pBABIP Rating and Defense

kq76 · 07-26-2023, 05:52 AM

So I've long thought Stuff is less important the better defense you have, but what about pBABIP? You see people say that a bad pBABIP stat may be an indicator of bad defense behind the pitcher, but is that really true? I think of it as more an indicator that the pitcher's pitches may just be easier to hit hard. But either way, if you have a good defense should you be as concerned about having pitchers with good pBABIP ratings or not? I think the answer is probably yes, but I'm not sure.

Syd Thrift · 07-26-2023, 06:28 PM

To be honest I think good defense should have more of an effect on pBABIP but the rating was put in to take care of pitchers who have better BABIP because of their stuff, which I find it downright bizarre that the devs haven’t included (there’s a modifier based on velocity and perhaps on Stuff of a couple specific pitches but the literature I’ve seen is that this effect seems to be best correlated with just straight up K rate) as well as pitchers like Jim Palmer, whom the people who work on the historical engine apparently wanted because “changing players defensive ratings messes up the stats”, which of course is a ridiculous comment since even the best teams only get 27 outs and as long as player A has ratings that are in line on a general percentage basis with player B, you wouldn’t see individual stats go crazy like that.

That said, I think you’re correct in that it probably doesn’t do a huge amount one way or the other. And I guess as it stands, studies of the game indicate that fielding already has a really large factor on the game, so let sleeping dogs lie, I suppose…

MathBandit · 07-26-2023, 07:59 PM

Quote:

Originally Posted by Syd Thrift

To be honest I think good defense should have more of an effect on pBABIP but the rating was put in to take care of pitchers who have better BABIP because of their stuff, which I find it downright bizarre that the devs haven’t included (there’s a modifier based on velocity and perhaps on Stuff of a couple specific pitches but the literature I’ve seen is that this effect seems to be best correlated with just straight up K rate) as well as pitchers like Jim Palmer, whom the people who work on the historical engine apparently wanted because “changing players defensive ratings messes up the stats”, which of course is a ridiculous comment since even the best teams only get 27 outs and as long as player A has ratings that are in line on a general percentage basis with player B, you wouldn’t see individual stats go crazy like that.

That said, I think you’re correct in that it probably doesn’t do a huge amount one way or the other. And I guess as it stands, studies of the game indicate that fielding already has a really large factor on the game, so let sleeping dogs lie, I suppose…

So, I used to agree with you that Rivera-like outliers were the huge aberration and for the most part Stuff is Stuff, but having watched Gausman be the runaway best FIP/K/etc pitcher in baseball for the last two years (2.54 FIP, next best is 2.88; 9.5 fWAR, next is 8.6; 25.5% K-BB, next is 24.7%, etc) while having literally the second-worst BABIP in baseball (.350, only Patrick Corbin with his 4.85 FIP is worse at .351) has me reconsidering that stance.

In that timeframe the Jays have been 14th in OAA, 21st in UZR/150; so not amazing defensively (thanks Lourdes and Teoscar) but certainly not abysmal either.

LeeD · 07-27-2023, 11:18 AM

The more you tinker with "double dipping" on certain aspects of the game (like having Movement and then also PBABIP), the more unintended consequences crop up. PBABIP is highly dependent on the team's defense and also on the pitcher's profile (is he a groundballer or a flyballer?). When you add it as an independent variable, as one Discord user remarked, it becomes "one stat to rule them all." This is how you win Iron tournaments with lineups full of guys who hit 8 homers but drive in only 12 runs. The home runs are solo shots because everyone is fixated on PBABIP as the "one stat" to keep the bases clear of runners...and it works!

Read the filler comments when OOTP is loading, and the old Bill James chestnut "Much of what we think of as pitching is actually fielding." pops up. I always silently add "Much of what we think of as fielding is actually world-class athletes making routine plays."

There are both procedural and philosophical arguments underlying the whole PBABIP situation. James looked at PBABIP years ago and concluded that it was random happenstance; i.e., it varied year to year from pitcher to pitcher. He based this on studying the guys who were supposed to be good at preventing base hits on balls in play (like Greg Maddux) but found random variance rather than consistency.

This obviously doesn't sit well with students of the game, as there surely must be some pitchers who are able to induce soft contact, and easier outs, more often than others. However, even Gausman's career doesn't support this, as his PBABIP is all over the place, from .290 to .364 and landing at .317 career. Sure, a lot of that is team defense, but it varied from .290 to .337 in his early years with the O's, and he was playing for the same team with largely the same defense over those years. The only regulars who changed were the corner outfielders. Other than that, it was Chris Davis at first, Schoop at second, Hardy at short, Machado at third, and Adam Jones in center.

