Never saw him deride. Ok, now that that’s out of the way…what has happened to Cohen?
Gary: Because we don’t talk about sabermetrics very often. And you know I think there are certain metrics that have come into play that are useful. But I think sometimes the people that are adherent to sabermetrics overrate the importance of them. In David Wright’s case, David has struck out a ton this year. Struck out 91 times. And yet he’s hitting well over .300. Now, one of the stats that the sabermetrics people like to throw at you is batting average on balls in play. And if you have a particularly high batting average on balls in play, they like to think that it’s because you’re lucky. In other words, if you have a high batting average on balls in play, it shouldn’t be that high, which means you’re having a fortunate year and you’ll come back down again.
Gary: Conversely, if a pitcher has a particularly low batting average on balls in play, they like to tell you it’s going to rise eventually. Well, to me that doesn’t make any sense. Certain guys hit the ball harder than other guys hit it. Certain pitchers induce more groundballs or more weakly hit balls than others. That’s part of what you’re trying to do. Am I totally off base with that?
Ron: No I totally agree with you, I think that for the average hitter, to have a high average putting balls in play, it’s probably because they do have some lucky hits. But certain hitters, like Wright, hit the ball hard almost all the time.
...I used to rave about Gary Cohen’s broadcasting and his quickness to embrace things like on-base and slugging percentages because those things made sense to me and I thought it was important (and commendable) that they be stressed above certain lesser things (batting average, RBI, and so on). I still like Gary a lot, but more for his game-calling and less for his grasp of the finer points of baseball analysis. Your average listener has no interest in understanding linear weights or regression analysis, but I think even casual fans can grasp the basics of things like BABIP (when explained properly) and FIP, and, contrary to what some may think, awareness and comprehension of these tools serve only to increase one’s appreciation for the sport and its players, not detract from it. It’s disappointing that Gary feels otherwise.
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
Sure. An example would be that the variation in IQ among professional athletes (presumably ranging from borderline retarded to genius level) is larger than the variation in IQ among professional physicists. Professional physicists HAVE to have a high IQ to do their jobs, which cuts off the left tail of the distribution and therefore reduces variance. Professional athletes are not selected on this basis and therefore have a higher variance. But no one would say that having a high IQ isn't important for being a professional physicist just because there isn't as much as much variance in it.
More generally, things that a particular profession selects for are generally quite important to that profession but the selection reduces variance in ability on that dimension for people in that profession. If this is an ability which "tops out" at a certain level, it may eliminate the variance entirely on that dimension for those people, but that doesn't mean the ability isn't critically important, just that it doesn't explain differences in performance amongst people who are already capable of performing that job.
On another note, something I'm wondering that I don't see answered in this thread (but maybe I just missed it) is how important is the part of BABIP which is orthogonal to K,BB,HR in overall pitcher success? That is, we know that pitchers vary on BABIP abilities and that for normal pitchers most of this variance is correlated with (normally we say explained by, but people seem to infer from that statement a causality assertion that isn't present when we use that term in academic circles) the DIPS stats, but that some of the variance in that ability is not correlated with those stats (even for normal pitchers.) I'm a bit rusty at this, so I'm not sure that I can just multiply the percentage of variation in BABIP not explained by DIPS stats by the relative importance of BABIP in actual outcomes.
I guess the simplest way to put what I'm asking is does anyone know what relative coefficient is attached to BABIP in a regression of RA/9 or some other pitching outcome on DIPS stats plus BABIP? This is "important" since if that coefficient is small enough we can effectively ignore BABIP when doing quick and dirty projections for established players.
I don''t buy it. I don't see any evidence that pitchers are selected on babip to a greater extent than say, strikeout ability. And there is a lot of variation in strikeout talent. Also, look at hitters on the mound. They should not have this ability that pitchers are selected for, and their babip is not 50-75 points higher than regular pitchers.
I put the range of talent somewhere in the .280-.320 range. With .325 about the upper limit you'll see for a guy who can throw 85-90 somewhere near the plate.
If only we can put to rest "luck". You would think scouts would say "luck" every 5th word. Geez, when will the stats geeks stop using "luck" and start working on explanations. Luck doesn't exist. There are reasons for everything and just cuz you can't define one, doesn't make it luck.
Just admit the vast majority of events involving humans are not easily defined. That is why we need scouts.
has there ever been anyone saying we don't need scouts? most people I know say we need scouts to rate based upon a players ability instead of concentrating on how they look in jeans... but not even Beane has ever said we don't need scouts, just that they are overrated, have a crappy old boy network in place, and that a vast majority of them are probably incompetent at their job since it's been regarded as an art instead of rigoursly setup as a science. Scouts are necessary, the problem has been that their jobs have been allowed to atrophy for so many decades that there is no legitimate way to grade a scout.
Good scouts are integral to baseball, to say different is just silly, but the problem is that the vast majority of scouts really aren't that good.
Nobody disagrees with your last statement, but if you replace the word "luck" with "events that are controlled at some level, but at the moment the control is undefined, or in some cases small enough that the net result is the same, or close enough to be below the threshold of the measurable effects", then the posts get kind of long. And EVERYONE knows what's meant anyway. The only people who don't are the people who object to the word "luck". I mean, if I'm on an insect forum I don't say "those little black insects with 3 body segments who live in a nest with a queen and who enter my house to find food, and when they find it they communicate to the rest of the nest where the food is and then proceed to make a long line to bring the food back". I say "ants". Because everyone knows what an ant is.
