Thursday, September 09, 2004
A Critical Account of Joe Morgan's Guide to Winning Baseball
From Joe Morgan’s chat on August 20:
“Who said I didn't respect Billy Beane. I said we have a different philosophy on how to win in the postseason. They haven't won in the post season. We've had two teams in the last two years win the championship that were wild card teams. That tells me anyone in the playoffs can win. Just b/c I don't kneel down at Billy's feet, doesn't mean i don't like him. I like Billy personally very much, but until he can manufacture runs in the post season by playing some small ball and bunt and steal and hit and run -- produce and score runs -- I am not going to agree with his philosophy. Look at the history. There's only one way to win. I don't know what you want from me. You should think about how teams have won and realize THAT is the way to do it”
“Well, first of all, I'm not a perfect player in [Billy Beane’s] system. I stole bases, I bunted, I did the little things -- and so did my teammates at the Big Red Machine -- to score runs.”
Joe Morgan makes two important claims here that I would like to contest by reviewing the historical record:
- Joe Morgan is not a Billy Beane system player because he stole bases and bunted.
- “There’s only one way to win” in the postseason and that way is small ball which entails bunting.
Let’s start with the disclaimers. First, I’m doing this largely just for fun. I think the vague internet baseball community of sabermetrically-minded individuals tends to pick on Joe Morgan a bit much, but I don’t think we should cease to do so. I’m not interested in Joe himself, I just think his data collection methods had a significant flaw somewhere along the way and so his observations sometimes don’t match reality.
Second, there are two types of bunting: bunting for a hit and sacrifice bunting. If Joe Morgan is talking about bunting for a hit, I’m not sure how that fits into the definition of small ball anymore than swinging for line drives. Furthermore, it would either be impossible or so time consuming as to be worthless to endeavor to separate out which bunt-for-hits were unsuccessful. So for the purposes of this post, I’ll take bunting to be equivalent to a sacrifice hit (or an “SH” in the box score, which does not include sacrifice flies).
Third, there’s not, to my knowledge, sufficient evidence to look at the historical record and delineate what proportion of attempts to get a sacrifice hit were successful and what proportions resulted in fielder’s choices, pop-ups, strikeouts, or double plays. I won’t endeavor to analyze that aspect of the sacrifice.
Fourth, I will use the phrase “reached base” to refer to (H + BB (+ HBP when I remember to check)), even though this doesn’t measure the number of times a player actually has an opportunity to steal a base. This is a shortcoming, but a pretty minor one.
I’m definitely not the first person to comment on Joe Morgan’s repeated statement of preference for “small-ball” and the lack of empirical evidence behind it, but I think it’s important to note when he cites empirical evidence that does not support his claim.
Part One: Joe Morgan’s Sacrifice Bunts, 1972-1977
Joe Morgan’s career really took off after he was traded to the Reds before the 1972 season. In 1972, he had his first season with an OBP over .400 and an SLG over .400, and he would accomplish that feat in each successive year through 1977. Combined with his defense at second base, a case could be made for him as MVP in each of those seasons, as Aaron Gleeman noted . During that stretch, he hit over .300 only once, so it’s very obvious that he drew an awful lot of walks. Given that the A’s clearly value defense very highly (a look at the Defensive Efficiency Ratio leaders the last couple years can attest to this), the notion that Joe Morgan is a player Billy Beane would value very highly is pretty solid.
Joe Morgan’s argument that he’s not a Beaneball player contains two components: stolen bases and bunts.
A. The Stolen Base
Morgan was a tremendous base stealer: he finished either second or third in major league baseball during each season from ’72 to ’76 (and finished sixth in the majors in thefts in ’77), with only Davey Lopes and Lou Brock finishing ahead of him multiple times and with Billy North and Mickey Rivers each having one single-season total surpassing Morgan’s during that period. What’s more, Morgan was very successful at stealing bases when he attempted them, being caught only 73 times during that six year period versus 359 successful attempts, an 83% success rate, well above the 75% or so break-average point.
That might make Joe Morgan seem like the anti-Beane since the A’s don’t attempt many stolen bases. But, as Mike Carminati pointed out in his response to Morgan’s chat, Billy Beane only wants players to steal if they’re likely to have success. A major reason the A’s don’t steal frequently is that they don’t have players who steal a lot of bases when playing for other teams.
