Monday, November 2, 2009
Why You Shouldn’t Even Consider Starting Chad Gaudin over A.J. Burnett on Short Rest
Although the readers on this site are pretty much unanimous in agreeing with the Yankees starting A.J. Burnett in Game 5 on three days rest instead of Chad Gaudin on something like 20 days rest, in other places the sentiment is a little more mixed. So I figured I should try and quantify it and see if there's any merit to starting Gaudin tonight.As I posted about in this entry, Gaudin averaged an RA of 5.55 over six innings against the Phillies, and the Yankees lost 60.2% of the time when he opposed Cliff Lee.
According to a study done in The Book, pitchers on three days rest were five percent less effective than they were on four or five days rest. I have Burnett's current projection as an RA of 4.06, so if we make him five percent worse, figure he's more like a 4.26 RA guy on three days rest. I'll also decrease his durability rating on the assumption that he won't be able to pitch as deep into the game. And yes, I'm aware that he has good splits in his career on three days rest, but I'm also aware that I trust what rigorous research with the entire population of MLB starters tells us over a handful of starts by one specific pitcher.
So, if we assume Burnett's a 4.26 RA pitcher and pitch him against the Phillies 1000 times, here's how it compares to Gaudin.
| Pitcher | IP | H | R | ER | BB | K | HR | RA | W% |
| Burnett (3 DR) | 6.0 | 6.0 | 3.3 | 3.1 | 2.4 | 5.8 | 1.0 | 4.95 | 46.0% |
| Gaudin | 6.0 | 6.1 | 3.7 | 3.4 | 3.0 | 4.3 | 1.0 | 5.55 | 39.8% |
W%: Percent of games won with Burnett or Gaudin starting vs. Lee.
IMO, that's a big enough difference in win probability to make the idea of starting Gaudin a bad one.
Update: Here's a chart mapping out the full rest of the Series using the two scenarios as suggested by Standard Deviance.
| Gm | matchup | PhIwpct |
| 5 | gaudin/posada | .602 |
| 6 | burnett/pedro | .429 |
| 7 | pettitte/hamels | .478 |
| Phi Series Odds | .123 | |
| Gm | matchup | PhIwpct |
| 5 | burnett/molina | .540 |
| 6 | pettitte/pedro | .455 |
| 7 | sabathia/hamels | .416 |
| Phi Series Odds | .102 |
The last column (Phiwpct) is the Phillies win probability for each game and then the series in total. The first scenario is Gaudin in Game 5, Burnett in Game 6, and Pettitte in Game 7. Burnett and Pettitte get their normal projection since they'd be on full rest.
The second scenario is the one we're looking at right now, with Burnett, Pettitte and Sabathia all on short rest. Their projections are for them to be 5% worse to account for that.
So in the first scenario, the Phillies have a 12.3% shot at the Series, and in the second scenario they have a 10.2%. Further evidence that what the Yankees are planning to do is the right move.
Comments
Well, now you’ve done and made my whole HOF post moot. Thanks, SG. I’ll expect an apology on the apology thread.
IMO, that’s a big enough difference in win probability to make the idea of starting Gaudin a bad one.
I’ll preface it w/ I am 100% in support of starting AJ. However, when running the sims, did you start Posada w/ Gaudin and Molina w/ AJ? If not, could you (and post results separately instead of updating)? Then we could also see how much of a difference Posada vs. Molina would be.
I don’t think you can necessarily go with a large sample size for recovery time. Physical recovery time is something that varies greatly in the population but is consistent for each individual.
I wonder how the book developed it’s statistic. I’ll go try to find my copy.
Thanks, SG.
I’m in the “don’t punt a WS game camp,” but to really make the statistical case for AJ over Gaudin, wouldn’t you have to calculate not just the probability of game 5 in isolation, but the probability of getting one win out of the next 3 games if you start Burnett tonight vs that same probability if you start Gaudin tonight? So…
Scenario A (the short rest scenario): probability winning one game out of AJ vs Lee, Pettite vs Pedro, CC vs Hamels
vs.
Scenario B (the Mike Francesa scenario): probability of winning one game out of Gaudin vs Lee, AJ vs Pedro, CC (or Pettite) vs Hamels
Oh, if it’s limited to Gaudin vs. AJ I take AJ. But I’m in favor of Marte—>Joba—>Coke—>Hughes or something like that.