That's the procedural argument. Then there's the philosophy of pitching, which has significantly changed over the years. Today, pitchers throw as hard as they can for as long as they can. That wasn't always the case. Prior to Koufax, most pitchers saved their best stuff for when there were men on base or when the game was on the line. That would obviously skew the PBABIP numbers for early-era players, except for these two guys, who were the strikeout kings:

Name: Johnson.jpg
Views: 315
Size: 6.8 KB

Name: Johnson.jpg
Views: 315
Size: 6.8 KB

Walter Johnson and Dazzy Vance. But even here, Vance's PBABIB varies quite a bit. It's only Johnson's deadball era numbers that are very consistent, hovering around .260 until baseball changed the rules after the 1920 season.

This guy, of course, breaks the game:

Name: Feller.jpg
Views: 310
Size: 7.2 KB

Name: Feller.jpg
Views: 310
Size: 7.2 KB

Bob Feller's PBABIP numbers are remarkably consistent and remarkably low thanks to Cleveland's excellent defense. That makes him one of a kind in more ways than one; but you'll have to search far and wide to find someone who was that consistently dominant in terms of both stuff and PBABIP. For the most part, James was right. It's fielding and random variance.

07-26-2023, 05:52 AM	#1
kq76 Global Moderator Join Date: Nov 2002 Location: Vancouver, Canada Posts: 11,498	pBABIP Rating and Defense So I've long thought Stuff is less important the better defense you have, but what about pBABIP? You see people say that a bad pBABIP stat may be an indicator of bad defense behind the pitcher, but is that really true? I think of it as more an indicator that the pitcher's pitches may just be easier to hit hard. But either way, if you have a good defense should you be as concerned about having pitchers with good pBABIP ratings or not? I think the answer is probably yes, but I'm not sure. __________________ My OOTP Wishlist \| My FAQ List OOTP Wiki \| Your Recommended Team Nicknames, By City (A Crowdsourced Project)

07-27-2023, 11:18 AM	#4
LeeD All Star Reserve Join Date: Jan 2022 Posts: 624	The more you tinker with "double dipping" on certain aspects of the game (like having Movement and then also PBABIP), the more unintended consequences crop up. PBABIP is highly dependent on the team's defense and also on the pitcher's profile (is he a groundballer or a flyballer?). When you add it as an independent variable, as one Discord user remarked, it becomes "one stat to rule them all." This is how you win Iron tournaments with lineups full of guys who hit 8 homers but drive in only 12 runs. The home runs are solo shots because everyone is fixated on PBABIP as the "one stat" to keep the bases clear of runners...and it works! Read the filler comments when OOTP is loading, and the old Bill James chestnut "Much of what we think of as pitching is actually fielding." pops up. I always silently add "Much of what we think of as fielding is actually world-class athletes making routine plays." There are both procedural and philosophical arguments underlying the whole PBABIP situation. James looked at PBABIP years ago and concluded that it was random happenstance; i.e., it varied year to year from pitcher to pitcher. He based this on studying the guys who were supposed to be good at preventing base hits on balls in play (like Greg Maddux) but found random variance rather than consistency. This obviously doesn't sit well with students of the game, as there surely must be some pitchers who are able to induce soft contact, and easier outs, more often than others. However, even Gausman's career doesn't support this, as his PBABIP is all over the place, from .290 to .364 and landing at .317 career. Sure, a lot of that is team defense, but it varied from .290 to .337 in his early years with the O's, and he was playing for the same team with largely the same defense over those years. The only regulars who changed were the corner outfielders. Other than that, it was Chris Davis at first, Schoop at second, Hardy at short, Machado at third, and Adam Jones in center. That's the procedural argument. Then there's the philosophy of pitching, which has significantly changed over the years. Today, pitchers throw as hard as they can for as long as they can. That wasn't always the case. Prior to Koufax, most pitchers saved their best stuff for when there were men on base or when the game was on the line. That would obviously skew the PBABIP numbers for early-era players, except for these two guys, who were the strikeout kings: Walter Johnson and Dazzy Vance. But even here, Vance's PBABIB varies quite a bit. It's only Johnson's deadball era numbers that are very consistent, hovering around .260 until baseball changed the rules after the 1920 season. This guy, of course, breaks the game: Bob Feller's PBABIP numbers are remarkably consistent and remarkably low thanks to Cleveland's excellent defense. That makes him one of a kind in more ways than one; but you'll have to search far and wide to find someone who was that consistently dominant in terms of both stuff and PBABIP. For the most part, James was right. It's fielding and random variance.