confidence game

yesterday's ballgame spurred me to investigate something i've been wondering about most of the season. of the 6 runs the cardinals yielded yesterday, 5 --- including all 3 of the ones that decided the outcome --- scored with 2 out. the walkoff single on friday night also came with two outs; ditto the walkoff single in pittsburgh on thursday. all year long, it seems like, the cards have been an out or a strike away from getting out of an inning . . . . . but by the time they finally do get out of it, 3 or 4 runs have come home. because these occurrences are so aggravating, i've thought that maybe we exaggerate the magnitude of this problem --- maybe we only remember the failures while taking the successes for granted. so i looked up some numbers, courtesy of the awesome team-splits data that Baseball Reference rolled out this year.

the split i focused on is the one where the cardinals lost their games yesterday, friday, and thursday: 2 outs w/ runners in scoring position. they aren't losing these games on walkoff homers; they're losing them on singles that knock guys in from 2d and/or 3d. zeroing in on that split, i looked first at the league-wide data. turns out that hitting drops sharply in 2o/risp situations --- these are 2007 national league numbers:

before we go on: you'll note that even though batting averages drop by 20 points with 2 out and RISP, on-base percentages increase by 29 points. that's because so many intentional walks come in 2-out RISP situations --- nearly 60 percent of all ibbs. if we subtract those plate appearances from the equation, the table makes a little more sense:

in plate appearances where they're trying to get the guy out, nl pitchers perform slightly better in 2-out RISP situations than in all other ones. why? maybe pitchers are more apt than usual to pitch around good hitters in those situations; maybe managers are more apt to play matchups in those plate appearances (although the hitters' managers would be just as likely to do the same, canceling out any advantage to the pitchers). perhaps the pitchers "reach back for something extra," as the cliché goes; a few more mph on the fastball, a little more bite on the breaking pitches. that would be an interesting study. . . . in any case, the splits for 2006 are roughly the same as the ones for 2007; B-R's leaguewide numbers don't go back any further than that.

in all situations other than 2 outs w/RISP, the cardinal staff is just a little worse than league average:

these are meaningful differences, to be sure; over that many plate appearances, they cost the cardinals about 20 runs over the average team. but now compare the cardinals' performance in 2-out RISP situations vs the league's:

great googly moogly. it should go without saying (but i'll say it anyway, just to drive home the point) that the cardinals have the league's highest batting average allowed in 2-out RISP situations; the next highest belong to colorado and florida (.272). in these 451 plate appearances, the cardinal pitching staff has coughed up 32 runs more than a league-average staff would have yielded; it is here that the stl pitchers have gone from substandard to truly awful. note that while nl pitchers, as a group, pitch slightly better with 2 out / RISP, the cardinals pitch considerably worse:

it's as if they've had latter-day matt morris out there for the run-of-the-mill plate appearances, but for the make-or-break ones they bring in jason simontacchi. what could explain it? one possible clue lies in this post, where i compared anthony reyes' pitch selection with RISP to his selection in all other at-bats. looks like i'm gonna have to revisit the issue, but expand the investigation to look at all the team's pitchers.

we might also pin the blame on the way-below-replacement-level guys who've absorbed so many beatings for st louis this year --- maroth, jimenez, cavazos, et al. but the truth is, those guys stink in all situations; there's little situational variation in their numbers. the guys with the big variations are --- well, you can probably guess:

maybe you anthony reyes skeptics can see why so many of us think he is redeemable. in the "all other" category --- 90 percent of the batters he has faced --- he is a solidly above-average big league pitcher. in the tiny sample --- 26 at-bats --- with 2 outs and RISP, he's charlie brown. fix that small problem, and you've got yourself a decent pitcher. to reiterate, these are not the cards' worst pitchers in 2-out RISP situations; they're merely the ones with the biggest differentials, the widest situational splits. wells and brad thompson have been terrible in 2-out RISP situations, but they've been almost as bad (in thompson's case, worse) in all other situations. reyes flores and johnson account for nearly 20 percent of the team's 2-out RISP at-bats --- and, accordingly, they account for the lion's share of the matt morris / jason simontacchi effect. only one of the cardinals' starting pitchers has truly been good in 2-out RISP situations: wainwright, the former setup man / closer, who is holding opponents to .223 / .350 / .269 in the key at-bats. . . . unfortunately, he is getting raked in all other at-bats (.298 / .351 / .431).

had enough? hold on, amigo; there's more. if you think the cardinal pitchers choke with 2 out and RISP, get a load of what the cardinal hitters are doing this season:

their batting average falls 60 points in those crucial 2-out rbi situations. in all other plate appearances, the cardinals bat 10 points above the league average and are within shouting distance of it in slugging; with 2 outs and RISP, they're sub-sub-replacement level. to reprise the previous analogy: it's as if mark grudzielanek or john rodriguez has been taking 90 percent of the cardinals' at-bats, but gary bennett has been pinch-hitting in all the big 2-out situations. . . . believe it or not, there is one team with a lower 2 out/RISP batting average this season: the florida marlins.

to put a cap on this, let's place the cards' performance side-by-side with that of their opponents:

by my rough calculations, as a team the cardinals are about 30 runs below average in the "all other" category (which comprises nearly 90 percent of all plate appearances) but 70 runs below average in the handful of at-bats that come with 2 outs and RISP. among individual players, the 2-out RISP sample sizes are so small that we could attribute any performance differential to luck (or "random variation," if you prefer) rather than skill. but at the team level, i don't think so; the sample size is too large. i think there's something going on here, something psychological. i think the cardinals, for whatever reason, have gotten anxious in big situations this season --- haven't made good pitches, haven't taken good at-bats. confidence is not a mirage; states of mind can and do impact performance, on and off the baseball field. tony's DUI, the season-opening punking at the hands of the mets, carpenter's injury, hancock's death . . . . the cards' state of mind took a beating very early this year, and it has never recovered.

but the main thing is, it wasn't a very good team to begin with --- 30 runs below average in your run-of-the-mill at-bats. that's the biggest factor undermining the cardinals' confidence: without carpenter, they don't have a chance. and the players know it.