Uberstats

I've been working on a uberstat database, calculating wins above replacement for everybody in the Retrosheet era. This includes park adjusted batting runs, which are based on custom linear weights, so the team baserun total adds up exactly to their actual runs scored. It was a bit trickier than it sounds, as I needed to remove pitcher hitting, or else I'm going to rate the AL players too low. This should be similar to the results of batting runs on baseball-reference, but sometimes there are decent sized differences. If a player consistently played on teams that scored more runs that you'd expect given their batting stats, the player will get extra credit.

I also removed baserunning and GIDP runs - say a team is 30 runs above average on the bases, and scores a total of 750 runs. I don't want to double count the baserunning, so I figure the LW values of the singles, walks, homers, etc. as if the team scored only 720. That way the batting runs + baserunning = actual runs.

Baserunning includes steals and caught stealing, as well as tagging up, going first to 3rd, etc. It tries to be a comprehensive evaluation. GIDP runs is based on how many DP's you hit into, given your number of DP opportunities.

For defense, there's TotalZone, outfield arms, and infield double plays. Plus catcher runs based on SB/CS, PB, WP, and errors.

All of these are converted to wins based on a custom runs per win figure for that league.

Position adjustment, per 150 games, is as follows (in wins): catcher +1.0, SS +.75, 2b/3b/cf +.25, RF/LF -7.5, 1B -10, and DH -15.

Finally, the difference between average and replacement level. This varies by my league strength calculations, between 1.8 and 2.2 wins per 150 games. For the 50's and 60's the NL was better. In the 70's and 80's it was about even, and in the 90's and beyond the AL has taken the lead. All of this is based on relative performance of players who played in both leagues.

The biggest surprise for me was Willie Davis topping the 70 win mark, which is a level of greatness. He played in an extreme pitchers park in one of the toughest decades to put up numbers, but in his context was an above average hitter, about 100 runs worth. He stole a lot of bases, ran the bases well according to retrosheet data, and rarely hit into double plays. His baserunning was worth another 100 runs, and he played outstanding defense in center field. He was about 350 runs above average, and in the 60's the runs to win conversion was pretty low. Add in replacement level for a National League that was superior to the AL at the time, and Willie gets 71.5 wins.

Pulling out my win shares book, I see Willie had 322. Just behind Ron Santo (325) and Reggie Smith (324) and just ahead of Nettles, Trammell, Simmons, and Torre. So perhaps Willie deserves more mention among the greatest players overlooked by the Hall of Fame.