- Who will become our new president?
- Will the 2008 Olympic medals from Beijing give our world's best athletes lead poisoning?
- And, most importantly, how will our 2008 BRASSball season shape up?
To answer this last question, I decided to run a little experiment using the concept of "Win Shares."
Introduced by Bill James in this 2002 book, Win Shares is a very complicated statistic that takes all the contributions a player makes toward his team’s wins and distills them into a single number. This single number represents the number of wins contributed directly by the player to his team, times three. In other words, a player with 3 win shares would contribute about 1 win to his team. Scores in the 20s are generally very good...scores in the 30s or above are MVP caliber.
In an effort to gauge the approximate strength of each BRASSball team, I took each team's roster (as of December 23, 2007) and assigned a Win Share value to each player. I then totaled each team's roster to determine a relative "strength" score for that team. Beginning with this post -- and continuing in additional posts to follow -- I'll be presenting my findings.
Before beginning, here are some final thoughts, assumptions, and admissions:
- I realize that many teams are still assembling their final roster, and that the upcoming draft will also change these scores. That said, these findings provide a valuable off-season baseline.
- I used Win Share figures from The Hardball Times website. The authors of this site have tweaked the original Bill James formula, so these Win Share scores may differ slightly from Win Share figures published elsewhere.
- Undoubtedly, it is probably not very "scientifically" sound of me to appropriate these MLB Win Share figures directly into our league. Given our league's smaller size (24 teams instead of 30) and different rules (DH for all teams, not just one league), I'm sure the "real" BRASSball scores would differ a bit. However, these MLB scores should still serve as a reasonable proxy for player value in our league.
- The Hardball Times breaks its Win Share scores into 3 areas: hitting, pitching, and fielding. This allows us to judge not just each team's overall strength, but also the caliber of their batters, pitchers, and (in theory) fielders separately.
- In real life, pitchers do receive "batting" Win Share scores. (For example, Arizona pitcher Micah Owings received 4 win shares with his batting alone. Meanwhile, his teammate Doug Davis received a -3 score.) Since our league does not require pitchers to bat, I zeroed out all hitting scores for pitchers.
- For the sake of completeness, I have included Fielding Win Shares in my analysis. I have done so even though Strat-O-Matic has its own unique fielding ratings. My reasoning is that: a) I've taken enough liberties with the Win Share concept already; b) that fielding is unarguably important in both MLB and Strat; and c) that any discrepancies between the Win Shares and Strat-O-Matic fielding systems would hopefully "even out" in the end. (Note: It does not appear that The Hardball Times assigned any fielding scores to pitchers, as all pitchers had "zero" scores for fielding.)
- Finally, in totaling each team's Win Share scores, it became apparent that these Win Share scores did not translate directly to one expected BRASSball win per 3 MLB Win Shares. For example, if I divided each BRASSball team's score by 3 to arrive at an "expected win" total, the average team would have won 90+ BRASSball games (which is impossible). I figure this discrepancy was due to: a) our league's smaller player pool changing the underlying formulae; and b) my elimination of batting Win Shares from pitcher scores. To compensate, I have assumed that one BRASSball win will require about 3.5 MLB Win Shares. This assumption delivers the desired BRASSball average of 81 wins per team.
Hopefully all of that (or at least most of that) made some sense. Also, note that this entire project is just for entertainment. I am as skeptical as anyone as to how this will all correlate to actual "on field" BRASSball performance!
Thanks for your patience, and I look forward to publishing my findings in the coming days. We'll start with the AL East and make our way through all six BRASSball divisions. We'll then look at the findings overall, and predict our upcoming season's strongest divisions and playoff contenders.
In an effort to gauge the approximate strength of each BRASSball team, I took each team's roster (as of December 23, 2007) and assigned a Win Share value to each player. I then totaled each team's roster to determine a relative "strength" score for that team. Beginning with this post -- and continuing in additional posts to follow -- I'll be presenting my findings.
Before beginning, here are some final thoughts, assumptions, and admissions:
- I realize that many teams are still assembling their final roster, and that the upcoming draft will also change these scores. That said, these findings provide a valuable off-season baseline.
- I used Win Share figures from The Hardball Times website. The authors of this site have tweaked the original Bill James formula, so these Win Share scores may differ slightly from Win Share figures published elsewhere.
- Undoubtedly, it is probably not very "scientifically" sound of me to appropriate these MLB Win Share figures directly into our league. Given our league's smaller size (24 teams instead of 30) and different rules (DH for all teams, not just one league), I'm sure the "real" BRASSball scores would differ a bit. However, these MLB scores should still serve as a reasonable proxy for player value in our league.
- The Hardball Times breaks its Win Share scores into 3 areas: hitting, pitching, and fielding. This allows us to judge not just each team's overall strength, but also the caliber of their batters, pitchers, and (in theory) fielders separately.
- In real life, pitchers do receive "batting" Win Share scores. (For example, Arizona pitcher Micah Owings received 4 win shares with his batting alone. Meanwhile, his teammate Doug Davis received a -3 score.) Since our league does not require pitchers to bat, I zeroed out all hitting scores for pitchers.
- For the sake of completeness, I have included Fielding Win Shares in my analysis. I have done so even though Strat-O-Matic has its own unique fielding ratings. My reasoning is that: a) I've taken enough liberties with the Win Share concept already; b) that fielding is unarguably important in both MLB and Strat; and c) that any discrepancies between the Win Shares and Strat-O-Matic fielding systems would hopefully "even out" in the end. (Note: It does not appear that The Hardball Times assigned any fielding scores to pitchers, as all pitchers had "zero" scores for fielding.)
- Finally, in totaling each team's Win Share scores, it became apparent that these Win Share scores did not translate directly to one expected BRASSball win per 3 MLB Win Shares. For example, if I divided each BRASSball team's score by 3 to arrive at an "expected win" total, the average team would have won 90+ BRASSball games (which is impossible). I figure this discrepancy was due to: a) our league's smaller player pool changing the underlying formulae; and b) my elimination of batting Win Shares from pitcher scores. To compensate, I have assumed that one BRASSball win will require about 3.5 MLB Win Shares. This assumption delivers the desired BRASSball average of 81 wins per team.
Hopefully all of that (or at least most of that) made some sense. Also, note that this entire project is just for entertainment. I am as skeptical as anyone as to how this will all correlate to actual "on field" BRASSball performance!
Thanks for your patience, and I look forward to publishing my findings in the coming days. We'll start with the AL East and make our way through all six BRASSball divisions. We'll then look at the findings overall, and predict our upcoming season's strongest divisions and playoff contenders.
No comments:
Post a Comment