Same kind of thing for hitters. We know that the range is considerably wider. Everyone agrees on that. But some numbers aren't sustainable over a season. So when a guy has a .450 BABIP through the first month, we know that he's not going to keep that up. So it's reasonable to say that he will regress.
*replace the word "luck" with "events that are controlled at some level, but at the moment the control is undefined, or in some cases small enough that the net result is the same, or close enough to be below the threshold of the measurable effects"
Luck is a model. It is often useful in projecting the performance of baseball players.
IIRC we've seen that when position players or washed-up pitchers pitch, they tend to have high BABIPs. Take John Smoltz last year; clearly he still had the ability to limit walks and get strikeouts, but his H/9 was through the roof, being driven by a .354 BABIP. (And his home run rate was a bit elevated.) Some of that could have been the AL factor, but only some.
But Rally, the data on non-pitchers supports exactly the opposite conclusion. Non-pitchers are just as much worse on preventing hits on BIP, compared to real pitchers, as they are with regard to striking hitters out or preventing HRs. Why does it require a 75-point difference in BABIP to prove that pitchers are selected on that basis, but a much smaller difference on the DIPS metrics?
I don't think this proves that the range in BABIP talent is as large as the range for Ks or HRs. But it shows that it's in the same ballpark. There are four "real" pitching skills, not three. I believe the range of skill in hit prevention, between HOFers and AAA pitchers, is the narrowest of the four skills. But it's not much narrower than walk prevention. And there are definitely pitchers for whom it is one of their two greatest skills (as I showed with regard to Seaver above).
I was going to look up Sid Fernandez's player card on BP, but they've completely rejiggered everything. Where did all the data go? BP??
Anyway, from looking this up a while ago, Sid Fernandez and Charlie Hough stuck out.
EDIT: by the way, GuyM, what are the calculations for turning those rate stats into runs?
This, of course, goes back to the original DIPS exception: knuckleball pitchers. Basically, BABIP is THE skill which enables knuckleball pitchers to excel at the major-league level.
Oh, and I'd echo Matt's edit. I've seen calculations that basically show that the range in "true talent" for all 4 pitcher skills - K, BB, HR, BABIP - are virtually the same when expressed in runs, but the math of how many runs .020 points of BABIP or 1.5 K/9 translates into has never been clear to me.
I think that the "DIPS exception" categories are interesting, but pitchers differ internally within those categories, not just among them.
Actually, that's what I took the "DIPS exception" categories to mean: these are the types of pitchers across whom BABIP varies the most. It's not that "knuckleballers" are harder to hit than non-knuckleballers, it's that the variance of how hard knuckleballers are to hit is more "real" than the variance among guys who throw 94 MPH fastballs. But maybe I've misunderstood that.
I was using back-of-the-envelope estimates:
Each additional BB/9 = -.30 runs
Each extra K/9 = +.28 runs
Each extra HR/9 = -1.4 runs
Each extra H on BIP = -.8 runs (you also have to estimate BIP/9, which for the non-pitchers is pretty darn high!)
I don't think the variance in skill is "virtually the same" for all skills. I think K/9 is largest, then HR, BB and BABIP probably in that order. If Tango stops by, maybe he can confirm (in addition to pointing out that using per-PA metrics would be superior in all cases!). To further complicate things, the truth is you can't completely separate these skills. For example, every single major league pitcher could reduce their walk rate, if they didn't mind giving up a .330 BABIP and recording fewer Ks. So when we say a pitcher has "good control," we really mean he can avoid walks without getting hammered.
What I find odd is the assertion that BABIP can be "explained" to some significant degree by K rate or K/BB. To the extent these skills are correlated, one could just as easily (and incorrectly) say that strikeout ability is "explained" by BABIP skill. There is variance in all 4 skills apart from their overlap. The valid reason to care more about K and K/BB is that we can accurately detect those skills much earlier in a pitcher's career -- they are indeed MUCH better indicators of overall skill for young pitchers. But one skill is not logically prior to, nor the cause of, the other skills.
Well, it's also highly unlikely that there's any player in any situation that really has a BABIP ability worse than the .330-.340 range, given that's how hitters with no professional pitching experience have generally fared. I'd love if we had home run derby location data - that would pretty much establish a hard floor.
There's some selection bias here, too. Old pitchers are generally assumed to be washed-up based on the performance resulting from extremely high BABIPs, so the fact that they lose their jobs after high BABIPs can't be used as evidence that those pitchers are washed-up.
I was objecting to the quote suggesting that minor league pitchers with babip around .350 or whatever aren't getting selected to the majors, so we see less variation at the MLB level. I'm saying those pitchers do not exist. That real pitchers toiling in the minors, and deemed worthy of professional employment at any level, should be at least as skilled as the backup shortstop who takes the mound in a blowout.
I am not saying differences in skill do not exist, but I'd put the range of ability somewhere in the +/- 0.020 range for babip, among professional pitchers.
This seems semantic. They are still being selected out, just earlier in the process.
Don't forget about Eephuscopalians and Guys Who Do Whatever This One Is Called: http://www.youtube.com/watch?v=UdYBEJzy-F0&feature=related.
I just pulled the .350 out of my ass. The "or whatever" was the general point: the right-hand tail (or left-hand, depending on how you want to view the graph) of the BABIP-skill population is being left behind in the minors. Of course, the same is presumably also true of K-rate, BB-rate, and HR-rate.
You must be Registered and Logged In to post comments.
<< Back to main