kmason, what dahsdebater is saying is if you look at a large enough sample size, all the other factors will average out. jdtapp or whatever his name was did a similar statistical analysis years ago and came to the same conclusion - that blk had a larger correlation to low opponent fg% than def, and it might have even been that it was a larger correlation than any other stat. however, i argue that unless recruit generation is totally random, this says little (and i doubt it is).
also, i won about 5-10 championships before i even started paying attention to shot blocking a little - i literally ignored it completely, like durability. so no amount of statistical analysis will ever convince me shot blocking is the most important defensive category.
my running theory is that admin was smart enough to know people would try to "stat hack" the game - i mean its certainly easily - and he made sure this would not 'break" the game. i already scrape the history of every team in HD, i could very very easily grab the stats and ratings of every team as well, and do a very large sample size stat analysis - so i feel its too easy not to safeguard against. plus i am not convinced that is not cheating, so i wont do it - i don't feel other coaches could do it and it has nothing to do with their coaching ability. and, i am pretty sure, the results would intentionally mislead at least in some areas. that is based primarily off the results jdtapp showed from his statistical analysis and how far from reality i perceived them to be. you can be damn sure if i made a sim engine, i would have run every one of these myself, and made sure some of them were misleading.
edit: just to clarify, i am not at all suggesting anybody who mined data to run these numbers did anything wrong. i am just not really sure what to think of it so i personally don't want to touch it.
1/19/2011 12:51 AM (edited)