Let's Debunk Myths About Short Rest

In case you've been living in a cave or under a rock, I've gone on record several times saying that the Yankees should use CC Sabathia on short rest throughout the playoffs.  He's their ace, their most durable pitcher, and throwing him out there as often as possible gives them the best chances of winning the World Series. 

But in this era of 5-man rotations, starting your ace on short rest is controversial.  The pros and cons have been debated repeatedly, but opponents of this idea usually fall back on something like this:

Usually, the ace of the rotation is the one pitching on short rest, yet the overall numbers for pitchers on 3 days rest still aren't good.  That's a terrible combination.  Starting on short rest must reduce a pitcher's effectiveness.

I just wasn't convinced that this was actually true, and thanks to the wonderful Play Index at www.baseball-reference.com, I can say that I was right.  It isn't true. 

There is no statistical indication that the variable of rest - in this case, specifically, comparing 3 days to 4 days - has any meaningful impact on how well or poorly a pitcher will perform on a given day.  Who you ask you to start on short rest matters, and how good they are matters.  So strap yourself in, and let the debunking begin!

Myth # 1 - If a manager needs to use a pitcher on short rest, he'll tend to go to his ace.   

False.

Here are the number of different pitchers who made at least one start on short rest by season for the post-strike era:

If the myth is true, this means that 106 "aces" pitched on short rest in 1995 (nearly 4 per team).  77 "aces" did it in 2004.  Even as the practice has fallen out of favor the past few seasons, anwhere from 40-60 different pitchers will still be asked to do it each year.  How good is the 60th best starter in baseball, you ask?  It's hard to say.  According to WAR, Scott Baker of the Twins was this year's 60th best starter, or if you prefer FIP instead, then it's Justin Masterson.  You get my point - these are good, serviceable, useful pitchers, but clearly not aces.  And neither of them pitched on short rest this year either, by the way.

Actually, of the 37 pitchers who did start on 3 days' rest in 2010, want to guess how many of them actually ranked in the top 37 in WAR?

Two.  Chad Billingsley and John Danks.

That means the rest of the WAR leaderboard, from Cliff Lee and Roy Halladay to Jason Hammel and Max Scherzer, 35 top starters in all, combined for zero short rest starts this season. 

In fact, over the past 5 years, only 40 pitcher-seasons have managed to rank in the top 70 in WAR and included a short rest start in the same season.  That might seem like a lot, but we're talking about hundreds of pitcher-seasons.  It's an average of 8 per year.  And if only 8 of the top 70 pitchers, on average, will throw on short rest in a given season, yet 40-60 total pitchers are doing it, we need to drop this notion that managers are only asking their best starters, because they clearly aren't.  They're actually going out of their way to avoid it.  Which brings me to my next point...

Myth #2 - The less rest a starter has, the less effective he'll be.

False.

Since the strike, here is the cumulative line for all major league pitchers:

And here is the total line for all short rest starts during that time:

The short rest ERA outpaces the short rest FIP, and over a relatively small sample (8,000 innings is small compared to 680,000), I'll give the FIP more credit, although that still makes the short rest stats 5-7% worse than their full rest counterparts.  But I'm not convinced this is as meaningful as it appears. 

Look at it this way: Major League pitchers combined to post a 4.07 ERA this season, but if you excluded Cliff Lee, Felix Hernandez, Roy Halladay, and most of the other 40 best pitchers from the tally, don't you think the new number would be quite a bit worse? 

What if you only ask bad pitchers to start on short rest?  We've already shown that elite pitchers rarely do it anymore (remember my first point), so, for the sake of argument, consider the 37 guys who did this year.  23 have them a career ERA+ under 100.  That's almost two thirds.  It's a group of pitchers that ranks as well below-average, so what would you expect to see?  Should we be surprised that short rest starters put up a FIP and ERA in 2010 that outpaced the league average by half a run when that group contains lots of guys like Jeremy Bonderman, Andrew Miller, Dontrelle Willis, Javier Vazquez, Nate Robertson, Kyle Kendrick, and Charlie Haeger, and virtually nobody like Cliff Lee, CC Sabathia, or Felix Hernandez?  Next you're going to tell me that Santa isn't real. 

That's probably why the short rest ERA is worse than the average ERA - good pitchers are rarely asked to do it, thus their stats are excluded from the sample.  And that's absolutely dumbfounding.  If there really is a mystique, a challenge, or difficulty surrounding short rest starts, why are managers almost always only asking their bad pitchers to do it?  Bad is bad, mediocre is mediocre, and no lack, abundance, or ample amount of rest is going to alter that condition very much.  However, if that's true, and I think you would agree that it is, then the converse must be true as well.  If a bad pitcher is bad on short rest, a good pitcher probably can do alright on short rest, which means teams really shouldn't be criticized  or second-guessed for starting their aces on short rest in critical situations, like a pennant race or the playoffs.  They should be criticized for not doing it.

Next time somebody flashes stats about short rest starts, proceed cautiously.  They have no bearing on how your team's ace will do, that is unless he's made 30-40 short rest starts of his own.