Pitching WAR in PT

[usa_hank]

There have only been 7 MLB pitcher seasons since 1918 where a pitcher had a 5.5 ERA and a WAR above 2.0, with at least 100 IP. In my current HD league, there are 2.

There have only been 20 MLB pitcher seasons since 1918 where a pitcher had a 5.5 ERA and a WAR between 1.0 and 2.0, with at least 100 IP. In my current HD league, there are 5.

(for context, the FIP on the 7 players in the HD league that meet the criteria above ranges from 4.24 to 5.04)

I bring this up because I've been concerned about the pitcher WAR stat in PT. It has always felt, well, useless in evaluating pitchers in PT. It always seems like my pitchers end up with a high ERA, ridiculously horrible W/L records, but respectable 2.0+ WAR ratings. I get that ERA is not technically a component of WAR, and I'm guessing that the WAR calculation used in PT/OOTP uses FIP rather than RA/9. But ERA still generally correlates to WAR, and the above MLB examples are very much outliers. So it's concerning to see so many outliers in a single season of PT. WAR should be a fairly reliable measure of actual (not theoretical) pitching performance, and it's really not in PT.

I wonder why the FIP for these HD players is so much lower than the ERA, which isn't nearly as true for the outliers in the MLB. I wonder if this indicates that PT overvalues the effect bad defense has on actual game performance.

That said, I also wonder if the parity of the PT league affects WAR in strange ways, as level of opposition is a component of WAR. Anyone else have any thoughts, or done any looking into this?

[usa_hank]

Sorry, I'm a stat nerd and just looking for some statistical insight to help me really build a better team, since it feels like my guys are always underperforming to their counterparts across the league.

[cwhitman]

OOTP does use FIP based WAR, like fangraphs does.

There just aren't incredible pitcher seasons, due to the normalization and how cards are bunched up.

I upload high diamond and PEL league stats every week on the app: might be worth looking at that to see what cards are doing the best at the highest levels. if you're consistently underperforming, i'd look at your platoons, as well as depth charts/rest to make sure your players aren't playing tired regularly.

[usa_hank]

I've just seen some pretty big standard deviation between performance of certain starting pitchers. Pedro Martinez, Satchel Paige, and Phil Niekro come to mind here. I don't understand how they can be this inconsistent between teams without there being an external "team factor." It's not just WAR and ERA, it's FIP here too.

Yet there are pitchers who seem to perform more consistency for most teams in my HG/LD/HD leagues: Whitey Ford, Kevin Brown, Vida Blue, Carl Hubbell, etc. Heck, even live Tarik Skubal performs more consistently than PMart, Paige, and Niekro.

Part of this is trying to piece together why that is.

[cwhitman]

catcher defense/framing perhaps? i also think stamina is something a lot of players really disregard. just throughout my years, though everyone has over and underachievers, there are some high level players i see constantly overachieve, and some constantly underachieve, and i definitely think there's a skill diff there to some degree

[usa_hank]

Also want to throw out another statistical anomaly. My Dice-K had a ridiculous 6.58 ERA with a .365 BABIP through the first month of the season before I yanked him out of my roster. However, his FIP was an astonishing 3.32. His WAR was 0.7. What? His rWAR looks more accurate to his performance, at a -0.2. I believe his FIP was skewed because he only gave up 1 HR through that span, despite giving up a ton of runs.

It's a bit weird to see a BABIP that high and a FIP that low, but an ERA extremely high. Think about it - if I had kept him in, and he performed the very same way the rest of his season, he would have been around 10th in the league at a 4.2 WAR, but with a horrible 6.58 ERA and a .365 BABIP. That's just completely off the charts and unreasonable. This is just one case.

It feels like FIP may not an accurate evaluation of pitching performance in PT for whatever reason, which is why a FIP-based WAR looks inaccurate to performance as well, and why rWAR looks better here.

[LeeD]

Trouble is that FIP is not, and is not intended to be, a measure of pitching performance as measured by traditional stats like ERA. It's a measure of what past pitching performance should have been if all other things had been equal. It's mainly used by teams as a predictor of future pitching performance absent the things pitchers can't control (fielding, framing. ballpark factors. etc.) and assuming the pitcher's performance level doesn't change.

There are a lot of ifs in there, and a lot of outside factors that will continue to change the outcomes. Main culprits are the same things that changed the past outcomes, with PT adding an additional drag in the form of league placement and ratings creep over the course of the yearly cycle.

If the pitcher's ratings and FIP are better than his performance, look at those outside factors. If he's a contact pitcher (lower Stuff and PBABIP), team defense is going to play a larger role in his performance. If he's vulnerable to the long ball (lower HRA), playing in power-friendly parks is really going to hurt.

Also look at stamina. Starters with lower-than-average stamina need lower pitch count limits and quicker hooks. League play is normally modern era, so even workhorses from bygone eras aren't going to toss complete games with any kind of regularity.

Another factor is bullpen quality. Bad pens allow more inherited runners to score (affecting ERA) and also influence AI decisions about whether it's time to yank the starter (potentially forcing starters to keep pitching even if tired). Also, tired pens will cause problems because they'll underperform their own ratings, causing a domino effect.

The cures:

Defense, meaning high team Zone Ratings and high Defensive Efficiency.

Ballpark. Pitchers in pitcher-friendly parks are more likely to overachieve, although this is also true for the visiting pitchers and has a downward effect on team batting. Still, there are some exploits available here depending on particular pitching staffs and lineup construction.

Catcher quality. Probably overemphasized, but playing catchers with very low framing rates doesn't help.

Better pitchers. At some point, everyone is expendable. It's built into the PT framework. Launch day is going on four months in the rearview, and there have been a lot of cards added to the mix. This applies to both the starting pitchers and the relievers they depend on.

[cwhitman]

FIP is also useless in PT because babip is a real rating, thus "fielding independent pitching" doesn't really exist.

[usa_hank]

Right about FIP. However, pitching WAR should be an evaluation of pitching performance, or at least a way of comparing players. I should be able to look at the WAR of player X and compare it to player Y to see if he's better or worse. Because FIP seems lacking application here in PT land, using a FIP-based WAR doesn't help.

[cwhitman]

runs based WAR is a stat that you can add as a column