Monday, October 08, 2007

An Open Letter

Dear results-oriented sportswriters everywhere,

You are wrong. C.C. Sabathia was still the right choice to start today's game.

Allowing ten baserunners in five innings does not make the starting pitcher the "Chevrolet Player of the Game" or a good clutch performer. Paul Byrd remains a fringe fourth starter who didn't pitch particularly well and was lucky not to get lit up by a strong lineup.

Also, on behalf of the other 29 teams in MLB, thank you for your upcoming articles on why Alex Rodriguez should leave the Yankees.

Forever yours,



Paluka said...

Byrd's post-game interview was amazing.He was trying to tell the interviewer that he didn't pitch very well, but the interviewer simply would not believe him.

The Vegan said...

Well Byrd didn't pitch well but he survived and didn't get lit up.

In any case, thanks for the winner!!

Eugene said...

I agree Byrd didn't pitch well at all, but that's results-oriented too. Let's ignore the result and concentrate on the theory. In the past eight years, pitchers have had an atrocious track record pitching on 3 days' rest. And Sabathia was definitely not babied in game 1. Forgive me if I'm wrong, but I doubt your models account for the fatigue factor -- there simply isn't enough data from relevant times (the past decade) to make a meaningful model, so I think it's not as clearcut as your numbers say. We need more data on how 3 days' rest affects the starter. I actually put down big money on the Indians today, because I thought not enough was being made out of Wang's start on short rest.

Paluka said...

I agree with the thing about 3 days rest. We don't have much good data on the issue, but the data we do have is pretty pessimistic.

j holz said...

"We need more data on how 3 days' rest affects the starter."

pick up copies of The Book and Baseball Between the Numbers for a couple of good discussions of the subject, complete with contrasting viewpoints.

also, remember the media is feeding you whatever stats best support their agenda. alex rodriguez isn't the AL MVP, he's the choker who can't get it done in october. starters on three days rest in non-playoff spots have done far better than 29-52, but you won't hear about that after wang got lit up.

no one was talking about josh beckett suffering from this problem when he dominated in the '03 playoffs.

Jonnymagic said...

Yeah I actually think theres a lot of evidence that 3 days rest doesnt make a huge amount of difference, esp if you keep the pitch count low. I remember reading an article on bp awhile back that said the #s seemed to suggest that clubs were better off using 4 man rotations and limiting their starters to 100 pitches, both from a win expectancy and risk of injury perspectives.

Paluka said...

But I think there is a difference between pitching on 3 days rest once versus doing it throughtout the season. Pitchers seem to react poorly to having their routine changed in any way. This is why John Smoltz thinks that being a starter is easier than being a closer. You are able to fall into a schedule and your body adjusts to it. Not sure how real this is, but i think the evidence does support him.

Eugene said...

I've got copies of both the Book and BBTN. I still don't think we have enough relevant data. IIRC, the 3 day rest data includes a time period where people were more likely to start on 3 days rest, and when lineups weren't so fearsome, so a starting pitcher could expend less effort.

The relevant data to me would be in games where the starter normally doesn't go on 3 days rest, and in a modern run-scoring environemnt (1995+?). I'll be the first to admit we have too small a sample size, but the data so far does not look promising in favor of 3 days' rest.

I didn't know that teams with starters on 3days rest in the playoffs have gone 29-52, but I was looking more a stat that is more indicative of pitcher quality. Something more like ERA rather than team won/loss record.

And fwiw, I also made a bet that Arod would outperform his per-game Yankee regular season WPA in the playoffs. I don't really pay attention to the crap the media is spouting. I still lost my bet, but at least it wasn't completely Arod's fault -- Wang, Posada, and Jeter were far more damaging.

Spencer said...

I just heard Joe Morgan doing Game 4 on the radio, and of course he quoted that since 2000 playoff starters on 3 days rest are something like 9-21, 4.86. I've heard this stated a bunch of times already this postseason as if it's conclusive proof that you cannot start a guy on short rest. The W-L is bad, but that means nothing, and the 4.86 doesn't sound so great either, but what I want - need - to know is WHAT THE ERA IS FOR PLAYOFF STARTERS ON 4 DAYS REST SINCE 2000. Without that context the stat that Morgan and everyone else keep quoting has no significance.