[Comment From AriAri: ]
Am I the only one that has a weird feeling that Edgar Renteria is going to become the first player to end two World Series with a walk-off?
The game was in Texas so this couldn’t have happened tonight but it was eerily close
fred
15 years ago
Gotta love the Dave C snarkiness telling a commenter to learn about the win/run expectancy model.
As Dan pointed out these are GENERIC models which assume league average hitter, league average pitcher, don’t care about handedness, the number of times through a lineup, the opposing pitcher or potential day to day issues most of us have no visibility into (is a pitcher or hitter dealing with a nagging issue? etc…)
Just because the win expectancy didn’t change in a given situation in a league average case doesn’t mean it won’t change in a specific case (especially when there are platoon, hitter/pitcher quality and other variables at play)
In the zeal to prove how non-intelligent many managerial moves are through the use of models, can folks at least begin to do this without applying those models blindly? Maybe at least acknowledge their limitations or attempt to factor in situation and the actual players involved?
The Lester thing was really distasteful.
[Comment From AriAri: ]
Am I the only one that has a weird feeling that Edgar Renteria is going to become the first player to end two World Series with a walk-off?
The game was in Texas so this couldn’t have happened tonight but it was eerily close
Gotta love the Dave C snarkiness telling a commenter to learn about the win/run expectancy model.
As Dan pointed out these are GENERIC models which assume league average hitter, league average pitcher, don’t care about handedness, the number of times through a lineup, the opposing pitcher or potential day to day issues most of us have no visibility into (is a pitcher or hitter dealing with a nagging issue? etc…)
Just because the win expectancy didn’t change in a given situation in a league average case doesn’t mean it won’t change in a specific case (especially when there are platoon, hitter/pitcher quality and other variables at play)
In the zeal to prove how non-intelligent many managerial moves are through the use of models, can folks at least begin to do this without applying those models blindly? Maybe at least acknowledge their limitations or attempt to factor in situation and the actual players involved?
Agreed. Expected value models kind of suck when applied to specific contexts. Even a systems dynamics model would be a major improvement.