it seems there is no single measurement of a players offensive value, assuming one considers stolen bases of offensive value.
Given the name, it does make sense to isolate out what happens after you get on base. The only thing about it is that you still get credited in wOBA for any triples and doubles which are more of a product of a smart/speedy batter’s base running acumen than his pure hitting ability.
Weighted On-Base Average is certainly a misnomer. I actually like that they have isolated base running. I just don’t know that changing what wOBA means was the best thing to do. It’s essentially a new stat now so I don’t know why they just wouldn’t create an entirely new stat instead, like wAVG perhaps. That way wOBA remains a measurement of OVERALL offensive value.
But wOBA wasn’t that before, because it didn’t include going from 1st-to-3rd on singles, breaking up double plays, scoring on sac flies etc. – those are things of offensive value.
Okay Aaron, there is an argument that could be made to include those things if possible, but that doesn’t mean wOBA wasn’t intended to be an evaluation of overall offensive value, and it certainly isn’t an answer to my questions.
SecondHand,
wOBA only “measured a hitter’s overall offensive value” on Fangraphs.
wOBA was actually created to weight different offensive events between batters/pitchers. Fangraphs changed the definition on this website but the “true” definition never changed. Fangraphs is now going back to what wOBA originally intended, and coincidentially the same definition used everywhere else.
This is a good move and does not change the intention of wOBA.
This makes a lot of sense to me, as long as all the data is still available to play around with.
So…. the new rankings are
wOBA 1. Cabrera .417, 2. Trout .409, 3. Fielder .398
wRC 1. Cabrera 137 , 2. Fielder .125, 3. Cano 124, 4. Trout .121
RAA 1. Cabrera 57.3, 2. Trout 48.2, 3. Fielder 46.1
wRC+ 1. Cabrera and Trout 166, 3. Fielder 153
UBR Trout 7.0, Cabrera -0.4
wSB Trout 5.0, Cabrera – 2.4
(just used Trout and Cabrera in base running stats for follow up on earlier fun stuff)
But it never included baserunning outside of SBs anyway.
Who were the players with the biggest changes in the effected stats?
Rajai Davis went from 1.8 WAR to -23.6 WAR
That’s freaking hilarious!
WAR isn’t affected by this change.
SB and CS was just moved from one category to another.
tbj is a buzzkill.
Shoot. Now I can’t handily point people to wOBA to explain why Mike Trout was better than Miguel Cabrera offensively this year. Of course, the SB and base running still counts, but I’ll have to make them add them in. Oh well.
Actually, I really like this change, just because it’s nice to have the components isolated. Takes a little more mental work, but that’s not necessarily a bad thing.
Point to wRC+
Also, if WAR is unchanged, why is Ryan Braun now at 7.9? He was at 8.0 before, I’m almost 100% certain of that.
He took a WAR cut in order to defer to Buster Posey’s MVP candidacy so he can proclaim he will be a 50/50 player next year, get injured multiple times, only play in about 106 games, and finish with less than double digit stolen bases.
If I’m not mistaken, there was one final UZR update after the end of the season. It’s possible that said adjustment is responsible for the change in Braun’s WAR figure.
It is possible that a player’s WAR might have changed ever so slightly, but any changes you see are most likely going to be rounding differences. It mainly has to do with the way IBB’s are being handled and their dynamic weighting.
Before if a player had a high SB total, it added to their wOBA and run value per plate appearance, and since the value of an IBB is set at a player’s individual runs per PA, players who had high IBB totals and high SB / low CS totals, will have the value of their IBB slightly reduced. This could result in a .1 change in WAR.
As for Braun, I’m actually pretty sure his total remained unchanged and with the final UZR update, I believe he was at 7.9.
is WAR precise enough that a change of 0.1 really means anything?
Nope.
So, stolen bases and caught stealing have nothing to do with runs created? Ok. Removing from wOBA I can sort of see, if I squint, but ending up with a “runs created” stat that ignores base running altogether is odd.
Much appreciated changes. This along with this pre-season’s inclusion of DRS alongside UZR show that FG is willing to keep an open mind on skills components that are ripe for various biases. Thanks Dave and Dave.
Is there any prospect of the WAR implementation of stolen bases/caught stealing being expanded to cover base/out state (and the particular base stolen)? i.e. a RE24 approach.
It is faintly ludicrous that a steal of home with 2 outs counts the same as a steal of 3rd with 0 outs. The system also makes occasional stealers with poor success rates look unreasonably bad: they are probably usually stealing 2nd with 2 outs, where the breakeven point is much lower.
This was a good article on here about the breakeven points:
http://www.fangraphs.com/blogs/index.php/breaking-down-stolen-base-break-even-points/
Context-neutral.
If you do that, you need to do it for every event.
I disagree, when it comes to steals context is everything.
We do context-neutral for batting because we assume the batter is trying to get a hit or draw a walk all the time, regardless of the base-out situation. And when they’re not, like sacrifice bunts, we generally discard the data.
