References to teams will use a 3 letter team id for this data model which comes from retrosheet.org. These team ids should be intuitively obvious to the most casual observer. SLN means St. Louis National and DET means Detroit. What do these two teams have in common? The WAR system underrated them both at a team level for batting; ranking DET #11 and SLN #19 when those two teams were #2 and #3 with respect to runs scored in 2013. This post will examine WAR batting evaluations for these two teams.
Let’s dive down into player ratings to see if there’s a pattern. It is not possible to compare WAA and WAR rankings directly since the two numbers have much different ranges and calculated differently. Each system however should provide an accurate ranking as to how they see players’ value in MLB. The entire purpose of a single number to discern a myriad of stats is to determine who provided more value to their team.
The tables below show the most egregious ranking differences between the two systems for SLN and DET. The first column represents the rank difference between WAA and WAR. The second and third columns show the ranks for their respective systems. I took the top 50 ranked players for each system, compared them, and sorted by highest differences. The numbers represent how WAR and WAA ranked that player amongst batters only.
St. Louis Cardinals
Update: The ranking numbers below might or might not be totally correct. Although baseball-reference WAA seems to strongly correlate with their WAR that might not always be the case. The script assigning WAR to batter players was extracted baseball-reference WAA value instead. Team and pitcher players were correct.
Update 2/3/2013: Rankings in both tables below now reflect the corrected WAR evaluation. WAR rankings for all of St. Louis did increase but not significantly. That St. Louis was undervalued in batting by the WAR system still stands.
The table below shows WAR evaluation along with the WAA for reference; batting averages, WAA, and WAR highlighted. Clearly WAR did not see much value in any of these players. All three of the below players made it to the top ten St. Louis Cardinals list posted here awhile back.
Update: WAR numbers were corrected on 1/30/2014 in below player tables.
Note: Ranges and distribution of WAR and WAA will be presented in a future post.
Detroit only had a couple of major differences. Miguel Cabrera was added here for reference. WAA ranked him #1 while WAR ranked him #6 with respect to batting. That’s not a significant difference but an interesting data point. All these players made it to the top ten 2013 Detroit Tigers list posted here awhile ago.
Update 2/3/2013: The differences don’t seem so extreme with the updated rankings based upon the corrected WAR information. The differences exist but accrue incrementally over the evaluation for the entire Detroit team. That WAR undervalued Detroit as a team still stands.
The above data are extreme examples of underrated WAR batting evaluations that caused SLN and DET to rank so much lower in the league with respect to batting than their actual run production numbers would indicate.
Since WAR incorporates fielding into its number would that have any effect causing SLN and DET to be ranked so low? According to the table in the previous team ranking post, SLN had +9.1 UR and DET had +13.1. This means they gave up fewer unearned runs as a team than league average which should indicate their fielding is above average. We’ll introduce this stat in more detail in a future post.
The next post will cover LAN (Los Angeles National) and PIT who were rated #2 and #3 with respect to batting but had below league average run production.