With Pitcher BABIP now a thing, FIP and Pitcher WAR are kind of pointless

[uruguru]

FIP and Pitcher WAR in OOTP are premised upon a pitcher not being responsible for BABIP. But in OOTP, pitcher BABIP has a big effect on pitcher performance.


Should the game be switching to Run-based WAR (rWAR) when evaluating pitchers for things like All-Star appearances and post-season awards?


What do you think?

[MathBandit]

For historical play, probably- though the game probably should have been using ERA for historical play anyways.

For modern and fictional play, absolutely not imo.

[uruguru]

Originally Posted by MathBandit

For historical play, probably- though the game probably should have been using ERA for historical play anyways.

For modern and fictional play, absolutely not imo.

Is Pitcher BABIP only for historical? If so, that would make sense.

[Rain King]

Originally Posted by uruguru

Is Pitcher BABIP only for historical? If so, that would make sense.

Yes, there is an option to override the rating for non-historical in the editor, but by default it is just blank and isn't used.

The game does use a pseduo-BABIP rating for modern/fictional players, but it isn't a visual rating. It is dynamic/hidden, calculated using their pitch repertoire/ratings. Assigning the actual BABIP rating in the editor would override that.

[cheddah]

Originally Posted by Rain King

The game does use a pseduo-BABIP rating for modern/fictional players, but it isn't a visual rating. It is dynamic/hidden, calculated using their pitch repertoire/ratings.

Do we have any more detail on this?

[Rain King]

Originally Posted by cheddah

Do we have any more detail on this?

If you do a search there are a few threads where people have done some thorough analysis of pitch types and what their affects can be, but it isn't information the designers give out.

[Cactusguy21]

Originally Posted by uruguru

FIP and Pitcher WAR in OOTP are premised upon a pitcher not being responsible for BABIP. But in OOTP, pitcher BABIP has a big effect on pitcher performance.


Should the game be switching to Run-based WAR (rWAR) when evaluating pitchers for things like All-Star appearances and post-season awards?


What do you think?

This is actually an issue in real life. Stuff+ and BABIP tend to have an inverse correlation, meaning that if you have better stuff, you tend to run a lower BABIP. This is the limitation of FIP-based statistics.

But the run based stats have obvious issues too. That's why I include both WAR on my pages.

[uruguru]

Originally Posted by Cactusguy21

This is actually an issue in real life. Stuff+ and BABIP tend to have an inverse correlation, meaning that if you have better stuff, you tend to run a lower BABIP. This is the limitation of FIP-based statistics.

But the run based stats have obvious issues too. That's why I include both WAR on my pages.

What I like about non-FIP stats, at least for starters, is that pitchers are ultimately responsible for all of the runners that score to the same degree that batters are responsible for all of the hits they get.

And we don't think twice about judging batters based on a non-FIP stats.

[kidd_05_u2]

Originally Posted by uruguru

What I like about non-FIP stats, at least for starters, is that pitchers are ultimately responsible for all of the runners that score to the same degree that batters are responsible for all of the hits they get.

And we don't think twice about judging batters based on a non-FIP stats.

Lol. A pitcher has the same defense behind him every game, so failing to account for whether the defense is good or bad is dumb. A batter faces a lot different defenses, so it is much more reasonable to assume defense is neutral for his outcomes across a full season.

[Syd Thrift]

The reality is that pitcher BABIP will for the vast majority of guys be very close to 1.000 anyway and so you will probably want to continue to use FIP since it's more evocative of actual performance than straight ERA. I personally don't have it on any of my screens but that's specifically because I play in an early 1970s environment where a. these stats literally were not available and b. people didn't really think of BABIP in that way (although I think there's always been a sense among actual baseball people if not necessarily fans that pitchers affect the game primarily by striking guys out and avoiding walks and homeruns). I also don't have WAR listed (I do have RC/27, which is an 80s stat but I think still based on the conceptions of the general time period).

I'm not sure which version of WAR is used in the game. If it's bbWAR, then it literally just looks at runs/earned runs and innings pitched and then adjusts for context. That's just, basically, a different way of presenting ERA (and as a counting stat). You do you but I'd just, you know, use ERA. fWAR is based on FIP. BBRef explains the differences here:

https://www.baseball-reference.com/a...ed_pitch.shtml

[uruguru]

Originally Posted by kidd_05_u2

Lol. A pitcher has the same defense behind him every game, so failing to account for whether the defense is good or bad is dumb. A batter faces a lot different defenses, so it is much more reasonable to assume defense is neutral for his outcomes across a full season.

There is way too much noise in defensive outcomes to say anything is "much more reasonable" than anything else. The error bars are so wide that suggesting that batters benefit equally from neutral defenses is crazy. This is ultimately the problem with all FIP-based metrics.