Steals are qualitatively different, a runner is weighing up the risk/reward before they decide whether or not to go.
Bonds’s wOBA and wRC values have taken a pretty big hit with the changes.
Good move, Dave. This makes total sense, separating baserunning from hitting. Fangraphs is a great site and continues to make improvements.
Since wOBA was created in The Book, then I have to correct the above reader who is trying to say what wOBA was intended to be.
wOBA in The Book was simply to weight each of the offensive events insofar that we needed them in various Batter/Pitcher studies. Obviously, SB, CS (and PK, BK, WP, PB, DI) had nothing to do with the results of batter-pitcher matchups.
Whereas OBP gave a weight of “1” to each of the BB, 1B, 2B, 3B, HR, we needed a better balance, hence the “weighted” on base average. BB went from 1 to 0.7, the 1B went from 1 to 0.9, HR from 1 to 2.0, etc.
***
Now, as for whether it’s better or worse to have SB, CS, etc in or not in wRC and wRC+, you can go either way there.
There’s something nice in being able to keep all the baserunning components as separate as possible until you need to bring them together at the very end.
I’m ambivalent there as to what’s better. But, I definitely think that if you include SB, CS in wRC, then you should include all the baserunning numbers like going first to third.
It’s easier to describe wRC as “runs as a batter”, than to say “offensive runs, except taking extra bases on batted balls”.
***
In any case, since David has given you all the numbers broken down, the reader can certainly lift his pinky finger and add things together when he needs to.
I agree. I like to be able to separate the various components. If one wants to see just batting numbers, why not have that available? If you want to see total offense combined, have that also.
I think this also makes the transition from traditional stats to saber stats a bit easier, conceptually, as you can see the natural progression more clearly.
Now, can someone please explain this stuff to Jim Leyland? Thank you!
I like the changes. I think they should also consider changing Fld in the Dashboard to Fld+Pos.
I love this. This came up in the Q&A recently and seemed to generate a positive response then as well.
This is a good move simply because the data now compares directly with Baseball-Reference.
Can you please still list the separate values for baserunning, i.e. split the two components (SB/CS v. other)?
Or have a baserunning tab? There are lots of interesting stats in baserunning.
There are separate values for SB / CS vs other. They’re listed in the advanced section under UBR (other) and wSB (SB / CS).
Awesome! Thanks.
I don’t pretend to be nearly as smart as 99% of you guys but, I understand an SB has nothing to do with getting on base. So I’m completely happy with taking it out of wOBA, however I can’t see how it’s being taken out of wRC and wRC+. I’m also with Tango on this, all baserunning should be put in, if you are going to add SB back in. Why wouldn’t we want an all encompassing hitting stat, AND one for ALL offensive value
Finally!!!
I have never understood the reasoning to including SB/CS in wOBA. To me having a stat that encompassed hitting only (i.e. batter versus pitcher) was the most logical stat missing from this website.
It is very easy to add in stolen bases and other baserunning metrics when evaluating a players total offensive performance but including them in wOBA was a disservice as you could not identify a hitter in one stat based on their pure hitting only performance.
Thank you!!!!
Breaking up wOBA this way seems to emphasize WAR as the final roll-up measure of a player’s contribution to their team and takes away any roll-up stat for total offensive contribution. True, as many have said, that the old wOBA didn’t include EVERYTHING that happened on the basepaths, but it was better than nothing. Given that measuring defensive contribution is far from perfect, I liked having the one offensive stat to look at along with WAR. Maybe I will get used to this in time, but I think I will miss the old wOBA.
It’s actually going back to the old wOBA. You are going to miss the not-old not-new wOBA.
Does UBR include all aspects of base running? Stealing 2nd base, Stealing 3rd base, Caught Stealing, Running 1st to 3rd, Breaking Double Plays, etc?
will wSB include Pickoffs that aren’t Caught Stealing, or just PCS? I know in the case of some players, like Elvis Andrus and Ian Kinsler in 2012, Pickoffs that don’t result in a CS can end up being a fairly large number. Per B-R, they had 9 and 8 Pickoffs respectively, but only 2 Pickoffs Caught Stealing each.
Forgive me if this is explained somewhere on the site, but does wSB take into account which base is being stolen (or at which base the player is being caught)? I would think stealing home would be far more valuable (and getting caught there far more detrimental) than the same events at second base. Or is it simply a measure above or below league numbers?
I’m gonna miss the old wOBA…
I like this change for wOBA, as it is a hitter/pitcher stat and that does not include baserunning. But removing baserunning from wRC makes no sense to me. How does baserunning not effect the amount of runs a player creates for his team? I have always thought that wOBA and wRC are a little too similar anyway, they are usually correlated pretty closely. Why not leave SB and CS in wRC+ and use wOBA as the primary BATTING stat and wRC as the primary OFFENSIVE stat? There should be a way to measure a player’s total offensive value/production without having to use WAR. wOBA and wRC both used to do it, at least 1 of them do it still and leave wRC alone.