Monday, September 26, 2011

Has Mariano Rivera statistically separated himself from his peers?

That is the issue raised by an article in yesterday's NY Times Mr. Young, Mr. Ryan, Mr. Rivera Will Be Joining You By DAVE ANDERSON.

He compares Rivera to Cy Young and Nolan Ryan. Young won 511 games and the next highest is Walter Johnson at 417. Ryan struck out 5,714 batters, 839 more than Randy Johnson. Does Rivera dominate his peers in a similar way?

It is hard to argue against his 0.71 ERA in 139 post-season IP. I doubt any other closer can come close to that. Probably no other has had a chance to pitch so much in the playoffs. But the article offered no comparisons. No other pitcher's post-season performance was discussed.

Rivera now has 603 saves and the only other guy with 500+ saves is Trevor Hoffman with 601. So let's see how those two compare in their regular season stats (Hoffman only had 13 post-season IP). The table below summarizes some of their stats.

Their save percentages are almost identical. I thought that maybe Rivera got into tougher situations because he has such low HR% and his ERA is so much better than Hoffman's (2.22 vs. 2.87) that it seemed like maybe he came in when the score was closer. But they each have about the same average leverage index. The 1.92 for Hoffman, for example, indicates that his typical game had nearly twice the pressure as average in terms of inning and score.

So these two guys pitched a similar number of innings under the same pressure and were equally successful in what they were asked to do. Now maybe Rivera was in more pennant races or playoff chases but the Yankees usually made it in fairly easily. Maybe the competition was a little tougher in the AL.

ERA+ is adjusted for league average and park effects. Rivera's means that his ERA so adjusted is a little less than half the league average while Hoffman is about 30% lower. So big edge for Rivera (another closer, Billy Wagner, in about 900 IP had a 187, not too far from Rivera). Wagner had a career save % of 86%.

They are close in WHIP. FIP or fielding independent ERA takes walks, HBP, SO, and HR into account (with IBBs taken out). Hoffman is not too far off. This comes from Fangraphs and I don't think it is park adjusted. A rough estimate is that Hoffman's parks were about 7.2% better than average for pitchers while for Rivera just about average. If I raised Hoffman's FIP ERA by half of that 7.2% I get 3.19.

When I do walk% and SO/BB ratio, IBBs are taken out and HBP are in. IBBs are also taken out when doing HR%. This edge for Rivera seems very large. Maybe it is even bigger being a righty in Yankee Stadium and other teams would have tended to send lefties up to face him.

Rivera is not too far ahead of Hoffman in AVG and OBP allowed. But the edge in SLG seems big, probably due to the low HR%. The edge is even bigger in road games.

One other closer, Dennis Eckersley is worth looking at. I found his best 6 seasons in FIP ERA and the simple average of them was 2.13. For Rivera it was 2.15. If I park adjust Eckersley here, he goes up to 2.19. Still not too far from Rivera. But that has to be adjusted for the league average. Those years for Eck around 1987-93. It looks like the league average has been about .44 higher in those years than it was over Rivera's years. Rivera's FIP ERA is about .474 of the league average while for Eck it is .535. If Eck pitched in those years of Rivera he would get 2.42. Not too far off (also recall that for Rivera, those are his best six years and but I just used his entire career as an approximation). Eckersely came to the closer well after the age of 30 (and maybe before that we did not really have closers). Eckersley had a career save % of 85%.

What is probably the amazing thing is for how long Rivera has been consistently great. He is almost 42 and is having another outstanding season.

Update: David Pinto points out that Rivera had 116 "long saves" (4 outs or more) while Hoffman only had 55. I left a response.

No comments: