just in time for the latest news about the cards' april rotation: a statty look at something called "starting-pitcher leverage" over at the Hardball Times. what's SP leverage?
according to derrick goold's sunday post at Bird Land about the rotation, tony and dave are practicing a form of leveraging. "The rotation is set as much for who pitches against what team as who doesn't pitch against certain teams or in certain ballparks," goold notes; this remark from La Tony reinforces the point:
carpenter | mets | astros | pirates | pirates | cubs | cubs |
wells | mets | pirates | brewers | cubs | reds | brewers |
looper | mets | pirates | pirates | cubs | cubs | brewers |
wainwright | astros | brewers | giants | reds | cubs | astros |
reyes | astros | brewers | giants | reds | brewers | astros |
THT's formula is a simple one: you simply take the average of a given pitcher's opponent teams' winning percentages, and divide that by the winning percentage of all opponents on the team's schedule. let's carpenter as an example, applying 2006 winning percentages just to keep it simple. his 6 opponents had 2006 winning percentages of .599, .506, .414, .414, .407, and .407 --- an average of .458. now divide that by the winning percentage of all st louis opponents --- the cards' first 30 opponents had an average 2006 winning percentage of just .468 --- and you get a quotient of .98. so carp's first six opponents are 2 percent less difficult than the average cardinal opponent during that stretch.
as measured by 2006 winning percentages.
but obviously, this isn't 2006; we just used those figures as an example. to make this slightly more meaningful, let's apply some reasonable projections of 2007 winning percentages and see which pitchers are being leveraged by that standard. i'll go ahead and use the w-l percentages posted in RLYW's Diamond Mind projection blowout last week; i trust those numbers as much as i trust any. here are RLYW's projections for the cards' first 30 opponents:
opp | games | est wpct |
---|---|---|
mets | 3 | .525 |
cubs | 6 | .519 |
brewers | 6 | .500 |
astros | 5 | .494 |
giants | 2 | .488 |
pirates | 5 | .457 |
reds | 3 | .451 |
over the 30 games, that averages out to a .492 winning percentage. now let's look again at which pitchers are facing which teams, and apply the THT leverage formula:
1 | 2 | 3 | 4 | 5 | 6 | avg | lev | |
---|---|---|---|---|---|---|---|---|
carpenter | mets | astros | pirates | pirates | cubs | cubs | .495 | 101 |
wells | mets | pirates | brewers | cubs | reds | brewers | .492 | 100 |
looper | mets | pirates | pirates | cubs | cubs | brewers | .496 | 101 |
wainwright | astros | brewers | giants | reds | cubs | astros | .491 | 100 |
reyes | astros | brewers | giants | reds | brewers | astros | .487 | 99 |
well, wasn't that a worthwhile exercise. . . . . . i'm so glad i went to the trouble. lest the whole thing (and this whole post) be a complete waste of time, let's make a logical adjustment here. leverage can be measured by things other than raw winning percentage. in this case, while the mets have the best projected winning percentage on the st louis schedule, they are definitely not the team the cardinals want the most leverage against --- because they play outside the division. the top nl central rivals --- cubs, astros, and brewers --- are the teams st louis wants the most leverage against. (i'll grant that the cards and mets might end up vying for the wild card --- but the same could be said of all the in-division foes, so that's a wash.) to reflect the greater significance of the games vs the cubs astros and brewers, let's assign a premium --- wild-ass guessing, i'm going to make it .050 of winning percentage --- and recalculate. here's the table, one last time:
1 | 2 | 3 | 4 | 5 | 6 | avg | lev | |
---|---|---|---|---|---|---|---|---|
carpenter | mets | astros | pirates | pirates | cubs | cubs | .520 | 100 |
wells | mets | pirates | brewers | cubs | reds | brewers | .517 | 99 |
looper | mets | pirates | pirates | cubs | cubs | brewers | .521 | 100 |
wainwright | astros | brewers | giants | reds | cubs | astros | .524 | 101 |
reyes | astros | brewers | giants | reds | brewers | astros | .521 | 100 |
dammit . . . . . all that pencil-scratching for nuthin'. that's not to say that tony n dave aren't seeking to exert leverage; they're simply applying different levers from the one described here. e.g., "the astros can't hit a curveball, so let's start wainwright vs them twice," or "wells has a 3.06 career era vs the cubs," that sort of thing . . . . . i prob'y should have spent my time looking into those. have at it in the comments thread, if'n you're interested.
i guess the most significant thing about the way the rotation has been set up is this: chris carpenter will get an extra day's rest 5 times in his first 10 starts. as we discovered during the postseason last year, carpenter thrives on extra rest. his era on extra rest last year (in 7 starts, including world series game 3) was 0.82; his whip was 0.585. he was completely unhittable on 5+ days' rest.
the other great thing about the rotation: it only has to cycle through one more time before the real games begin. st louis plays the marlins today; gameday link here.