Time to do some research

If anyone knows where someone has already performed this analysis, please let me know.

But I believe one of the most asenine strategies employed by NFL head coches is going for two after a touchdown with a four point lead late in the fourth quarter, as Herm Edwards just did against the Pack.

Intuitively, this makes no sense. Six-point lead versus five-point lead doesn't matter much when you expect the other team to only have one more solid possession. I understand the logic for it--two field goals by the other team will only tie the game, not win the game. But when you consider the success rate of the two-point conversion, it seems that the team is only setting iteslf up for a situation where the other team can go score a touchdown and put them at a three-point deficit rather than having to convert their own two-point conversion to make a three-point deficit.

I wonder, at what minute marker is the value of a six-point point lead over five points no longer worth the risk? As someone who believes the two-point conversion is a strategy that should be used only very late in games, I can't see there being a very large time window where six versus five is worth the risk, if there is one at all.

But seriously, if anyone knows whether there's hard research out there on this and other two-point conversion-related strategy, please let me know. Otherwise, I know what I'm doing on my lunch break everyday next week.