some questions and answers at the all-star break

1. how often has a la russa-managed team rallied from an all-star deficit of 7.5 games or more to make the playoffs?
once. the 2001 cardinals reached the break in 3d place at 43-43, trailing the cubs by 8 games and the 2d-place astros by 5. as late as september 7 they were still in 3d place, 6.5 games back. at that point they went on a 9-game winning streak (which began right before the september 11 attacks and resumed after baseball's week-long hiatus), followed that up with a 6-game streak, and finished the season tied with houston for the national league's best record. chicago finished 3d.

three other la russa teams overcame lesser deficits. the 1992 oakland athletics were 2 games back of the defending champion twins at the all-star break, with a 51-36 record; they went on to win the division by 6 games. the '89 athletics were in 2d place, 1.5 games back of the angels with a 52-36 record; they took the crown by 7 games. and tony's first playoff team, the 1983 white sox, were 3.5 games behind and in 3d place at the all-star break, with a 40-37 record; they blew past the angels and rangers in the 2d half to win the division by 20 games.

tony has never led a team from a sub-.500 record at the all-star break to a playoff berth. that leads to question #2:

2. how many teams have gone on to make the playoffs after reaching the all-star break with a sub.-500 record?
i only checked back to 1995, the beginning of the wild-card era. three teams did it in the inaugural wild-card year, which you'll recall was strike-shortened: the yankees (30-36), dodgers (34-35), and mariners (34-35) all overcame losing first-half records to play baseball in october. since then, it has only happened twice:

• the 1997 astros got to the all-star break 2 games under, at 43-45; they were in 2d place behind the division-leading pirates, who were at .500. houston went 41-33 in the second half to win the division with an 84-78 record.
• the 2003 twins were 44-49 at the all-star break, 7.5 games back of the surprising kansas city royals. they played .667 ball down the stretch to finish with 90 wins and a second straight division title.

three other playoff teams reached the recess at an even .500, all from the nl central: the 2004 astros, 2003 cubs, and aforementioned 2001 cardinals. another nl central team, the 2005 astros, was only one game over at the all-star break (44-43) but went to the world series. so if we want to be hopelessly optimistic about this (and why shouldn't we? we're fans), we could say that in 5 of the last 10 years, an nl central team got to the playoffs despite reaching the all-star break in the rough vicinity of .500. . . . if we want to take a more sober view, we'll acknowledge that only 2 teams have reached the break as far south of .500 as the cardinals and gone on to make the playoffs --- the 2003 twins and 1995 yankees. since the latter took place in a strike year, i don't know how applicable the precedent is. the storyline of the '03 twins is slightly more hopeful, but not much. the twins were chasing a kansas city team that was 10 games over at the break but had a pythagorean record of .500; their luck ran out down the stretch, and they went 32-38. the brewers' record more accurately reflects their real ability (they're only outperforming their pythagorean mark by 1 game), so they are not likely to oblige the cardinals by folding in the 2d half of the schedule.

a couple of teams came close to pulling this off just last season, by the way: the phillies were 7 games under at the all-star break (40-47) and remained in contention until the final week of the season, while the astros (there's that nl central again) got to the recess 3 games under (43-46) and weren't eliminated until the last day.

3. how many games has chris carpenter's absence cost the cardinals?
various ways we can estimate this; the simplest is to look at runs allowed. as a cardinal, carpenter has averaged about 7 innings per start and allowed an average of 2.6 runs (that's total runs, not just earned) per start. to date he has missed 16 starts, which (per the averages) translates into 112 innings pitched and 42 runs allowed. it's impossible to assign those 112 innings to a single replacement pitcher, because the rotation has been such a scramble and because many of the innings carp would have thrown have been picked up by the bullpen. so let's just use the staff average: the cardinals in the first half allowed an average of 65 runs per 112 innings pitched. if we simply take the difference (65 runs minus 42), we can estimate that carpenter's absence has cost the cardinals 23 runs --- approximately 2 games in the standings. to put this another way, if carp had taken his regular turns and all else remained equal, we might expect the cards to have a team era of about 4.32 (which would rank 9th in the league) and have a record of 42-43, 5.5 games out.

here's a second way to look at it. the cardinals have won 66 of carp's 94 starts since 2004; that's a .702 winning percentage. so they could expect to have won 11 of the 16 starts he has missed. how many games did they actually win in those starts? to answer this question, we'll actually have to untangle the rotation and identify the specific replacement starter(s) who pitched in carpenter's stead. for 3 games, it was keisler: they went 2-1 in those starts. then brad thompson took over; the cardinals have gone 8-2 in his 10 starts to date. that takes care of 13 starts, in which the cardinals went 10-3. the remaining 3 starts were picked up by some combination of wells, reyes, and wellemeyer; the cardinals collectively are 10-25 in their starts, so let's just call that 1-2. tally it up, and the team is right where it should be: they've won (roughly tallied) 11 of the 16 starts that carpenter missed, the same number they could have expected to win had he made the starts himself.

if that's a counterintuitive outcome, consider two factors. first, the cardinals scored 66 runs in the 13 starts made by thompson and keisler, or just over 5 runs per game; this, more than the pitching, explains the team's success behind the replacement starters. second, even if carpenter had been healthy, the cardinals still would have had their overall record dragged down by the fruitless starts of kip wells and anthony reyes, and the same combination of bad pitching / bad run support / bad fielding. i think it's extremely unlikely that either of those pitchers would have been dumped from the rotation any sooner had carpenter been pitching; on the contrary, la duncan might have felt more secure in the rotation's stability, and hence might have given greater latitude to the floundering starters. wellemeyer might not have been picked up at all; there wouldn't have been any need to claim him, because thompson would still be in the bullpen.

of course, carp's presence would likely have a ripple effect benefiting the whole pitching staff --- he takes the pressure off the bullpen, the other starters, the offense, everybody. that's a valid argument, and i would tend to agree with it. but the effect of his loss is probably not as dramatic as we might like to think. this team probably would be laboring even if carpenter had been healthy all year.

4. how much has rolen's slump cost the cardinals?
let's start by quantifying the depth of the slump:

			AB	R	H	2B	3B	HR	RBI	BB	AVG	OBP	SLG	|	RC	BR
2007			253	35	67	13	2	4	37	29	.265	.342	.379	|	34	35
2006 first half			284	59	94	28	0	14	57	30	.331	.398	.577	|	66	62
1st half career avg			263	43	76	19	1	12	49	34	.287	.375	.500	|	49	49

he's 8 runs and 12 rbi behind his career first-half averages; that's 20 runs right there, or 2 wins. the projections for base runs (BR) and runs created (RC) are consistent but show slightly more modest deficits, on the order of 15 runs. compared to his first half of last season . . . . well, there is no comparison. but there is, alas, an-all-too pertinent comparison between rolen's first half in 2007 and his 2d half in 2006:

			G	AB	R	H	2B	3B	HR	RBI	BB	AVG	OBP	SLG
2007			73	253	35	67	13	2	4	37	29	.265	.342	.379
2006			66	237	35	60	20	1	8	38	26	.253	.333	.447
total			139	492	70	127	33	3	12	75	55	.258	.337	.411

that looks like a typical year from todd zeile, who averaged 17 homers, 74 rbi, and 65 runs scored per 500 at-bats to go along with career averages of .265 / .346 / .423. zeile did have a long and successful career; a team could do much, much worse. but still . . . .

oh by the way, rolen is hitting .190 / .333 / .259 against left-handers this season; second straight year he's had a pronounced reverse split. freaky.