Thread: Article: Changing Luck

You can view the page at http://twinsdaily.com/index.html?554-Changing-Luck

Luck sounds about right :

Hendricks .397 BABIP
Liriano .356 BABIP
Blackburn .336 BABIP
Marquis .336 BABIP
Walters .316 BABIP (before today's game)
Pavano .304 BABIP
Diamond .225 BABIP

I don't think that a high BABIP is necessarily bad luck for the pitcher or a reflection of bad defense. Sometimes, it just means that hitters are tattooing the ball.

I think any pitcher good enough to wear a big league uniform is capable of putting together two good outings, and certainly even the worst team in the league, in any year, is capable of going 5-5 in a 10-game sample.

It would be nice if two unheralded pitchers go on to have a good season and lead the team back towards respectability. But... what are the chances?

Luck sounds about right :

I've become confused at the idea that BABIP is always supposed to converge around one number, from either the pitching side or the hitting side. If you're the type of pitcher who doesn't have huge strike out numbers and whose goal is to induce weak contact, then your goal is to *maintain* a low BABIP. Surely there are many successful - not dominant but decent - pitchers who make careers out of getting guys to hit weak ground balls. The idea that any pitcher with a below-average BABIP - and any hitter with an above-average one - is lucky does not seem patently obvious. Are there studies that form the basis for this bit of sabermetric lore?

Originally Posted by jeffk

I've become confused at the idea that BABIP is always supposed to converge around one number, from either the pitching side or the hitting side. If you're the type of pitcher who doesn't have huge strike out numbers and whose goal is to induce weak contact, then your goal is to *maintain* a low BABIP. Surely there are many successful - not dominant but decent - pitchers who make careers out of getting guys to hit weak ground balls. The idea that any pitcher with a below-average BABIP - and any hitter with an above-average one - is lucky does not seem patently obvious. Are there studies that form the basis for this bit of sabermetric lore?

You're mostly right. Generally there is an average and most players tend to stick close to it, more for hitters. But in reality, there's probably a slightly different "true" number for each person. It's just that it doesn't tend to differ much. For instance, Mauer has consistently (always) had a BABIP higher than average. He hits a ton of line drives compared to average, those are harder to field, etc...basically just what you said. So yeah, when a particular player over and over again goes above (or below) the average, then one can surmise that there's something different about them that makes them different. But overall, people tend to cluster close to that overall average in the long run.

They were pretty unlucky i agree...but I might call it coming up tails more like four times in a row rather than five, and one of those tails wasn't exactly bad luck.

I think they came up heads a couple seasons in a row on Pavano, and thought they could go on flipping heads indefinately. I'm not sure that's bad luck.

And Marquis...I think that might more accurately be described as betting on heads while flipping a two-tailed coin.

Originally Posted by jeffk

I've become confused at the idea that BABIP is always supposed to converge around one number, from either the pitching side or the hitting side. If you're the type of pitcher who doesn't have huge strike out numbers and whose goal is to induce weak contact, then your goal is to *maintain* a low BABIP. Surely there are many successful - not dominant but decent - pitchers who make careers out of getting guys to hit weak ground balls. The idea that any pitcher with a below-average BABIP - and any hitter with an above-average one - is lucky does not seem patently obvious. Are there studies that form the basis for this bit of sabermetric lore?

Not every player above or below the average is lucky or unlucky. Every player, given enough time, will establish their own BABIP baseline. Pitchers tend to settle closer to the league average because they face all different types of hitters. Batters will have a wider range based upon their individual skill sets (groundball/flyball, fast/slow, hard/weak contact, etc).

A guy like Ichiro posts BABIP's in the .350 neighborhood, while a guy like Joe Crede was typically .250-ish. Relievers typically post lower BABIP's than starters.

I really don't like using BABIP for pitchers. I think it's a useful tool for hitters, but there are much more useful sabermetrics that account for luck like xfip in terms of pitchers.

The problem with your concept is that you pretended it was a 50-50 coin flip, when the odds were always stacked to the tails side. If you pretend that all of these factors were a coin flip, then it is disappointing that almost all of them have come up tails. But, if you had more realistic odds then you should not be surprised that Pavono continued to decline, Liriano is struggling even more with consistency, Scott Baker continues to have arm problems, and Jason Marquis isn't that good.

And, I think that this was also the outlook of the Twins management. It was all a coin flip, and now we have the real results. We should have been looking at this season as a rebuilding year from day 1. Instead, we played the middle road, hoping beyond hope that all of those coin flips would come our way. I am not saying it was catastrophic, but it just delays the overall rebuilding process.

Originally Posted by jeffk

Are there studies that form the basis for this bit of sabermetric lore?

Jeff, check this out... http://www.fangraphs.com/fantasy/ind...p-spreadsheet/. It can be used to calculate what one would expect a player's BABIP to be based on their batted ball profile (% line drives, % fly balls, etc). BABIP - xBABIP would give you a better indication of a player's "luck" than just a surface level assessment based on their current BABIP.