ZIPS and the 2008 rotation

ever wonder how bob gibson's career would have looked if he'd spent all those years in a radical pitcher's park like petco? Beyond the Boxscore did; here are the results.

what-if pitching scenarios are the theme of the day around here. if you don't like lotsa numbers crunched, lotsa tables, and/or lotsa ZIPS projections, go away and come back some other time. as for the rest of you, get out the ol' pencil and cue up the back of an envelope; time to do some figurin'.

my goal here today is to gauge the cardinal rotation in its current state. how does it compare to last year's? can we expect it to be better or worse? by how much? and under what circumstances? since dan szymbroski has generously provided us, at this early date, with the cardinals' ZIPS projections, that's the dataset i'm gonna start with. i urge you not to take these figures as gospel truth; they're not presented in that spirit. the exercise is so speculative and omits so many variables that we can't trust it that far. however, even allowing for a wide margin of error, this kind of guesswork is not entirely without value; while any given scenario is unlikely to be very accurate, by examining a range of scenarios we can derive a useful set of parameters and probabilites.

so let's get started. here's how the cardinal rotation performed last year:

2007 ROTATION

the "others" include carpenter, mulder, maroth, wellemeyer, thompson, keisler, and troy percival (last day of the season, remember?). these results redefined awful; the starters in 2007 were 0.25 earned runs per game worse than the shockingly bad 2006 rotation, which itself had been thought to represent a nadir. we should remember that as we move through this exercise; although it might seem there's nowhere to go but up, things can always get worse. . . . . anyway, let's look now at the ZIPS projections and compare them to last year's rotation:

PROJECTED 2008 ROTATION

ZIPS projects the 5 returning starters to make 136 starts, leaving 27 for other pitchers. well actually this isn't quite true; pineiro is only projected to 22 starts, with 14 relief appearances. but since we know that's not pineiro's intended use pattern, it's illogical to accept that projection. so i'm going to assign him 27 starts and maintain the rest of the pitching line (you're getting an idea of how scientific this is . . . . ), which leaves him with a projected average of about 5.2 innings per start --- exactly what he averaged last year. we also know that carpenter will return at some point, i plugged him in for 8 starts, ignoring his ZIPS projection (which is ignorant of his health status) and applying wainwright's rate stats to him instead. for the "others" category, i simply assumed the same rate stats as last season and pro-rated them for the 18 unclaimed starts. based on these assumptions, the rotation ought to improve by 0.22 points through sheer inertia; that's better than a 20-run improvement, or 2 wins, without lifting a finger --- sheer regression to the mean.

now let's make a couple more assumptions. despite being no great fan of pineiro's, i think ZIPS is far too pessimistic about him; in this league and this ballpark, he ought to do better than a 5.28 era. so i'm going to assume that he matches his FIP (st louis only) from last year, 4.80. if we make this change, our projected improvement stands at nearly 30 runs, or 3 wins --- without adding any new personnel. this is hardly a radical assumption; regression to the mean is a very well-documented, and very powerful, tendency. and realize, a 30-run improvement would still leave the cardinals with a rotation well below the league average. the suggestion is not that the rotation will become good if we trot the same old stiffs back out there; it'll just be less god-awful.

ok, now let's start changing personnel. suppose the cardinals sign curt schilling; he bumps reyes (the low man on the totem pole) from the rotation. schilling's ZIPS projection isn't available, but his era's the last two years have been 3.97 and 3.87; if he stays in that range, this change improves the rotation's era to 4.60, and the bottom line improves by another 15 runs. the rotation is 4.5 wins better than it was last season.

suppose the new pitcher is aj burnett rather than schilling (and never mind how we get burnett; just go with it). ZIPS projects aj to a 3.79 era next season --- but that's in a hitter's park in a dh league. put him in busch vs nl opposition, and that projection looks more like 3.35. but let's be conservative and say 3.50 ---- he bumps reyes, and here's how the rotation looks:

now we're at about a 55-run improvement and close to a league-average rotation (the median nl rotation last year had a 4.55 era). intuitively, that makes sense to me: i'd expect a rotation of burnett / wainwright / looper / pineiro / mulder to be no worse than league average. but is a 55-run improvement in one year realistic? let's add a little context: between 2003 (when the cards' starters ranked 10th in the league) and 2004 (when they moved up to 4th), the rotation improved by roughly 45 runs. between 1999 (9th in the nl) and 2000 (3d), the rotation improved by about 60 runs. the cubs' rotation improved by about 90 runs between 2006 and 2007. this is not at all far-fetched.

the question is, where does that get us? don't forget, the cards' run differential (minus 104 runs) last year was a lot worse than their actual record; so even if we assume a 55-run improvement in the rotation, that leaves the team at minus 49 runs --- they would have to improve by 49 runs elsewhere (offense, defense, bullpen) just to break even. in other words, a 55-run improvement by the rotation --- whopping though it would be --- won't be nearly enough to make the cardinals into contenders. we're going to need every single one of these guys to beat his projection --- a 4.50 era from mulder, a 4.25 from pineiro, 4.60 from looper, 3.80 from wainwright, 3.30 from carp --- to get anywhere close to contention. if i throw the just-cited figures into the spreadsheet, the rotation era comes down to 4.28, a 75-run improvement over last year. that leaves the team within 30 runs of .500; if we assume further rosy scenarios (bounceback years from edmonds rolen and kennedy), then the cards have a winning club.

we can't say all of these things won't happen; we can certainly say a whole lot of things have to break the cards' way if they're going to contend in 2008.