How Many Games Will the Mets Win?

Now that the Mets have acquired Johan Santana (pending a contract agreement of course) most baseball analysts see the Mets as the NL East favorites, and many as the favorite to represent the NL in the World Series. This seems to be a fair expectation for the Mets who were the best team in the NL for 5/6 of the season last year, but it lacks statistical support. So then how many wins should the Mets, statistically, win?

Win Shares: Win shares is the Bill James statistic that measure how much each individual player contributes to a team win. Each win share is equal to 1/3 of a win according to James. The players below are who I expect to be on the Mets 25 man roster for the majority of the year.

Players Win Shares (2007)

Line Up
Jose Reyes -----------------24
Luis Castillo----------------16
David Wright---------------34
Carlos Beltran--------------25
Moises Alou----------------12
Carlos Delgado-------------13
Ryan Church---------------16
Brian Schneider -------------11

Bench
Ramon Castro---------------6
Endy Chavez (2006)---------13
Angel Pagan----------------5
Marlon Anderson------------4
Damion Easley--------------8

Starters
Johan Santana--------------17
Pedro Martinez (2005)-------19
John Maine----------------11
Olliver Perez---------------10
Orlando Hernandez---------9

Bullpen
Billy Wagner---------------12
Aaron Heilman--------------8
Duaner Sanchez (2006)-------6
Pedro Feliciano--------------6
Matt Wise------------------4
Scott Schowenweis-----------2
Jorge Sosa ------------------6

Total Wins-----------------99

I added up the total amount of win shares that each player on the current Mets' 25 man roster had in their last healthy year, which came out to 297 win shares, and divided that by three to estimate the total number of wins, 99. Obviously certain players will do better next year, and certain players will do worse, but it is probably safe to assume that the progression of players who are either still improving or who are due for a rebound year will offset the regression of certain injury prone or ageing players.

Pythagorean Record: Is used to calculate how many wins a team should have.

Formula: (Runs scored^1.83)/ ((Runs scored^1.83)+(Runs allowed^1.83))

Runs scored by current Mets' 25 man roster: 790
Runs allowed by current Mets' 25 man roster: 559

Pythagoren Winning Percentage: .653
Pythagorean Record: 106- 56

Once again only using players on the projected 25 man roster, and only using their last healthy year I found the Mets to win 106 games. Now the pythagorean record is only a projector of how many games a team should win and, unlike win shares, does not take the actual win total into account.

Both of the projected win totals will be skewed because it is impossible to project injuries, and the performance of the subsequent replacement player. You do not know who is going to step up and become a major contributor, nor do you know who will have a major drop- off year. Of the two methods the win share total is probably the more accurate projector, but the pythagorean record does give a good estimate of the team's potential. For comparison's sake the Philadelphia Phillies' current 25 man roster would only win 86 games by the win share method. This is probably a valid estimate considering the Phillies won 89 games last year and lost a major contributor in Aaron Rowand without really doing anything to make up for his departure. However they have a pythagorean record of 101-61, which means that a full season of Brett Myers and Kyle Kendrick with ERAs in the mid 3s and a full season of Cole Hamels pitching Cy Young caliber baseball would give them the potential to win 101 games, which is still 5 games worse than the Mets potential.

Overall 99 wins seems like a pretty logical projection for next year, but 106 is certainly not out of the question for a team as talented as the Mets. Either way all indicators point to a very successful season for the New York Mets.