i like eflhoca's theory for HD at first glance, it makes sense. metsmax's rebuttal makes sense in that by benching a player in foul trouble you're basically hedging your bets (sacrificing your team a bit 'now' to make sure you don't get completely screwed 'later' by decreasing the chance of double-foul-out).
I have a different take, though. ever notice how basketball "blowouts" are typically measured in point differential, as opposed to ratio between scores? a blowout in a basketball game might be 60-40, a huge blowout 80-40. for total points scored, that's about average for basketball, or at least in the ballpark. the average football game might be 27-17. nobody is calling that one a blowout, even though the scores are a 1.58/1 ration (compared to the 1.5/1 ratio in the 60-40 mini-blowout). a huge blowout could be 35-7, a 5/1 ratio, whereas the huge basketball blowout is only 2/1. hockey, baseball, soccer, etc. could all end up with these huge football-like scoring ratios and not be considered as huge of blowouts as an 80-40 basketball game.
the reason for this obviously comes down to how the game is scored and the consistency in which 'quality play' leads to quantifiable 'points'. in basketball, you outplay a team for a half and it will absolutely show up on the scoreboard; in football, some bad luck might leave the game within a score despite outplaying your opponent.
because of basketball's continuous scoring, a team that starts out the game on a 30-10 run (and completely earned the 3/1 scoring ratio) will NEVER maintain that scoring 'ratio', even if they had the talent advantage to do so. why play hard when you're up 20? you may have the skill to maintain that 30-10 'ratio', but you're much more likely to ease up and tie the other team the rest of the way, winning 70-50. basketball has a tendency to 'stay close' more than any other game BECAUSE the score is so fluid.
my point? what you do on the first possession is actually not worth as much as the last possession, because basketball games have a way of staying close. the benefit of having your best man out there in the first half could just as easily be mitigated early in the second half because of how big leads cause teams to ease up. the scores of HD games would lead me to believe they MAY set the game up like this beyond the 'play backups in blowout' option.
if they don't set the game up like that, then my whole theory still holds true with RL bball, at least.