Talk:Estate/@comment-4111213-20110717202557/@comment-4148196-20110718025227

Not one needed repair yet? I'd been wondering about that. What if the repair mechanism works by increasing the likelihood of needing a repair the longer a structure has gone without it? That seems a logical way to implement repairs realistically, plus it explains why a new building never needs repairs. The ~7% value we see could simply be the average value as we sum up the increasing likelihood of a repair on a building (which resets when a repair is done).

The next step would be to keep data on a building's second collection, then its third, etc. which you are in a decent position to do.