averagejoe99 wrote:

... this still sounds like a case of "Limit Bench Playing Time" being used and the CM having no pinch-hitter available who isn't ahead of their real-life plate appearance "pace" vs. whatever side the opposition pitcher hurls from. The CM would like to PH for the pitcher but it can't and so, maddeningly, it doesn't... only to then turn around and replace the pitcher in the next half inning. The CM can sometimes work around this with some (earlier) double switches but, otherwise, it's stuck.

If you play with that switch turned on (some leagues do), make sure you have at least a few guys with available PA's vs. both sides on your bench. Or, as I do, go in and change individual player limits to 999 (or at least some number higher than real life) vs. either/both side. (Just make sure you farm or replace such a guy before he gets you into overuse trouble in your league.)


I understand how LBPT works and in the 35 seasons our league has operated it has never been turned on.  So that couldn't be the explanation. 

I want to make it clear that I wasn't really complaining, just making an observation.  Computer simulations have a limitation compared to human decisionmaking because (so far at least) humans are better at considering and balancing multiple objectives, especially when those objectives are contradictory.  Computer programs will handle those situations with a more linear approach to sequential decision rather than examining all the potential objectives and decisions in one bundle of variables.  In the situation described by Wreckilarian, the CM's process for deciding if the pitcher should bat probably determined that with nobody on base, a pinch hitter would have less impact on the team's odds of winning compare to the impact of leaving the pitcher in.  But when the situation changes from 0 on and 0 out to 1 on and 0 out, the run expectancy almost doubles, so now perhaps the CM sees a bigger advantage in improving the chances of scoring one more run than it does in keeping the starter in to pitch the 9th.  Or even if it doesn't, it might have another routine for determining when to pinch run for someone to reduce the chances of injury.  A human would have already decided whether or not he cared more about leaving the pitcher in than about increasing the odds of another run when he made the pinch hit/no pinch hit decision, and probably wouldn't change that assessment just because the pitcher gone on base.  So I wasn't trying to say that there is something amiss in DMB because of such (infrequent) perplexing decisions, other than to observe that it is something that will always occur with computer simulations of complex decisions.  DMB seems to have a lot fewer of those kinds of little puzzlers than other games, at least the ones I've played.

On the other side of the ledger, there are some mistakes that humans frequently make that the CM probably probably won't make. In the example I described above, a human who decides not to pinch hit for a pitcher because the base-out situation is unpromising might then refuse to consider using a pinch runner if he gets on even if one would be called for, simply because to do so might imply that his decision to let the pitcher bat was a mistake.  Humans can get locked into irrational decisionmaking too, just for different reasons.