THE BOOK--Playing The Percentages In Baseball

Quick stats comparison.  I’m going to use BP’s PRAA because they break the season by team, which is easier for me to research quickly.

Sutcliffe - 18 PRAA (150 IP)
Alexander - 22 PRAA (88.3 IP)
RJ - 27 PRAA (84.3 IP)
Sabathia - 25 PRAA (130.7 IP)
Lee - 02 PRAA (79.7 IP)

Not sure if those numbers include the playoffs, however.  I suspect they do not, but I just wanted to point out I’d agree with any answer except Lee.  Shows how most people just vote for what’s fresh in their minds.

I’m a little young for Sutcliffe and Alexander.  I also have only the fuzziest of memories of RJ for some reason.  I remember some post-season dominance.

I was pretty torn between Lee and CC.  Lee was just so good in the playoffs, but the Brewers don’t come within 3 games of making it without CC stomping all over the NL before finally running into the Phillies AL quality lineup. 

I picked Lee.  Playoffs got greater weight than the regular season from me.

OK, I found out how to get the partial season splits and combine it w/ postseason data.  Felt a little lazy after my first comment.  So here’s a comp, postseason numbers included (note that we’re switching stats, to Fangraphs’ WPA, which accounts for the leverage of a given situation):

Sutcliffe - 1.26 regular season WPA (not found for postseason, which was 13 IP, 3.38 ERA)
Alexander - 1.99 WPA (regular season only; not found for his postseason, but he had a bad postseason and would be negative wpa i think)
RJ - 3.03 WPA (3.01, 0)
CC - 2.82 WPA (3.17, -.35)
Lee - 1.66 WPA (.20, 1.46)

Is including leverage really the right thing to do?  These are all starters; they all entered each game with the same leverage (or close to it, depending on which was home and away).

Probably not the best thing to do, but A Team made a comment about weighting playoffs more heavily.  In retrospect, no, it’s not really that relevant, but it still shows how far Lee lags behind some of the other guys (unless you REALLY weigh playoffs, at like 2x regular season games).

I only used that because a) coffee hadn’t fully kicked in, and b) it was an easy counting stat that I could pull from fangraphs that included same-season splits, and playoff numbers.

Obviously all of this can’t be attributed to Randy Johnson, but before his first game on the Astros, they had played .600 ball. From August 2nd (his first start for them), through the end of the year, they played .692 ball.

Tom: you ask two different questions here, and depending on which one you’re looking for it may change my answer.

The poll asks for “best pitcher performance,” which is asking solely about how well the traded pitcher pitched. But in the body, you ask, “who was the best mid-season pitcher acquisition?”

The best performance in my mind goes to Randy Johnson in Houston. But they were up by 3.5 games at the time of the trade and finished the season 13 games up. They probably would have made the playoffs without RJ.

The best acquisition might be CC Sabathia because of how much he had to do to get the Brewers into the playoffs, pitching on 3-days rest several times at the end of the season. A 1.65 ERA in 130 innings doesn’t hurt his case either.

I really weighted playoffs then, as the mental adjustment I made was probably more than 2x a regular season game.  Obviously I didn’t look at any numbers, when answering questions like these where the correct answer isn’t important, I like to trust my gut and then see where it got me afterwards.  I think of it as a test.

The mental adjustments I made were to basically try to estimate the amount of dollars the player was worth over the guy he replaced.  I (perhaps unfairly) dismissed Sutcliffe and Alexander out of hand since the years in question were my age -3 and 0 years.  I then pretty much dismissed RJ because I only remember both him and Schilling dominating the playoffs and could not estimate a dollar sign.  So I kind of bastardized the poll and lopped off 60% of it.

That said, Lee replaced Moyer or Rodrigo Lopez and contributed something like 2.5 wins above those players.  Toss that on the win curve and that’s wins 91.5, 92, and 93...which is right on the sweet spot of the curve.  I decided to call each of those wins worth an average of 5 mil or 12.5.  I’m not trying to be accurate with those numbers, just create a relationship that can be used to compare Lee’s value to the Phillies to CC’s value to the Brewers.  I don’t know how to calculate Wins Above Rodrigo in the playoffs since he was really taking starts from Blanton and Happ, but I arbitrarily decided he was worth about 1 win and that win was worth 6 mil.  So my total for Lee was 18.5.

To quickly go through CC, I thought he was worth between 3 and 3.5 wins in his stomping of the NL with a slightly negative value in the postseason.  So I’m going to say 3.25 wins * 4 = 13.  I arrived at 4 instead of 5 because Milwawkee is a smaller market and they were at a slightly earlier part of the win curve (86.75-90).  4.5 might be a fairer comparison, but the answer doesn’t change.

And of course this mental calculation was done in a moment and not half as thought out as this.  If I had my vote back it would go to RJ I think.

Now you know more than you ever wanted to about how I answered the question.

See, I’m not even remembering the right season for Johnson...I was wondering why I couldn’t remember the D-Backs trading for him.

I went with CC.  I’m a little surprised that Sutcliffe isn’t getting a bit more love in the poll, though.

I voted for Sutcliffe because I assumed he had a huge advantage in innings pitched over the other guys. I voted based purely on memory of the performances without looking up any numbers though. I’d probably change my vote to Sabathia based on my criteria.

I’d count it like this for the playoffs:
2x: round of 16
3x: round of 8
4x: round of 4
5x: final round

So, all World Series games count as 5x.  The LCS count as 4x.  And the division series count as 3x.  And, if ever they add more teams, or you want to do this for the other leagues, you count those games as 2x.

Where’d I come up with those numbers?  Half from my b-tt, and half based on how much I think we’d pay for tickets in each round.

Maybe it should be 1.5, 2, 3, 5?  Something like that…