Lookout Landing: An SB Nation Community

Navigation: Jump to content areas:

Around SBN: Staturday: Does Jason Giambi help the A's? Bar-right-arrows

Deconstructing WHIP

Hmssurprise_tiny by Graham on

WHIP sucks.

What's wrong with it? Let me give you a list.

1) Assumes hits given up are entirely under the pitcher's control.

Did you know that defence isn't entirely static? Funny that. Hits don't occur at the same rate in front of different fielders and in different ballparks, and the variation in defence/park effect far outstrips any variation in hit prevention skill amongst ML calbire pitching.

2) Assumes innings pitched are entirely under the pitcher's control.

Sort of the same as point one but so crucial I decided to mention it twice. Outs are determined by pitcher skill, defence, and park effects. Pretending that the latter two don't exist when trying to measure pitcher effectiveness is a bad idea.

3) Improper mapping of different types of hits to different run values.

WHIP values a single the same as as a home run. At least ERA doesn't have this problem.

The problem with throwing together a set of numbers that look vaguely similar and calling them statistics is that generally the result means, in layman's terms, sweet bugger all. For all intents and purposes, WHIP is ERA except worse. At least ERA tries to measure runs, while all WHIP does is... measure walks and hits per inning pitched.

It's crap like this that gives real analysis a bad name. Fantasy baseball has a lot to answer for.

1 recs | Comment 113 comments

Read Related

Story-email Email Printer Print

Comments

Display:

While you shouldn't use WHIP ever

Isn’t it a bit better than ERA, because it at least tried to measure on-base percentage, as opposed to runs scored?

Furcal

by JI on Dec 2, 2008 10:04 PM PST   0 recs

WHIP is my favorite math-phobic journalist's mocked statistic

WHUP, WHOP, FLOP, FLIP, SHIP, WHAT, WHIN, et cetera. I made all of those up, but they all seem like plausible variations of dumb acronyms created by stat-blind, Moneyball-bashing “baseball analysts”, who somehow still manage to retain their jobs despite churning out articles full of misinformation and back-handed swipes at people who use math to describe things.

But the hilarious part is that no one who knows a thing about baseball analysis believes WHIP is even marginally worthy.

by JLC on Dec 2, 2008 10:50 PM PST   0 recs

I never heard of WHIP in my life until I started playing fantasy baeball.

Then when I looked up just what the hell this stat meant, the first thing that came to mind was, “wow, that’s pretty retarded. They really use this to determine fantasy champions and cash prizes?”

I haven’t played fantasy baseball in two years. Coincidentally, that’s around the same time I discovered sabermetrics.

Every day I hear about Seattle sports' failures. Every night I fall asleep to the sound of my own tears.

by Benne on Dec 2, 2008 11:16 PM PST   0 recs

WHIP

more useless than even OPS?

Adopted Giant: Aaron King

Wearing the crown by 2011. Or at least the LOOGY hat

by baetown415 on Dec 3, 2008 12:04 AM PST   0 recs

Doesn't OPS correlate well with runs scored on a team level?

I remember a 2006 or 2005 article from THT about this and it being comparable to RC and Base Runs in that regard.

Ah. Here

by Fez on Dec 3, 2008 1:15 AM PST to parent up   0 recs

Yes

Despite its faults, OPS is fine.

by katal on Dec 3, 2008 5:49 AM PST to parent up   0 recs

It's got an R^2 of .989

from 1967-2007 for the AL.

by Fett42 on Dec 3, 2008 5:59 AM PST to parent up   0 recs

Just to clarify:

This relationship is between teams runs scored and team OPS?

by FlaskInSafeco on Dec 3, 2008 9:26 AM PST to parent up   0 recs

The THT article that Fez linked to above

found an R2 correlation of .9128 between team OPS and team runs scored for all teams from 2000-2004, which is in line with the correlations if seen calculated for this in other places. I find it hard to believe that there was an R2 of .989 between runs scored and any other offensive statistic. That’s a pretty ridiculously high correlation. The nice thing about OPS is that it predicts runs scored on the team level nearly as well as more advanced offensive stats and it’s much simpler to calculate.

by FlaskInSafeco on Dec 3, 2008 9:57 AM PST to parent up   0 recs

That .989 correlation is for OPS to Runs per Game for the entire AL (not any individual teams) 1967-2007

I did that as part of a mini-project for a basic stats class. Using R/G probably decreases some of the variation due to different amounts of games played. I know .989 is really high, but that’s what the numbers say.

by Fett42 on Dec 3, 2008 11:55 AM PST to parent up   0 recs

Oh, OK for the entire league.