We use AJ today, we lose him for the rest of the series. We use the relievers today, we get them back for game 6.
And game 7 becomes: CC + Andy + Mo. I like that.
[5] Given the constitution of the Yankees pen, an all reliever game might work reasonably well, particularly in an NL park, but that’s just way too unorthodox a strategy to expect a manager to implement.
I don’t see why Burnett couldn’t pitch in relief in a theoretical game 7. And I don’t know why you’re so gung-ho about Pettitte in relief.
[7] Agree on both counts.
However, when running the sims, did you start Posada w/ Gaudin and Molina w/ AJ?
Yeah, I had Posada starting with Gaudin and Molina with A.J. The simulation was set up to pinch-hit for Molina in non-blowouts, so he essentially got between 2 and 3 PAs per game. Here’s what the different scenarios spit out.
<u>Posada + Gaudin</u>
4.70 RS/G, 5.75 RA/G
<u>Molino + Burnett</u>
4.53 RS/G, 4.92 RA/G
To continue the HOF discussion: Mike, I know you’re making funny in post [172] of the previous thread. However, I think Rice’s induction was produced less by Gammons and more by 1) steriod-era backlash, and 2)anti-sabermetrics backlash.
In other words:
“Rice did it all withourt steroids! Waaa waa!” and
“I hate the nerds who tell me he wasn’t as good as I think he was! Waa waa!”
That these two things could coincide at a time when Rice had been on the ballot for a while made him a perfect cause for conservative baseball thinkers to champion.
the readers on this site are pretty much unanimous in agreeing with the Yankees starting A.J. Burnett in Game 5 on three days rest instead of Chad Gaudin on something like 20 days rest
Nit-picking of course, but I’m not sure this is quite true. There aren’t that many of us here, and at least two are against and maybe a couple more think it’s a pretty close call. And while I’m at it, I’ll add that I’m not completely in love with it myself. I just think it’s the best option.
I’d rather not plan for Game 7 when we’re up 3-1. Lee will probably pitch well tonight but there’s a chance he doesn’t shut us down. He’s been very good, but it’s not like he has ‘99 Pedro type of stuff.
Isn’t the question 46% plus .54 x the probability of getting one win with Andy/CC against 40% plus .60 x the probability of gettig one win from AJ/CC with Andy in reserve for game 6 or 7
SG, did you take into account Gaudin only pitching 2.3 innings in the last five weeks? I’d have to think his effectiveness and stamina would be off his normal levels.
Also, does the Gaudin projection include his negative splits against lefties, or just the generic strngth of the Phillies lineup?
[11] For me the primary question revolves around how Burnett and Pettitte will respond to three days rest, and since this seems to be the question with the least quantifiable answer, you gotta go with AJ.
Also, what chance does Gaudin really have of going anywhere close to deep in a game he starts? I’m thinking he’d have five innings, at most. At which point, he’d have to relieve himself.
Of course, the vast majority of my angst is tied up in how well Lee pitched in game 1 and irrationally assuming that’s what he’s going to bring tonight.
I’m in the “don’t punt a WS game camp,” but to really make the statistical case for AJ over Gaudin, wouldn’t you have to calculate not just the probability of game 5 in isolation, but the probability of getting one win out of the next 3 games if you start Burnett tonight vs that same probability if you start Gaudin tonight?
I’d think any scenario that sets up a Sabathia start is going to be the best one, but let me see if I can figure this out.
Nit-picking of course, but I’m not sure this is quite true. There aren’t that many of us here, and at least two are against and maybe a couple more think it’s a pretty close call.
I’m about to ban everyone who’s against it to make my comment true. Those of you who think it’s a close call, choose wisely.
I say let Clay Bellinger pitch his way outta this curse!
Sabathia is the most valuable starting pitcher the Yankees have had since Ron Guidry.
46% is way higher than I would have guessed, which is good. The lineup and opposing starting pitcher make this feel like a total reverse lock.
I don’t see why Burnett couldn’t pitch in relief in a theoretical game 7. And I don’t know why you’re so gung-ho about Pettitte in relief.
Burnett will certainly be available for at least a couple batters in a possible Game 7. I would think it would be all-hands-on-deck, minus Pettitte, if he pitches a possible Game 6.