In the time they’ve been an above-.500 team (1999-2004), the A’s have only had eleven players above the Official John Jaha Stolen Base Total Cutoff Line (13 stolen bases in a season). Tim Raines played for the A’s in 1999, the year he turned 40, and he reached base reached base 55 times and had 5 SB attempts. Tony Phillips returned to the A’s in 1999 after a 9-year absence, reached base 170 times, and attempted 14 steals. Prior to that, he’d reached base 3344 times and attempted 277 steals. Randy Velarde was with the A’s in 1999 and 2000 at ages 36 and 37 and again at 39 in 2002, reaching base 346 times and attempting 30 stolen bases; for his career, he reached base 1634 times and attempted only 115 steals, a rate substantially below his rate with the A’s in his super-decline phase. Rich Becker was with the A’s briefly in 1999 and 2000, reaching base 80 times with 6 SB attempts after having reached base 726 times to that point in his career while attempting 83 SB at a 73% clip. Becker did, however, steal more frequently while with the A’s than he had in 1998 and in his 1999 stay with the Twins, and he attempted even fewer stolen bases with the Tigers in 2000 before ending his major league career. Johnny Damon reached base 231 times and attempted 39 stolen bases while with the A’s in 2001, down a bit from his 55 attempts after reaching base 280 times the previous year, but his stealing a bit less frequently with the A’s makes sense since his base-stealing skills were in apparent decline—he was successful only 69% of the times in 2001 versus 85% over the previous two seasons. Ron Gant had two brief stints with the A’s at ages 36 and 38, reaching base 30 times and attempting 2 steals, and he hadn’t attempted more than 20 stolen bases in a season since turning 31. Ray Durham was with the A’s briefly in 2002 at age 30, reaching base 84 times and attempting 8 steals, much less frequently than he did before that point but much more frequently than he has since leaving the A’s. Chris Singleton reached base 102 times as an A and attempted 9 steals in 2003, when he turned 31. Previous to that, he had reached base 717 times and attempted 99 swipes with a 74% success rate. Mark McLemore is a month from turning 40, and he’s 0 for 2 in stolen base attempts as an A, and he’s been on base only 80 times; McLemore was 18 for 28 on SB attempts in 2002 and 5 for 10 in 2003. Mark Kotsay has attempted 11 steals this year after attempting 11 in 2003, and Kotsay has attempted more than 20 steals in a season only once. This year, Eric Byrnes has swiped 13 bases for the first time in his career on a career-high 14 attempts.
I don’t think there’s anything in that record to suggest that the A’s significantly curb the stolen base attempts of good base-stealers on their team. One could infer, however, that the A’s are simply disinclined to pursue good base-stealers and thus would not value Joe Morgan. There are two primary responses to this. The general one is that the ability to steal bases is simply overvalued on the marketplace. The ability to steal bases is of fairly marginal utility in creating runs but is highly valued in the standard general manager’s evaluation process. The reverse is true of Joe Morgan’s other atypical offensive attribute, his ability to draw walks.
The second answer is that there simply aren’t many talented base stealers to be pursued. As of September 9, there are only 21 players in the major leagues with 20 or more stolen bases and only 8 with 30 or more, and three of those 21 players (Juan Pierre, Chone Figgins, and Brian Roberts) have been caught on over 30% of their attempts. What’s more is that Beane in 2001 traded for the player who had led the league in stolen bases the previous year and the A’s didn’t significantly curb his base stealing while with the team. Joe Morgan possessed a skill that Billy Beane has not pursued much, but very few players possess that skill and the few who do probably tend to be overvalued.
B. The Sacrifice Bunt
Mike Carminati already did a lot of the work here, providing a chart with the A’s SH% (sacrifice hits divided by plate appearances) from 2000-2004 in his aforementioned Joe Morgan Chat Day from August. In those five years, the A’s have had SH% of .40%, .39%, .32%, .36%, and .32% (was .28% when Mike did his chart), respectively. Only the Red Sox have a lower SH% this year (although I haven’t calculated the SH% for NL teams if you adjust for the pitcher’s hitting, nor do I plan to). The A’s clearly have an organizational philosophy that eschews sacrifice bunts.
