More XR

The Top Ten XR/27 Scores from People with Between 100 and 502 PAs (Non-qualifiers who played at least a bit).

1. Frank Thomas - 8.98
2. J.T. Snow - 8.33
3. Joe Mauer - 7.78
4. Larry Walker - 7.55
5. Todd Hollandsworth - 7.33
6. Robb Quinlan - 7.30
7. Mark Sweeney - 7.04
8. Eli Marrero - 6.81
9. Trot Nixon - 6.67
10. Jason Bay - 6.59

The Top Ten XR Scores from Qualifiers

1. Barry Bonds - 18.86
2. Todd Helton - 10.46
3. Jim Edmonds - 9.66
4. Albert Pujols - 9.31
5. Lance Berkman - 9.25
6. J.D. Drew - 9.23
7. Bobby Abreu - 9.02
8. Scott Rolen - 8.79
9. Travis Hafner - 8.35
10. Melvin Mora - 8.29

The highest ranking Met:

75. Mike Cameron - 5.48

When you look at it that way, I don't know if Art Howe should have been fired.....

The "We Don't Know Why We're in the Majors but Jim Bowden Could Still Give Us $16.8 Million" Bottom Ten XR with 502+ PAs

1. Alex Cintron - 3.29
2. Alex Gonzalez - 3.42
3. A.J. Pierzynski - 3.55
4. Reed Johnson - 3.69
5. Cristian Guzman - 3.72
6. Angel Berroa - 3.75
7. Craig Counsell - 3.76
8. Eric Hinske - 3.81
9. Tike Redman - 3.88
10. David Eckstein - 3.94

And, the honorable mention on the Bottom-10, who was so bad that he was #12, even at Coors' Field....

12. Aaron Miles - 4.02.

There you have it. XR's not park-adjusted, so you can see that there are a lot of Rockies at the tops of lists and such. It could be park adjusted, but I didn't have the figures and I wanted to post this stuff.

XR/27 says how many runs on average a team would score with 9 of ________ in the lineup. The best lineup in baseball would be 9 Barry Bondses. Not only would they score almost 19 runs / game, the games would last forever, considering how many walks he gets.

This is ACTUAL XR, though, not XRR, b/c XRR would have Bonds rated quite a bit higher. IBBs count less than BBs in the XR formula.

The one other thing? Barry Bonds added more than 8 wins (via XR) to the Giants compared to a slightly below-average LF (Jose Cruz Jr, the closest match to a replacement type player I could find). Not even in the running without him. Probably one that doesn't need the stats to verify it, but it's good to see the numbers.

Methodology: I used Baseball Prospectus's league average totals for LF, which were .277/.366/.484. Replacement level is about 35 less on each number, as a fair estimate. So it's .242/.331/.449. I then looked through the database of 2004 and tried to find a player (at any position) who was most comparable to those numbers, and substitute him into the Giants' team total for Barry Bonds. Fun stuff. The best match? Jose Cruz Jr, who, if he were in left field, would be about at replacement level, maybe a bit better b/c of his steals. That said, I thought that it was a good comparison.

Problems: This study has a few problems....I normalized Cruz's plate appearances to Bonds's, but the problem is that Bonds's offense was a major reason why there were so many extra plate appearances for the Giants. This works on a few assumptions: equal park factors (which is not the case), correct replacement-level numbers, the difference in steals, stuff like that. GIDP, sac bunts, sac flies, are all random. Cruz was just the closest match I could find. If you'd like to REALLY look at wins added, check out eXtrapolated Wins, which uses XR and replacement level players to actually calculate the difference. I will work on this myself (if I have time) after I finish writing my biology paper.

The XR rankings are correct, though, assuming that the Baseball Database from Baseball1.com is accurate (I'm confident that it is) and that my spreadsheet doesn't have any errors in it (I'm pretty sure of that, too).

And, in perhaps the biggest surprise of this, Edmonds probably deserved to be in the MVP running a lot more than people thought he did. Edmonds didn't get a top-4 vote, and he ranked 3rd in XR. Adrian Beltre was 11th, btw.

I'm not going to be tearing through stats like this come the regular season, but it's what I've been working on, so I figured that I should post it.