That certainly makes it more believable. I buy it. If that’s the case I wonder if the more sophisticated stats do better when regressed over the entire league. I mean, from .989 there’s really not much room for improvement.

by FlaskInSafeco on Dec 3, 2008 3:46 PM PST to parent up   0 recs

OPS is OK.

It doesn’t weight things properly, but it’s at least measuring the two most important things a hitter can do — get on base and hit for power.

by Teej on Dec 3, 2008 8:06 AM PST to parent up   0 recs

In Moneyball they mentioned that OBP is roughly three times as important as SLG,

which is something I hear repeated all over the place. How do you suppose the Oakland guys figured this? I suppose they might have created a formula for a stat we could call “adjusted OPS (aOPS)” and say that aOPS = c*(OBP) + (SLG), where c is some constant. Then if you ran a bunch of regressions of team aOPS vs. team Runs Scored for different values of c, you could find an optimal c that gave you the best correlation, and the Oakland guys might have found that c ~= 3. Maybe I’m doing this wrong. Thoughts, anyone?

If you did something like this, I’d be curious to see how much the constant fluctuated over different seasons (i.e. is OBP more or less important in relation to SLG during higher or lower run-scoring environments?). Also, if we are fairly confident that using 3*OBP + SLG will give you a better predicter of runs scored than just OPS, why isnt this stat used more often? It’s not that much more difficult to calculate. I also wonder how a stat like that would stack up against some of the more sophisticated offensive measurements, like runs created.

by FlaskInSafeco on Dec 3, 2008 10:11 AM PST to parent up   0 recs

GPA is what you're after

Matthew wrote about it a while ago here

by Graham on Dec 3, 2008 10:13 AM PST to parent up   0 recs

Thanks. I always thought saying OBP was 3 times as important as SLG seemed a bit much.

Any idea of how GPA stacks up against BaseRuns or some of the other more sophisticated offensive metrics in terms of predicting runs scored? Also, it seems like in this study GPA was regressed against BaseRuns, and not runs scored. Any particular reason for this? I would think that you would want to be aiming to predict actual runs scored.

by FlaskInSafeco on Dec 3, 2008 10:29 AM PST to parent up   0 recs

i was talking more about on an individual level

but i guess i’ll keep OPS around for occasional use

Adopted Giant: Aaron King

Wearing the crown by 2011. Or at least the LOOGY hat

by baetown415 on Dec 3, 2008 2:46 PM PST to parent up   0 recs

As flawed as WHIP is, its rise gives me hope.

When I first started hearing about it in the early 90’s most people ignored it. Now it is commonly accepted.

Keep pimping tRA and in a decade it may make its way into the common culture.

by Sec 108 on Dec 3, 2008 6:06 AM PST   0 recs

It's funny to me when I read random internet comments deriding tRA for being a made up stat

and the comment then uses WHIP to show how tRA is flawed.

WHIP is exactly the sort of statistic these people should be yelling against. It’s arbitrary and backed up by no useful analysis. It is a stat for stat’s sake. WHIP is what’s ruining baseball.

by Matthew on Dec 3, 2008 7:55 AM PST   0 recs

Ruining baseball?

That’s a bit hyperbolic. WHIP has one single use – fantasy baseball. It works in that context. To steal Graham’s phrase, fantasy baseball means sweet bugger all in the context of actual baseball; let the WHIPheads have it in their fantasy leagues, and leave it at that.

Nice Guys Finish Third - Hopelessly lost, but makin' good time.

by pdb on Dec 3, 2008 8:03 AM PST to parent up   0 recs

I get what Matthew is saying though.

I know a lot of people who are heavy into fantasy baseball, and the statistics that their leagues use are the only statistics they give a crap about. So when you try and tell them that Player X isn’t very good, they use Player X’s fantasy value to point out how wrong you are. It’s immensely frustrating.

by acblue on Dec 3, 2008 10:15 AM PST to parent up   0 recs

Yeah, I get that too

but if a person’s baseball experience/conversational knowledge is limited to a player’s fantasy baseball usefulness, I tend to not listen to them very much in the first place.

Nice Guys Finish Third - Hopelessly lost, but makin' good time.

by pdb on Dec 3, 2008 10:17 AM PST to parent up   0 recs

There's no defense in fantasy baseball.

You could start Albert Pujols at second

Furcal

by JI on Dec 3, 2008 10:17 AM PST to parent up   0 recs

I run a fantasy league, and it's definitely a challenge to separate the two.