Joe’s comments led me to believe he had a lot of sacrifice bunts. This is not, so far as I can tell, accurate. Morgan had 51 sacrifice hits in 11329 regular season plate appearances, an SH% of .45, only slightly higher than the A’s as a team have been. In his six year peak, he had seven SH in 3902 PA, an SH% of .18, nearly twice as low as the A’s lowest figure as a team. In 1975 and 1976 when the Reds won the World Series, Morgan had zero sacrifice hits.
Morgan’s implication that his sacrifices helped the team score runs seems dubious, at least for the six-year peak period I examined. (Be forewarned that this is, of course, a tiny sample size.) Four of the seven times Morgan sacrificed the Reds inning ended without scoring a run after the sacrifice. Two other times the Reds likely would have scored if Morgan had done anything but hit into a double play. On April 25, 1972, Morgan sacrificed Pete Rose to second, and Rose scored on Johnny Bench’s 2-out double. On September 22, 1972, Morgan sacrificed Rose to second and then Tolan and Bench followed with walks before Rose scored on a fly-ball and Tolan scored on a single. The one time Morgan indisputably aided a score was on July 18, 1974, when Morgan sacrificed Rose to second with one out in the first inning, and Rose scored on a single by Bench in a game the Reds lost by one run. That example probably demonstrates why even a successful sacrifice bunt that leads to one run scoring isn’t a terrific idea, since Morgan had a 40% chance of reaching base, which would move Rose to second while elongating the inning.
Morgan had two sacrifices in the playoffs in this span as well. The first occurred in the 1972 NLCS. In the first inning, Morgan sacrificed Rose to second. Rose was thrown out at third when the next batter, Tolan, hit a groundball to the shortstop (although Retrosheet notes there was an argument over the call). Tolan scored in the inning after a Johnny Bench single and a Manny Sanguillen throwing error. The Reds won the game, 7-1.
Morgan added another sacrifice bunt in the 1973 NLCS with none out and Pete Rose on first. Morgan reached on an error, but the Reds didn’t score in the inning.
Clearly, Joe Morgan was not someone who laid down a lot of bunts, and his claim that this aspect of his game makes him an imperfect player in Billy Beane’s system is erroneous.
Part Two: The Big Red Machine and Sacrifice Hits, Regular Season
Morgan made the contention that there is only one way to win in the postseason—namely, “smallball”—and referenced “The Big Red Machine” as a team that did the little things needed to win. I’ll use a liberal definition of “The Big Red Machine” that encompasses Morgan’s eight years with the team, 1972-1979. This team was very successful.
Here are the figures for the position players from those Reds teams (it’s really not fair to compare the sacrifice hits the Reds’ pitchers made to the DH-empowered A’s totals):
1972: 35 SH, 5554 PA; 0.63 SH%; .616 WPCT
1973: 36 SH, 5847 PA; 0.61 SH%; .611 WPCT
1974: 22 SH, 5910 PA; 0.37 SH%; .604 WPCT
1975: 26 SH, 5935 PA; 0.43 SH%; .666 WPCT
1976: 17 SH, 6073 PA; 0.28 SH%; .629 WPCT
1977: 17 SH, 5810 PA; 0.29 SH%; .543 WPCT
1978: 33 SH, 5674 PA; 0.58 SH%; .571 WPCT
1979: 31 SH, 5774 PA; 0.54 SH%; .559 WPCT
In general, the Reds sacrificed with slightly more frequency than the A’s have, but it’s definitely in the same ballpark. In 1976 and 1977, they did it less frequently than the A’s have during Beane’s reign, and in 1976 they won the World Series. This evidence certainly doesn’t suggest that sacrifice bunts are a key component of regular season success.
What’s more is that the Big Red Machine fit Morgan’s Beane profile (drawing walks and hitting home runs) quite well. They ranked first in the NL in walks in ’72, ’74-’76, and ’78-‘79 and ranked second in ’73 and ’77. In home runs, they ranked first in ’76, second in ’74 and ’78, third in ’75 and ’77, fourth in ’73, and fifth in ’72 and ’79.