Because in the end, we will judge hitters who benefit from a .320 BABIP against neutral defenses as better than hitters who get a .270 BABIP. Not just better, but significantly better even though a huge portion of that .050 BABIP difference is just luck.

"Lol."

[kidd_05_u2]

Originally Posted by uruguru

There is way too much noise in defensive outcomes to say anything is "much more reasonable" than anything else. The error bars are so wide that suggesting that batters benefit equally from neutral defenses is crazy. This is ultimately the problem with all FIP-based metrics.

Because in the end, we will judge hitters who benefit from a .320 BABIP against neutral defenses as better than hitters who get a .270 BABIP. Not just better, but significantly better even though a huge portion of that .050 BABIP difference is just luck.

"Lol."

Hitters have significantly more influence over their Babip than pitchers. This fact, combined with what I said above about defenses, is why FIP-like statistics are much more needed for pitchers than hitters.

This doesn't mean one shouldn't look into Babip, exit velocities, and other stuff for hitters, obviously.

[kidd_05_u2]

Run-based statistics for pitchers are also affected by sequencing luck. Yet another factor that doesn't come into play for the batter statistics that matter.

[md40022]

The best part about this is exit velocity factors into critiquing how an opposing defense affects BABIP as well (speaking IRL)...... and just a few threads down on this very same message board exit velocity has been written off by many as meaningless.

Ball coming off bat and going from point A to point B at a faster speed between Hitter 1 and Hitter 2 = Defender not getting to point B quickly enough to field hitter 1's ball, equaling a base hit for hitter 1. As where the same defender fields hitter 2's ball, which was hit to the exact same spot in the field, and throws him out = Higher babip for Hitter 1.

[uruguru]

Originally Posted by kidd_05_u2

Run-based statistics for pitchers are also affected by sequencing luck. Yet another factor that doesn't come into play for the batter statistics that matter.

sequencing luck? Are you suggesting that base situations have no effect on hitting and pitching effectivness?

[Syd Thrift]

Originally Posted by uruguru

sequencing luck? Are you suggesting that base situations have no effect on hitting and pitching effectivness?

There’s a world of difference between “no effect” and”a different effect for different pitchers”, which has yet to be demonstrated AFAIK. Certainly a lot of sequencing “luck” is in fact luck even though most pitchers are a little worse with a man on first. Also, in a lot of those cases a guy who allows a lot of baserunners on base will in fact have more of those men on base situations, which means that sequencing isn’t going to be so much of a factor.

Sequencing is absolutely an issue you see with relief pitchers and sequencing “luck” is often a thing that sees a guy have a low ERA one year and a high ERA the next. Again, if you want to use ERA, feel free. I don’t use FIP myself specifically because I want to reduce my access to modern knowledge about the way the game works. It’s certainly not useless and sequencing luck is a huge part of why it’s more reliable than straight up ERA.

[MathBandit]

Originally Posted by uruguru

sequencing luck? Are you suggesting that base situations have no effect on hitting and pitching effectivness?

Careful. It sounds like you might be saying that hitting with RISP is a skill lol, which I assume you don't mean.

[uruguru]

Originally Posted by MathBandit

Careful. It sounds like you might be saying that hitting with RISP is a skill lol, which I assume you don't mean.

No, "hitting with RISP is a skill lol" would be me suggesting that some hitters are better at it than others. There is no evidence in the statistical record of clutch hitting.

But of course the base situation is always going to affect pitching and hitting. It doesn't take a rocket scientist to see that. Having a force at a base, for example, can make the difference between an out and a hit. Having first base open with a runner on second encourages a pitcher to pitch around a hitter, increasing the chances of a walk.

[Syd Thrift]

In that respect, sure, FIP stats aren’t perfect but they’re stable more quickly than ERA because by concentrating on the peripherals they don’t ding a guy for unlucky sequencing. I’m not sure they’ll ever quite figure out how to account for situational hitting, exactly (although I suspect that the slow move to physics will mean that statheads of the future won’t feel the need to), but I don’t think perfect needs to be the enemy of good here,

[Rain King]

Originally Posted by Syd Thrift

In that respect, sure, FIP stats aren’t perfect but they’re stable more quickly than ERA because by concentrating on the peripherals they don’t ding a guy for unlucky sequencing. I’m not sure they’ll ever quite figure out how to account for situational hitting, exactly (although I suspect that the slow move to physics will mean that statheads of the future won’t feel the need to), but I don’t think perfect needs to be the enemy of good here,

Case in point.

https://fantasy.fangraphs.com/pfip-p...fip-framework/