Stephen Tomlinson (can I say the great Stephen Tomlinson?) has released the 2003 edition of his long-running series of "Stat Reports" on his website. Stephen's essays, which cover all of the AL and NL in addition to in-depth analysis of the Jays, are available for all years going back to 1977 and are a great capsule summary of each season.
Josh Phelps played in 119 games? Sure didn't feel like it.
Craig, you are very generous. Thanks.
P.S. There are some minor inconsistencies in the format due to what data was readily available at the time and how developed my scripts were (the first article went up in January 1998, before more general sites like baseball-reference.com emerged). For example, pre-1997 articles are missing the NL summary.
I hope to do an updated version some day. Retrosheet offers a lot of new possibilities, particularly for 1977-92. I'd like to do better defensive stats (more than just games played by position, e.g. something like James' defensive winning percentage). And especially I'd like to do win probability added (for all players, not just relievers, including defense and offense, i.e. like win shares but better). WPA might not have as much predictive value as traditional averaged stats, but it is better at explaining where the actual wins and losses came from. (Of course, it'd be fine if someone else would do this first, as I likely won't get to it for a few years.)
P.S. There are some minor inconsistencies in the format due to what data was readily available at the time and how developed my scripts were (the first article went up in January 1998, before more general sites like baseball-reference.com emerged). For example, pre-1997 articles are missing the NL summary.
I hope to do an updated version some day. Retrosheet offers a lot of new possibilities, particularly for 1977-92. I'd like to do better defensive stats (more than just games played by position, e.g. something like James' defensive winning percentage). And especially I'd like to do win probability added (for all players, not just relievers, including defense and offense, i.e. like win shares but better). WPA might not have as much predictive value as traditional averaged stats, but it is better at explaining where the actual wins and losses came from. (Of course, it'd be fine if someone else would do this first, as I likely won't get to it for a few years.)