Identical Lineups

[Josquin]

Originally Posted by SpacePope

A database of performance of pitchers across many seasons of many perfect leagues. Also, there is a formula you can apply to any pitcher that's extrapolated from that database, which I think is already discussed in a topic called Pitcher's Ratings or something like that.

Long story short, MOV > CTRL > STUFF. Maddux's ratings are insane.

Yeah, that's a linear regression to predict FIP from STU/MOV/CON, I believe? But how do you know that FIP is the stat you want to rank pitchers by?

[SpacePope]

Originally Posted by Josquin

Yeah, that's a linear regression to predict FIP from STU/MOV/CON, I believe? But how do you know that FIP is the stat you want to rank pitchers by?

If there is a better stat, then use the better stat.

In any case, even if FIP weren't the best stat to use it's, at least, a very good stat to use; and it has Maddux solidly in the top 5.

[chazzycat]

FIP makes sense, because it's basically how the game models pitching. The calculations for FIP involve just 3 components, and they are the same ones used for the main pitcher ratings: strikeouts (stuff), homers (mov), and walks (con). So using FIP is a pretty natural fit.

[Josquin]

Originally Posted by chazzycat

FIP makes sense, because it's basically how the game models pitching. The calculations for FIP involve just 3 components, and they are the same ones used for the main pitcher ratings: strikeouts (stuff), homers (mov), and walks (con). So using FIP is a pretty natural fit.

Below are the career stats for a Maddux in my current league. Can you guess in which season he had the lowest FIP?

[chazzycat]

Not with much confidence...I don't have all the formulas in my head....that's what FIP is for. Translating HR/9, BB/9, and K/9 into something on the same scale as ERA for easy comparison.

If I had to take a guess probably 2036

[Josquin]

Originally Posted by chazzycat

Not with much confidence...I don't have all the formulas in my head....that's what FIP is for. Translating HR/9, BB/9, and K/9 into something on the same scale as ERA for easy comparison.

If I had to take a guess probably 2036

Indeed. Looks like FIP doesn't translate very well into fewer runs allowed. I don't even have Maddux in my top 20.

[SpacePope]

Well, hard to argue against 6 championships. What are you using to evaluate pitching then?

[chazzycat]

Originally Posted by Josquin

Indeed. Looks like FIP doesn't translate very well into fewer runs allowed. I don't even have Maddux in my top 20.

it's hard to tell if you're joking...I hope you are...

I mean yeah, that season, his FIP didn't correlate to ERA. But with larger, multi-season samples it definitely will do a better job.

It seems abundantly clear the main driver of his ERA in 2036 was just giving up more hits. Hits are outside the control of the pitcher (as modeled in the game) and so it doesn't really make sense to look at that year as being bad. He was just unlucky.

[Josquin]

Originally Posted by chazzycat

it's hard to tell if you're joking...I hope you are...

I mean yeah, that season, his FIP didn't correlate to ERA. But with larger, multi-season samples it definitely will do a better job.

I am not joking.

It seems abundantly clear the main driver of his ERA in 2036 was just giving up more hits. Hits are outside the control of the pitcher (as modeled in the game) and so it doesn't really make sense to look at that year as being bad. He was just unlucky.

Could it be that he is more exposed to the vagaries of BABIP because his stuff rating is relatively low?

[SpacePope]

That's my understanding - Maddux will have a high BABIP in Perfect without very strong defense. But almost all the teams that can afford Maddux are also running elite defense.

EDIT: this is why I think tournaments will have less established metas. In most formats, unlike Leagues, you wont be able to afford elite defense without sacrificing significant offense. So the weight given STUFF, MOV, and CTRL will be more dynamic.

[Josquin]

Originally Posted by SpacePope

That's my understanding - Maddux will have a high BABIP in Perfect without very strong defense. But almost all the teams that can afford Maddux are also running elite defense.

BABIP is independent of the pitcher, so it has nothing to do with Maddux. But if you don't strike out a lot of batters, then a high BABIP will do more damage (because there will be more balls in play). And that's why STU is very important.

[chazzycat]

Originally Posted by Josquin

I am not joking.

Could it be that he is more exposed to the vagaries of BABIP because his stuff rating is relatively low?

Well now we are getting somewhere. Without a doubt, lower stuff would lead to more balls in play and hence greater susceptibility to swings in BABIP.

But in general that would all wash out with a large sample like the ones that were used to generate the FIP spreadsheet.

I also don't really think cherry picking one season of 13 where the FIP doesn't match results proves anything. It's proven FIP is more predictive than ERA.

If you're using something better than FIP, we'd all love to hear it.

[Josquin]

Originally Posted by chazzycat

If you're using something better than FIP, we'd all love to hear it.

At this point, all I think I can say is that pitching isn't necessarily "solved" in PT, and arguments could be made that FIP doesn't represent the only credible way that pitchers can be ranked. I've tried to measure the pitcher's run suppression ability more directly, but whether it's better than FIP remains to be seen.

[SpacePope]

Originally Posted by Josquin

BABIP is independent of the pitcher, so it has nothing to do with Maddux.

It has everything to do with Maddux, the fielders behind him, and the average BABIP of hitters he's facing. So yeah, not just Maddux, but is has more than a little to do with him.

[Josquin]

Originally Posted by SpacePope

It has everything to do with Maddux, the fielders behind him, and the average BABIP of hitters he's facing. So yeah, not just Maddux, but is has more than a little to do with him.

If that's the case, then which pitching rating controls the pitcher's portion of BABIP?

[SpacePope]

Originally Posted by Josquin

If that's the case, then which pitching rating controls the pitcher's portion of BABIP?

I get you. You're talking about the actual percent and I'm talking about the impact of his BABIP on how you rank him. But yeah, afaik, all the pitcher ratings do is determine how likely a PA results in a K, BB, or HR.

EDIT: actually, all I said was "Maddux will have a high BABIP in Perfect without very strong defense." My guy, you got us going in circles. Either tell us the secret sauce to 6 PT championships or not!

[chazzycat]

Originally Posted by Josquin

At this point, all I think I can say is that pitching isn't necessarily "solved" in PT, and arguments could be made that FIP doesn't represent the only credible way that pitchers can be ranked. I've tried to measure the pitcher's run suppression ability more directly, but whether it's better than FIP remains to be seen.

Fair enough. I agree with that. I don't base my decisions on the FIP spreadsheet, that's for sure. It can be a useful baseline reference though, and it's easy to calculate since it uses the same inputs as the game's pitcher model. That's the main point I was trying to make. FIP is not the end-all be-all of pitching metrics.

[humboldt1498]

Just to add fuel to this fire, my Maddux has a 3.04 ERA and a 2.54 FIP over 2053.2 innings in Perfect with an elite defense.

[Josquin]

Originally Posted by humboldt1498

Just to add fuel to this fire, my Maddux has a 3.04 ERA and a 2.54 FIP over 2053.2 innings in Perfect with an elite defense.

While my entire rotation has ERAs lower than their FIPs.

[chazzycat]

Yeah, one of my teams does that consistently too. Wanna know their secret? Set park factors to .9