Thanks for continuing to add fun (and informative) stats! I'm all for a little quantitative subjectivity and this set of measures is perfect for that. Observers do not have to value wins and losses the same if they have different criteria for what merits tournament selection.
For example, some mid- and low-major teams are not given the opportunities to build a large Win Quality stat; some plaudits have to be given to North Texas, the 29th ranked Loss Quality team. Given their schedule to date, they have only had the chance to rack up 4.0 units worth of Win Quality, bringing down their Resume Quality. Is their earned-to-opportunity ratio of 2.1/4.0 good? (Probably not, but it's certainly better than Wake and Georgia and about the same as Arizona! Though it's somewhat glaring that even without facing the toughest opponents, they lost all 4 of their most difficult games.)
Gonzaga is not in the top 35 of the resume quality rankings; yet, the team is recognized as one of the top 10 teams (#7 at the time of this comment) at evanmaya.com. I realize that different metrics are being compared but there needs to be a moreodicum of consistency.
Gonzaga is one of the biggest discrepancies you will ever see between just how the wins and losses look on paper and how good the team actually is. It's really quite remarkable.
With that said, if Gonzaga is a top 10 team, they should be able to win enough games beyond what would be expected from a bubble team in order to get their resume quality closer to their team ranking.
The NCAA publishes the raw WAB numbers daily. I would think that the NCAA WAB is derived from the NET, which is supposedly a predictive rating. There were 49 teams with positive NET WABs through 1/29/25.
I think you need both. The resume-only metrics are the most fair way to decide who should get into the tournament, but the team strength models are helpful to figure out the seeding to make the bracket matchups more fair.
That's so fascinating. I never thought about using the different type of models to figure out seedings and matchups VS simply deciding who gets into the tournament but now it seems so obvious I feel a little silly.
For example, if Gonzaga ends up in the tournament, it would be pretty unfair if they were given an 11 seed to whoever the 6 seed is. The Zags would almost certainly be favored in that game. They are top 10 at my site still!
Thanks for continuing to add fun (and informative) stats! I'm all for a little quantitative subjectivity and this set of measures is perfect for that. Observers do not have to value wins and losses the same if they have different criteria for what merits tournament selection.
For example, some mid- and low-major teams are not given the opportunities to build a large Win Quality stat; some plaudits have to be given to North Texas, the 29th ranked Loss Quality team. Given their schedule to date, they have only had the chance to rack up 4.0 units worth of Win Quality, bringing down their Resume Quality. Is their earned-to-opportunity ratio of 2.1/4.0 good? (Probably not, but it's certainly better than Wake and Georgia and about the same as Arizona! Though it's somewhat glaring that even without facing the toughest opponents, they lost all 4 of their most difficult games.)
I haven't run the numbers, but I'm fairly confident Indiana State would have easily made the field according to Resume Quality last year.
Gonzaga is not in the top 35 of the resume quality rankings; yet, the team is recognized as one of the top 10 teams (#7 at the time of this comment) at evanmaya.com. I realize that different metrics are being compared but there needs to be a moreodicum of consistency.
Gonzaga is one of the biggest discrepancies you will ever see between just how the wins and losses look on paper and how good the team actually is. It's really quite remarkable.
With that said, if Gonzaga is a top 10 team, they should be able to win enough games beyond what would be expected from a bubble team in order to get their resume quality closer to their team ranking.
The NCAA publishes the raw WAB numbers daily. I would think that the NCAA WAB is derived from the NET, which is supposedly a predictive rating. There were 49 teams with positive NET WABs through 1/29/25.
Evan, Auburn beat Purdue in a neutral site game.
That's my bad, updated the article, thanks.
Evan, does your win quality metric factor in the margin of victory (or loss) as part of its calculation?
It does not! Purely based on wins and losses.
Thanks, Evan - what do you think about models that factor that in? Do you not think it's useful or helpful?
I think you need both. The resume-only metrics are the most fair way to decide who should get into the tournament, but the team strength models are helpful to figure out the seeding to make the bracket matchups more fair.
That's so fascinating. I never thought about using the different type of models to figure out seedings and matchups VS simply deciding who gets into the tournament but now it seems so obvious I feel a little silly.
For example, if Gonzaga ends up in the tournament, it would be pretty unfair if they were given an 11 seed to whoever the 6 seed is. The Zags would almost certainly be favored in that game. They are top 10 at my site still!