This post will focus on the career of Christy Mathewson to further explain what WAA represents when judging value of baseball players. Christy Mathewson pitched from 1900 – 1916 for the New York Giants (SFN) and had one of the top careers of all time; ranked #36 with a 92.5 career WAA.
There is an interesting blip in the year 1906 of his career. Christy pitched a 22-8 season yet his WAA=-1.7 is below average meaning he hurt his team by 1.7 games. Stats like WAA provide a lens through which a collection of data can be seen. At first glance one would look at a pitcher’s W/L record to make a judgment as to that player’s worth. In this case the win loss record deceives.
Oddities like this make me check the code and math generating these tables to see if a bug exists. This post will explain the result of my analysis and hopefully provide insight to the philosophy behind WAA and other player rating methods to be introduced later.
First a shout out to baseball-reference.com which I use a lot for reference. Baseball-reference has far more information about players, teams, history, and everything baseball. This site cannot possibly reproduce their fine work. Tables presented here are of one format, the reasoning for which will be explained in a future post. Baseball-reference has a stat called WAR which resembles WAA but calculated very differently and will produce very different rankings at times. Christy Mathewson’s full career stats can be seen here on their site. His value amongst other 1906 players according to WAR can be seen here (click on WAR to see rankings). Even baseball-reference ranks him #19 amongst pitchers with a WAR=2.5 so they see a problem with this 20+ game winning pitcher too.
Let’s take a look at Mathewson’s career trends by listing his stats 2 years before and after 1906.
Mathewson came off a 31 win season in 1905 posting a historical high WAA=12.8 and then dipped negative in 1906 after which he climbs to a top tier player again in the next two years. Notice the correlation between the ERA column and WAA. In 1906 Mathewson posted an ERA more than double that of 1905 and 50% higher than 1904 and 1907.
Pitchers as a class are the only players responsible for team runs against column. ERA directly relates to the number of runs against a pitcher gives up.
ERA = RA/Games where Games = IP/9.
The above formula is the definition of ERA. The more runs a pitcher gives up directly relates to team losses. Let’s take a look at Mathewson’s ranking amongst the league to provide context with respect to his ERA.
Top 5 Pitchers:
Christy ranks 143/175 in the league according to WAA, much of it to do with how many runs he gave up. Maybe part of why he won so many games is through his above average batting.
Usually pitchers are negative in this area. Wins and losses are determined algorithmically by a rules committee to give prizes to pitchers for their participation in each game. Mathewson got lucky to win 22 games. He played for the Giants who finished with 96 wins, second behind the Cubs’ 116 so Mathewson had a lot of hitting behind him to compensate for how many runs he cost his team. It is extreme that a team with 96 wins finishes 20 games out of first in a 154 game season. It might not have mattered how Mathewson pitched that year with such disparity between team talent.
WAA and Mathewson’s league ranking was appropriate after analyzing this oddity.