http://ftw.usatoday.com/2017/04/roy...-unc-ncaa-tournament-success-tom-izzo-coach-k
^ Couple interesting links on who is better. I would offer some math that may clear this up a little bit. Now consider that Roy's wins have ALWAYS come as head coach of a team in a power conference, not mid level programs but blue blood programs where the competition is very strong EVERY year. In other words, no weak conference fluff to Roy's record like would be there if Roy coached Memphis, UMass, a SEC school, or Army.
The math:
Roy has a career win total of 816 having head coached 1032 total games, his winning % is 79%
K has won 1071 games having head coached 1401 games for a winning % of 76%
Now that is a very large sample size for both coaches so the % are not likely to change for either of these hall of fame coaches. K has coached 369 more games than Roy has so if we apply Roy's expected winning % to that extra number of games so that the games each have coached is equalized, it would add ( 369 games x Roy's career win % of 79%) would add another 291 wins to Roy's current total = 1323 wins for Roy to Ks current total of 1071.
Now it is not likely that Roy coaches another 369 + how ever many more K coaches to see the exact final records be equally considered, best we do do at this point is extrapolate statistically having a large sample set in order to draw reasonable conclusions.
Bottom line, I will take Roy! LOL