I thought I'd share the results of some of my latest baseball nerding sessions. The rankings are after the jump.
Here's my methodology. I've constructed 25-man rosters for each of the teams. On the offensive side, each position player gets a % of expected plate appearances (e.g., 11% for a star near the top of the lineup, maybe 2-5% for a bench player), and then, if I think appropriate, a +/- of a percentage point or three for "platoon" players (e.g., Kubel will see more PA vs. RHP and Harris more against LHP). The idea is to try to tune the lineup for pitcher handedness like the manager likely would. Using PECOTA projections with added adjustments for each player's projected splits vs. RHP and LHP, I calculate a team OPS versus RHP and one versus LHP and 2 corresponding runs-scored figures against a neutral opponent.
Each team then gets assigned 5 starting pitchers and 7 relievers, and a corresponding team ERA for each of the 5 starters based on their expected average number of innings pitched, with the 7 relievers covering an equal share of the remaining innings. I then figure defense for the team (based just on the 8 starting position players, from CHONE) and figure team runs allowed.
The 3 best projected relievers on each team get special consideration as high-leverage pitchers (since teams with excellent closers and setup men will tend to win more games than their RS/RA figures indicate). This results in an adjustment of usually less than a percentage point in team win percentage (highest: Dodgers: +1.2%, lowest: Astros: -.067%).
The first set of rankings are power rankings pitting each team against the same neutral opponent in a neutral run environment. The second set, the divisional rankings, are actual simulations of each team's 162-game schedule, pitting each team and its 5 starting-pitcher squads against a randomly assigned series of starters from the actual teams they'll face this season. So the second set reflects each team's strength of schedule, as well as the handedness of pitchers they are likely to face and their corresponding OPS splits. Since starting pitchers are randomly assigned here, there is a little "play" in the rankings; each team's projected wins might adjust by +/- half a game or so based on how lucky they get with opposing starters.
I'm sure this isn't a perfect projection system, but I think it has its strengths, and I have a lot of fun with it. I'd love to hear where people think it's about right and where it's off wildly. Awesome.
|Rank||Team||RSvRHP||RSvLHP||RS||RS rank||RA||RA rank||Def||Def Rank||Win%||W||L|
|2||New York Yankees||5.61||5.38||5.54||1||4.26||6||-7||17||0.58||94.0||68.0|
|3||Tampa Bay Rays||5.29||5.52||5.36||3||4.26||5||15||7||0.56||91.2||70.8|
|4||Saint Louis Cardinals||5.19||5.36||5.24||5||4.19||2||32||2||0.55||89.6||72.4|
|6||Los Angeles Dodgers||4.91||5.20||4.99||14||4.33||7||-11||21||0.53||86.2||75.8|
|17||San Francisco Giants||4.67||5.01||4.77||24||4.40||13||-22||29||0.49||79.5||82.5|
|18||New York Mets||4.92||5.26||5.02||10||4.66||21||-5||15||0.49||78.9||83.1|
|20||Los Angeles Angels||4.96||5.14||5.02||12||4.71||23||-20||27||0.48||78.1||83.9|
|23||Chicago White Sox||4.59||4.87||4.67||29||4.53||17||-12||23||0.47||76.5||85.5|
|24||Kansas City Royals||4.77||4.74||4.76||25||4.65||20||-18||26||0.46||75.3||86.7|
|26||San Diego Padres||4.70||4.64||4.68||28||4.63||19||-8||18||0.46||74.2||87.8|
|29||Toronto Blue Jays||4.75||4.94||4.81||22||4.86||28||-24||30||0.44||71.8||90.2|
|AL East||Red Sox||95.9||66.1|