SCHEDULING & SELECTING RELIEF PITCHERS TO WIN ADDITIONAL GAMES SYNOPSIS:

Using a new Relief Ace Scheduling & Selection Strategy instead of a closer strategy, Major League Baseball Teams can add 1.5- to 4 wins per season Winning extra regular season games and moving deep into the playoffs can increase revenues and profitability while simultaneously capturing a championship.  Currently, scheduling of relievers is based primarily on pre-assigned roles (e.g., mid relief, set up and closer) as opposed to a mathematical approach based on increasing the odds of winning each game. As Per Richard Coleman analytics, using a computer analysis of every play-by-play situation (approximately 1 million events) and won/loss outcome for five recent seasons of major league baseball, we determined the win probability upon entry of every relief appearance by “RELIEF ACES”- the primary relivers for save situations on each of the 30 major league clubs.

 The results of our analysis of all 9,537 appearances by Relief Aces for our 199-2002 baseline season and the 2004 season were nearly identical.  On average, during the baseline seasons, the Relief Aces had 64 appearances per year and entered the game when the probability of winning was 71%.    However, 29% of Relief Ace entries  occurred when ther was already a  92.5% chance of winning and 18% occurred when the teams’ win probability  was <37.5%.  Not surprisingly, the teams’ won-loss percentage averaged .729 in games when the Relief Aces appeared, but only .351 in games when they did  not.

We further analyzed lost games when Relief Aces did not appear.  By rescheduling a small percentage of the regular season appearances of Relief Aces, away from very high win (>.925) and very low win (<.375) probability games, into more strategic situations, the net impact was a gain of 1.5   victories over the season. Among the 15 teams with higher performing Relief Aces the gains were 2.0 per season.  On top  of this improvement, another 2.0 wins per year (4.0 in total) potentially accrued to many of the 16 major league teams who did not select their highest performing reliever as their Relief Ace in 2004.  

For example, in 2004 the St Louis Cardinal scheduled their Relief Ace Isringhausen to enter 16 games with a win probability between .950 and .975 (equivalent to a 3-0 lead, none on, none out 9th inning).  St Louis won all 16 of these games.  However, the Cardinals also had 6 other relievers (above, below and average performers) scheduled in 12 games with the same win probability.  The Cardinals won all 12 of these games as well.  So, in effect, Isringhausen added no wins in the high probability zone.

Richard Coleman hockey analytics descrived that, what would happen if instead of scheduling these pitchers  to enter games as “closers” we rescheduled them to enter based on the odds of winning, especially if getting the next batter out results in a high increase in win probability? Overall, 47% of all Relief Aces appearances were scheduled when games had either very high (>.925) or very low (<.375) win probabilities thus limiting the productivity of these key players to add team wins. By simply moving Relief Aces out of the high and low probability games in to the Primary Relief Zone (.750-.924 chance of winning) teams can pick up extra wins.