|
In general, baseball statistical rankings have not factored in the postseason. If you look at the WAR leaderboards on Baseball-reference, Fangraphs, or Baseball Prospectus, they focus on the regular season. Bill James, in Win Shares, covered the regular season. Same with Pete Palmer and his Total Player Rating. My stats are no different, from my original ratings to WAR 2.0 I have focused on regular season. That is, until today. WAR 2.0 pages are now available for all postseason games from 1912 on. Postseason WAR pages are on the same index as the regular season player links, side by side. There are good reasons not to do this. Sample size is a big issue. Opportunities are not even. In the regular season everyone could get a chance to play 162 games. Not everyone does. I played zero games last year, because I am not good enough at baseball to earn the playing time. Mike Trout is very good at baseball, better than almost everyone. He played only 29 games last year, because he couldn't stay healthy. If a player was both good enough and healthy enough to play 162 games, then the opportunity is there in the regular season regardless of whether he plays for a pennant winner or last place team. This is not true in the playoffs. From 1957 to 1960, Ernie Banks averaged 155 games per year. This is very impressive for a schedule of 154. The record must include a few tie games. He hit .293 with an OPS+ of 153, an rBat+ of 155, and averaged 44 homers with 123 RBI per year. He played shortstop. The fielding metrics liked him, and the gold glove voters agreed, at least in 1960, awarding him the gold glove. It's hard to find many players with a higher peak, for both production and durability, than Ernie Banks. And for all that, he never played a single playoff game. Beyond that, the playoffs have expanded over time. Up through 1968 only the two best teams played a single series of between 4 and 8 games. Then we added a round and had 4 playoff teams. Then 8, 10, and now 12. By any counting stat, the modern players have an advantage over the old-timers in playoff opportunities. Before 1968 the only players with impressive playoff career totals were Yankees. They only got one round, but at least they played that one round every year. Derek Jeter is the leader in postseason games played, with 158. This is partially attributed to his durability. The Yankees played 161 playoff games from 1996 to the end of his career. Derek played in the first 158 of them. He was injured in the second round of the 2012 playoffs and missed the last 3 games. His postseason career record looks like another full season from him. He is also the beneficiary of a great run of opportunity. Compare his postseason schedule to that of Mickey Mantle (65 games), Joe DiMaggio (51), or someone who did not have the good fortune of playing for the Yankees like Willie Mays (33 games). While acknowledging the opportunity issue, I'd rather have the record than not have it. I was curious on how players would rate in fielding and baserunning metrics. If you are using metrics for something like deciding who goes into the Hall of Fame, I don't know the best way to do that with postseason. Add it all in? Ignore it because of the opportunity issue? Count it double or triple because postseason games are more important? Do as thou wilt, it's all up to you. For the most part, I made as little changes to the code as possible. In some cases I round the numbers and lose some decimal points. That makes sense for regular season, but for something like the position adjustments most people are going to come out as zero. Maybe it would be better to give it an extra decimal place. If I did that I'd get plenty of complaints from people thinking that should not be done. Batting and fielding runs are centered at zero for each year. I definitely can't do leagues separately. If I did then I'd have Phil Rizzuto showing up as average every year because his fielding stats were the same as the AL playoff league average every year. In pre-1969 years I've got Rizzuto compared to the average of Rizzuto and Pee Wee Reese, or whoever the Yankees played that year. In later years things work out better because Trea Turner is being compared to the average of 12 playoff shortstops. One possible solution could be to use multiyear averages for the league level, but doing that would mean a lot of work on my part to redo the code. And that is not happening. One place I made a change to the code is for pitchers, when I get the expected runs allowed based on opponents faced. For this I use regular season ratings. One team that shows why this is necessary is the 1966 Orioles. They swept the Dodgers, pitched 3 shutouts, and allowed only 2 runs in 4 games. I use the Dodger's regular season level of 3.7 runs per game to establish the expectation. When Jim Palmer threw a complete game shutout, he gets credit for 3.7 runs saved above average. Much better than calling it only 0.5 runs. Doing that will distort the runs across the whole team, but it all washes out in the reconciliation, and once that is done player postseason wins will add up to the team total. The other change I made is to replacement level. For regular season games, I use the .294 replacement level, the same as Baseball-reference and other sites. In the playoffs you are facing a higher level of competition, so I use a replacement level of .220. I'm definitely leaving the door open for someone else to put in the work and come up with better playoff ratings. Just want to get something out there.
This page was last modified 12/19/2024 |
