I'm no statistician, but I'm curious how the statistical playing field is evened out between baseball eras. There are plenty of teams with 1907 Addie Joss rostered and for good reason. He allowed just .08 HRs per 9 innings pitched. But it was a season in which the median number of home runs hit by an American League team was 12 , or .023 HR per 100 at bats. Compare those stats to the 1975 Tom Seaver, who allowed .35 HR per 9 IP in a league where the median number of HRs was 101, or 1.81 HR per 100 at-bats. According to my calculations, it was 78 times harder to hit a home run in 1907 than 1975. Adjusting Seaver's stat to 1907 expectations would give him a HR/9 IP rate of .004. So why is it that Joss' worst sim HR per 9 rate is .22, compared to Seaver's best of .42?
Can anyone explain? Thanks in advance.

I made an error in computing the 1907 HR stats. The HR per 100 AB rate should be .23, so Seaver's adjusted HR/9IP rate would be .04. Sorry about the mistake, but my question still stands.

Congrats, you've figured out the concept of normalization.

Look at the "+" stats for normalized numbers. The normalized numbers show how well the pitcher did compared to the league average. 100 = league average. Higher is always "better".

That said... normalization for HR's is not done nearly as well as it should be. For some reason, the raw number seems to matter a lot more then the normalized number does when it comes to HR's by pitchers.Now, by the same token, I believe you will find Joss giving up (a lot) more 2B and 3B...

Just use the # stats to compare across eras.