Saturday, March 27, 2010

Similarity Scores Lineup: Boston Red Sox

Baseball-reference.com has a great feature called similarity scores, where it finds 'similar players' in baseball history to a chosen player. So let's propose a 'hypothetical' similarity lineup with the Sox player and a similarity 'shadow.

C Victor Martinez   Joe Mauer (really)
1B Kevin Youkilis  Adrian Gonzalez (yeah)
2B Dustin Pedroia  Aaron Hill
SS Marco Scutaro  Pokey Reese
3B Adrian Beltre  Mike Lowell (meet the new boss, same as the old boss)
LF Jacoby Ellsbury  Denard Span (and the Twins wanted Ells)
CF Mike Cameron  Tom Brunansky (end of the season catch fame)
RF J.D. Drew Carl Everett (not in temperament)
DH David Ortiz  Mo Vaughn (who woulda thought)
P  Josh Beckett  John Lackey (number 1 on sim score!)
P Jon Lester  Daisuke Matsuzaka (doesn't look Japanese)
P Jon Lackey  A.J. Burnett (is what it is)
P Jonathan Papelbon  Takashi Saito (why can't we get players like that)

The point? Sometimes our "endowment bias" sees players differently than they 'are'. For example, most everyone would agree that Albert Pujols is one of the top players in the game. Who is on his similarity score list?
This is a little bit of apples and oranges, because above is the list by age. The higher the number, the more similar the player. For example, Reggie Smith had Fred Lynn (959) as his number one sim, one of the closest I've ever seen. Roger Clemens number one is Randy Johnson at 850. Luis Tiant has Catfish Hunter (an easy Hall of Famer) at 942 and HoFer Jim Bunning at 931, although Tiant hasn't gotten within a sniff of the HOF.

No comments: