March 23, 2023
InsideHitter.com published its first player evaluator for the season this past Sunday. IH applies a proprietary algorithm to measure each players contributions for the season and gives an “overall” impact rating to each player at the D3 level. With over 1,700 rostered players with statistics, and removing players who did not reach a minimum of 25 sets played for the season, we pulled the top 6 players from each team and ran the math based on ratings, rankings, and percentile of the players. Our traditional rating and ranking algorithms are based on team statistics, margin of victory, schedule strength, and numerous other factors related to overall statistics. This analysis just looks at player talent and individual statistics.
In the look below, we are pulling the overall ratings of the top 6 players and analyzing averages for the top “starters” who have impacted their teams. You would expect that the teams with the top players in the country would yield the top-rated teams in the country. But this is an interesting look for games played thus far with many of the top ranked teams falling into line at the top, but some outliers are certainly apparent.
The analysis would be different if based on smaller or larger player counts. We chose 6 to represent the starting lineups but in most cases, this was not the case. The 25-set minimum came into play as some teams did not have 6 players that met the criteria.
The Delta % is a key factor in this analysis as you can see the spread between the best player on the team and the lowest rated player in the top 6. You can also make an assumption that a team is only as good as its worst rated starter. Teams like Juniata, Stevens, Vassar and NYU are clearly deeper than most teams and Stevens is showing the most depth with a deep list of players exceeding the 80th percentile of players overall.
As we head toward year end playoff activity in each conference, this analysis can be used to predict potential upsets in conference playoffs if players continue to perform to their potential. But we all know how a team gels together will dictate their success rather than individual performances.
All comments welcome. The algorithm is still in beta and will continue to be refined in the future. Last season, the IH algorithm performed at a 86% game prediction win rate and we have experienced similar results this season.