First results from my tests (I need to further update my NHL data but I have enough data for a first look):
In the game I track how many games a player misses due to injuries (so I take only major injuries into account that make it impossible to play). The info I store is "how many linked games does a player miss", so in the end I know exactly how many player had e.g. an injury that made them miss 3 linked games, 5 linked games or 56 linked games. I also have the total number of injuries and I have them separated by goalies, defender and forwards.
You will not like the results but we are actually quit a bit below the real world data (especially goalies).
We are pretty spot on for different length types (1-5 games and 1-14 games compared to 6-82 and 15-82 is almost the same as in reality).
I will continue to update my NHL data and, if the trend holds true, update the injuries in game. But I will also look into making injury frequency optional (basically like OOTP with a Realistic setting and a default average value which is below that and 2-3 more steps). That way everyone should be happy.
