Monday’s column on the fate of World Baseball Classic pitchers triggered a surprisingly skeptical reaction from the readers. I say “surprisingly” because the type of material that I deal with in this space is usually too technical to get anybody really riled up. It might have been a selective sampling issue–it’s a lot more fun to write a columnist if you have a beef with what he’s saying. But the reaction on the internal BP mailing list was pretty skeptical too.
I think there is a little bit of denial going on here. The starting pitchers involved in the World Baseball Classic posted a 5.49 ERA over more than 735 collective innings in April. This is a very strong fact. That is a lot of performance, equivalent to three or four full seasons worth of performance for a durable starting pitcher. If you took a pitcher with a 4.24 ERA (the weighted average PECOTA projection of the WBC starters) and had him throw 735 innings in a Diamond Mind Baseball simulation, the odds of his finishing with an ERA of 5.49 or higher over that stretch would be many thousands to one against. Even a simple binomial distribution tends to support this. The odds of 19 of a group of 26 pitchers coming out on the high side of their weighted mean PECOTA projection on the basis of randomness alone are about 200-to-1 against.
Now, none of this logic holds if PECOTA is the source of the error, rather than the pitchers themselves. But if we take this same group of starting pitchers and simply take the weighted average of their major league ERA in 2005 (this excludes Lenny DiNardo, who hardly pitched in the majors last year), we come up with a 4.05 ERA, about twenty points lower than their collective PECOTA projection. These pitchers are performing about a run and a half worse than they did one year ago. Something rather dramatic has happened with this group of pitchers, and the only common thread that I can deduce between them is that they pitched in the World Baseball Classic.
The other, less skeptical line of questioning concerned what effect the WBC has had on the position players who participated. I’m going to run through the roster of 77 hitters in one fell swoop, comparing actual results against PECOTA. The statistic of choice is MLVr since, with a little bit of algebra, we can easily transform it to a plus-minus rating that represents the number of runs that a player has performed above or below his projection thus far on the season. (Another convenience is that MLVr is calibrated around league average, so the higher offensive levels that we’ve seen so far this year will automatically be accounted for). All statistics are through Monday night’s games.
Player Team PA MLVr- Actual MLVr- Pred. +/- ----------------------------------------------------------- Suzuki JPN 155 -.068 .025 -4 Koskie CAN 92 .110 .012 +2 Morneau CAN 111 -.201 .095 -8 Orr CAN 40 -.477 -.112 -4 Bay CAN 143 .189 .227 -1 Guiel CAN 15 -.870 -.135 -3 Stairs CAN 40 -.264 .043 -3 Stern CAN 21 -.660 -.064 -3 Amezaga MEX 31 -.271 -.208 -0 Cantu MEX 64 .092 .013 +1 Castilla MEX 110 -.230 -.073 -4 Castro MEX 86 -.343 -.145 -4 Barrett USA 96 .134 .020 +3 Schneider USA 107 -.279 -.121 -4 Varitek USA 103 -.077 .117 -5 Jeter USA 138 .398 .087 +11 Jones, C USA 80 .183 .198 -0 Lee, D USA 58 .487 .282 +3 Rodriguez, A USA 133 .145 .310 -5 Teixeira USA 153 .080 .252 -7 Utley USA 129 .329 .123 +7 Young USA 157 .200 .129 +3 Damon USA 141 .179 .049 +5 Francoeur USA 133 -.190 .043 -8 Griffey USA 36 .120 .204 -1 Holliday USA 140 .241 .154 +3 Wells USA 139 .451 .127 +11 Winn USA 138 .015 .049 -1 Jones, A NED 130 .266 .156 +4 Saenz PAN 55 .471 .017 +6 Lee, C PAN 140 .464 .128 +12 Lopez PUR 74 -.204 .061 -5 Molina, Y PUR 106 -.553 -.130 -11 Rodriguez, I PUR 119 .113 .017 +3 Valentin, Ja PUR 58 -.131 -.018 -2 Cintron PUR 51 -.239 -.052 -2 Cora PUR 32 -.324 -.123 -2 Delgado PUR 139 .394 .198 +7 Perez, E PUR 51 .438 .127 +4 Valentin, Jo PUR 29 -.661 -.101 -4 Beltran PUR 93 .530 .122 +9 Cruz PUR 108 -.030 .043 -2 Ledee PUR 33 .001 -.037 +0 Matos PUR 39 -.610 -.032 -6 Rios PUR 109 .628 .003 +17 Williams PUR 89 -.210 -.049 -4 Paulino DOM 50 -.206 -.047 -2 Belliard DOM 128 -.107 -.047 -2 Beltre DOM 136 -.273 .057 -11 Feliz DOM 132 -.164 -.030 -4 Ortiz DOM 143 .226 .296 -2 Polanco DOM 116 -.224 .056 -8 Pujols DOM 140 .741 .467 +10 Reyes DOM 150 .032 -.068 +4 Soriano DOM 143 .193 -.001 +7 Tejada DOM 142 .443 .133 +11 Alou DOM 94 .610 .147 +11 Encarnacion DOM 129 -.232 .003 -8 Guerrero DOM 135 .211 .267 -2 Pena DOM 78 .207 .194 +0 Taveras DOM 141 -.102 -.133 +1 Piazza ITA 89 -.059 -.068 +0 Catalanotto ITA 91 .513 .032 +11 Delucchi ITA 31 -.461 .134 -5 Blanco VEN 28 -.662 -.180 -3 Hernandez VEN 118 .145 .032 +3 Martinez VEN 138 .451 .053 +14 Alfonzo VEN 45 -.863 -.102 -9 Cabrera VEN 127 .485 .260 +7 Guillen, C VEN 126 .232 .067 +5 Perez, T VEN 89 -.387 -.205 -4 Scutaro VEN 45 -.266 -.086 -2 Vizquel VEN 129 .211 -.104 +10 Abreu VEN 136 .137 .184 -2 Chavez, En VEN 62 .001 -.141 +2 Ordonez VEN 127 .287 .064 +7 Rivera, J VEN 42 -.223 .010 -2
As I’m sure you’ve detected, the overall trend is neutral. A small majority of position players–42 of 77–have underperformed PECOTA. On the other hand, a few of those who have outperformed PECOTA (Alexis Rios, Carlos Lee) have done so by a large margin, and the overall plus/minus rating is in the black at +44. The hitters are fine, as were the relief pitchers–this problem seems to be confined to the rotation guys.
I think it’s worth remembering what happened in 1995, once the players finally got back on the field after the lockout ended. Through games of May 26th of that season–roughly the first month of that late-starting baseball year–the league scored an average of 10.1 runs per game. After May 26th–in spite of warmer weather–the league average dropped to 9.6 runs per game. There is some precedent to the notion that any kind of awkward start to the season tends to help the bats.
What to expect going forward? My hunch is that the pitchers will regress back to their PECOTA projections–or at least back toward the mean–over the course of the next month or so. Carlos Zambrano looked much more like himself in his last outing; Johan Santana looked like he was going to no-hit the Tigers. April may have served as a sort of trial-by-fire, extended spring training, but these pitchers have now gotten in four or five weeks of work. Still, the poor performances of WBC starting pitchers has already taken several wins from the records of teams like the Orioles and Twins, and that is probably too much damage to be undone, particularly when each of those squads has a whole host of other issues. Let me put it like this: I’m looking forward to the 2009 Classic, but if Justin Verlander stubs his toe and has to pull out of the event, I wouldn’t mind it one bit.
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.
Subscribe now