Value Add Injury Adjusted Bracket 28 of 32, 99th Percentile

The theory of Value Add Basketball is to calculate the number of points each player is worth to his team above a replacement player, and so far applying it to this year’s bracket correctly predicted 28 of 32 first round games, good for the 99th percentile of all brackets submitted to ESPN (see all picks posted March 14).

The bracket, published here before the tournament started, missed only in picking wins by Maryland, SMU, Miami and Seton Hall. Of the 10 systems covered by the Breitbart pre-tournament piece, the only other system to match the 28 of 32 mark was the system adjusting for team match-ups that favor one team (e.g. 3-point shooting team against poor 3-point defense, etc.). The “Team Match-ups” system also has all Sweet 16 teams still intact, while the Value Add Injury system lost only SMU from it’s Sweet 16.

The system is two wins better than “going chalk,” the practice of simply picking the better seed in each game which actually does better than almost all brackets, and even better than the Vegas odds. The following are the results of all systems provided prior to the tournament. The six systems that went 27 of 32 are in the 95th percentile, while the two systems that went 26 of 32 as well as “going chalk” would put you in the 85th percentile of all ESPN brackets.

All 11 systems missed USC’s stunning come back win over SMU, the 13th time USC has come back from double digits this year (best in the country) and the first time in seven years a team from the play-in game was able to win their first round game as well. The missed picks from Friday are shown on the table below, while the Thursday misses were listed in Friday’s Breitbart follow-up.