Part Three: The Big Red Machine and Sacrifice Hits, Postseason
The Big Red Machine appeared in eight postseason series, winning five. Here’s the run-down of their sacrifice hits:
1972 CS (W): 3 SH
1972 WS (L): 5 SH
1973 CS (L): 3 SH
1975 CS (W): 0 SH
1975 WS (W): 2 SH
1976 CS (W): 1 SH
1976 WS (W): 0 SH
1979 CS (L): 1 SH
The Reds had 15 total SH, or 1.875 SH per series. Their record in games with SH was 8-5, and their record in games without SH was 14-10. Six of their sacrifice hits came in the 9th or later with a one-run defecit or a tie game. In 1975-76, the two years that they won the World Series, all 3 SH came in the 9th with a one-run defecit or a tie game and none out. Two were by fifth outfielder Ed Armbrister pinch-hitting for the pitcher and one was by Cesar Geronimo hitting in the 8th spot. Geronimo’s sacrifice moved Ken Griffey to second, aiding the Reds in scoring the go ahead run in game seven of the ’75 World Series. Armbrister’s sacrifice in the ’75 NLCS with a runner on first didn’t lead to a Reds score, and his sacrifice in the ’76 World Series with runners on first and second moved the runners over though they likely would have scored on Ken Griffey’s single anyway.
In the 12 times outside of ’75-’76 that the Reds had an SH, the lead runner scored only four times, and whether the run would have scored otherwise in each case is debatable.
The A’s have only had two sacrifice hits in their twenty postseason games in the Beane era, both by Ramon Hernandez. One occurred in the third inning of game four in 2000 with none out and runners on first and second; the A’s didn’t score in the inning but went on to win 11-1. The other occurred in game five against Boston last year with no outs, the A’s trailing by a run, and runners on first and second. The A’s, as you probably know, did not score in the inning. One wonders whether the A’s would sacrifice in other high-leverage situations, but they simply haven’t had any. In each of their two-run playoff losses, the A’s haven’t had anyone on base in the 9th with fewer than two outs, and they also had no one on base in the tenth or eleventh in the 11-inning game 3 loss in 2003. They also had no baserunners in the 9th in the 4-3 loss in game 4 in 2003. In Game 3 in 2001, they trailed 1-0 in the 9th when Jermain Dye doubled with one out and Eric Chavez coming to the plate. Chavez struck out, but I don’t see why anyone would have him sacrifice. In Game Five in 2002, the A’s cut the Twins lead to one on Mark Ellis’ one-out homer, but Terrence Long flied out for the second out before the A’s got another baserunner. In their game two win in 2003, they pulled even in the ninth when they decided not to sacrifice with runners on first and second with one out, though one could argue that they should have to put the go ahead run in position to score on Durazo’s single (not that that argument is particularly strong). In the tenth inning of that game they played some “smallball”, sending Hatteberg from first with one out and the result was a strike-em-out throw-em-out double play. They had a runner on first in the eleventh with none out and didn’t score. In the twelfth they had a runner on first with none out and did score. There’s no evidence to suggest that a lack of sacrifice bunts was the A’s downfall in any of their playoff losses.
Furthermore, in the two years the Reds won the World Series they won it in large part by doing the things Morgan says can’t be done in the playoffs: drawing walks and hitting home runs. They had 665 plate appearances in those four series and drew 59 walks and hit 18 home runs. That’s a walk every 11.3 PA and a homer every 36.9 PA. Over those two years, the Reds averaged a walk every 9.4 PA and a homer every 48.9 PA. Considering that the quality of pitching is better in the postseason, it’s pretty hard to conclude that hitting home runs, drawing walks, and only giving up outs in very high leverage situations is a strategy that can’t win in the postseason, and how Joe Morgan could apparently recall the secret to the Reds’ success as being other than continuing to get on base and hit the ball out of the park is for someone else to figure out.
Joe, if Billy Beane could sign you to a reasonable contract he would be positively giddy. He’s endeavored to win a World Series by assembling a team very similar to the Reds teams that you helped win World Series—teams that succeed not by high batting average and bunts but by drawing a ton of walks, hitting home runs, and playing solid defense. It’s hard for me to see why you insist that the philosophy you ascribe to him can’t work when it clearly worked for your ’75-’76 Reds, and I’m fairly positive that your 1975 World Series ring would rest on Carlton Fisk’s finger right now had Sparky Anderson agreed with you and ordered his players to lay down bunts and swing away instead of taking one of the 25 walks issued by Boston pitchers in that series.