I look at stupid stats because I have to, but you have to keep them out of your real-baseball evaluation.

by Teej on Dec 3, 2008 11:13 AM PST to parent up   0 recs

I will never understand how 'WHIP' became associated with sabermetrics

and how dumb writers mock sabermetrics by saying all they care about is WHIP. Bizarre.

I also don’t get how it became such a fad in fantasy leagues. Why? What’s the point of it? The only thing that makes sense to me is trying to have a stat that you can draft around, but still includes sweet, sweet volatility. I don’t know why RA doesn’t provide that, but there you go.

The more you look at it, the more hilarious it gets. It’s not even a useful shorthand for anything. OK fine, it maps -sort of – to OBP. But….OBP doesn’t need to be shortened. It’s short! Just fucking use it! As cobbled-together and sort of messed up as OPS is, it works and DOES highlight two components of run scoring that mean more (esp. combined) than the ubiquitous batting average. WHIP obscures far more than it illuminates, and really needs to go away.

by marc w on Dec 3, 2008 10:08 AM PST   0 recs

Why? What’s the point of it?

To measure how often pitchers allow baserunners. If your league is awarding points for best OBP, it makes sense to award points for preventing baserunners.

Furcal

by JI on Dec 3, 2008 10:19 AM PST to parent up   0 recs

I've only played fantasy baseball once -- last year.

Like most fantasy games, the stats used in our league had little bearing on winning actual baseball games: sb (but not cs), avg, hr, rbi, runs, era, whip, saves, wins, and strikeouts. You were awarded a “win” or “loss” for each statistical category over the course of a week in a head-to-head matchup with another team.

While I would definitely agree that the usage of these sorts of stats in fantasy games contributes to the ignorance of the average sports fan, I would argue that these stats made the fantasy league more interesting from a gaming standpoint. I mean, could you imagine a fantasy league where the offensive categories were wOBA, Eqa, OPS, BaseRuns, and VORP? While this might be a better measure of baseball knowledge, it would probably be pretty boring. Good players would be generally good in all categories, and bad players would generally suck in all categories. In the dumber version of fantasy, you have to think about how to game it to pick up the bullshit categories like sb and saves.

Anyway, all that said, I really try to avoid fantasy sports. It takes up time with really irrelevent stuff, and I hate even the slightest possibility of feeling conflicted while watching a game because of fantasy reasons.

by FlaskInSafeco on Dec 3, 2008 10:43 AM PST to parent up   0 recs

I don't mind fantasy sports, but don't play it much

I was in an AL-only, very sabr-friendly league that used Tango’s run values for events (1b, 2b, k, HR, etc.). A lot of fun, I suppose, but it’s a hell of a lot of work to update line-ups every day.

also, I got my ass kicked, so that was less cool. Injuries. Damned injuries.

by marc w on Dec 3, 2008 11:48 AM PST to parent up   0 recs

We subbed out BA for OBP in my league, and it's awesome.

But the rest of the 5×5 categories don’t have easy replacements like that, unfortunately.

by Teej on Dec 3, 2008 11:14 AM PST to parent up   0 recs

Yes....why?

How/when did this happen?

by marc w on Dec 3, 2008 11:46 AM PST to parent up   0 recs

I'm not sure what the point of this FanPost was, to be honest

you’re pretty much preaching to the choir here

by seattlebruin on Dec 3, 2008 12:43 PM PST   0 recs

Something to link to.

I love posts like this because Graham is better at explaining these things than I am, so I can just say “read this” when I get into arguments.

by acblue on Dec 3, 2008 12:45 PM PST to parent up   0 recs

Heh, just saying

the Triumvirate has us all conditioned to the point where when we see a pitching metric the first thing we do is ask “does it include hits?” and if it does, consider it stupid

by seattlebruin on Dec 3, 2008 12:49 PM PST to parent up   0 recs

And even stranger...

when I hear the letters tRA… I start salivating. What’s up with that?!

by johnbai on Dec 3, 2008 1:07 PM PST to parent up   0 recs

Graham's been conducting an elaborate experiment on you.

Trajan is my favorite emperor.
Have you tried the new trattoria?
I’m excited to see GMZ pull of some trades.
Where have you traveled (besides Belgium), John?

by marc w on Dec 3, 2008 3:00 PM PST to parent up   0 recs

?

Furcal

by JI on Dec 3, 2008 1:52 PM PST to parent up   0 recs