Since pitchers overall on 3 days rest are 5% less effective, then what exactly contributes to it? Loss of velocity? Arm fatigue? Less control of breaking pitches, and if so why? Maybe it’s something very hard to quantify, such as mental preparation.
What would really be great is a breakdown of the types of pitchers that have started on short rest and how they did. Do fastball/curve pitchers suffer more than fastball/changeup? or fastball/slider/cutter,etc?
It would also be an interesting study to actually survey a bunch of starting pitchers on how they think short rest actually affected them. If someone like Burnett says that he feels just as comfortable on three days rest, it’s really no question of who starts tonight. Give him the ball.
Wouldn’t it be great if the matchup actually was Burnett/Molina?
I think psychology may factor in here as well. I don’t know if it can be proved statistically, but all things equal, I’d say a team that’s up 3-1 has a better chance of winning Game 5 than they would a Game 7, simply because the confidence and momentum would be on the side that had been behind.
I think if Burnett tells you he his ready to go you don’t hesitate. You start him, especially with the off day tomorrow so the bullpen can pitch 3-4-5 innings today and still be good to go wednesday*
*Please let us hope we don’t need that game.
In the “red” scenerio, why does Pettitte get the start over CC?
So it looks like a Gaudin, A.J., C.C. rotation give Philly a 10.74% chance of winning. I just thought I’d throw that in there.
[24] i think that CAN be proven. I think i’ve seen somewhere that teams deep in a hole “give up” a bit and win less than they should.
I’m about to ban everyone who’s against it to make my comment true. Those of you who think it’s a close call, choose wisely.
That’s censorship!
At least have a poll first.
[27]
And I wonder how much having pettitte available in the pen for 6 OR 7 is worth….
[25] Unfortunately, since the Yankees have a -100% chance of winning tonight’s game, there’s a 200% chance that there will, in fact, be a game on Wednesday.
Why not Gaudin/Burnett/Sabathia? I get 0.107 from the above, which is > 0.102 but within noise. Of course it wouldn’t be surprising to see Gaudin greatly underperform due to rust and handedness.
Yeah, I was thinking they’d give Pettitte the start in Game 7 because of his place in team lore and the fact that he’d be on full rest, but logically, starting Sabathia on short rest instead makes more sense if it gets that far, and is probably what they would do.
Molina tonite-who didn’t know that
Molina tonite-who didn’t know that
I like our bench.
32 is right-if Guadin started it would be AJ/CC with Andy in the pen.
We better produce a lot of runs from 1->4 since 5->9 is Swish, no Ribby Robby, Gardy, Molino
Thanks for running the extra numbers, SG.
Interesting results, particularly if you give Sabathia game 7 in the red scenario. As [27] calculates, that makes the difference between the two probabilities pretty tiny.
Still, I’m in the AJ in game 5 camp: in the postseason, you try to put the ball in the hands of your best pitchers as much as possible.
[32] Oops, missed [27].
I don’t understand why Gardner should make a difference wrt Melky, even though the latter has been playing pretty well. It’s true he’s coming off the bench as it were, but from a rate POV he’s been not too far behind on offense this year and slightly better overall, last I checked.
Have you guys seen the awesome new meme that the Yanks are ruining what should have been a great World Series by winning it too quickly? Those bastards! Next they’re going to cuckold all of us, steal candy from our children and kick that nice old lady’s walker out from under her.
Where did Wojo get the $208M total salary number? Cot’s has them down for $201.5M this season. Now, I know that’s only $6.5M, but these millions start adding up to real money at some point.
I think it has to be picked up by more than one foolish ESPN scribe to be a meme, right?
Replacing Melky with Gardner is essentially a wash, with Melky a bit better offensively and Gardner a bit better defensively and in baserunning, although needing to use Gardner against a lefty is suboptimal.
Pena replaces Melky
Pena replaces Melky
I guess. Good for him.
Let’s go Yankees!
Next entry: Yankees (103-59) @ Phillies (93-69), Sunday, November 1, 2009, 8:20pm **Game Chatter**
Previous entry: 2009 World Series Odds after Game 4
There are currently 72 visitors who are not logged in.
There was a record 241 simultaneous visitors on May 2, 2011 at 11:54:25 pm.











