Debunking some Playoff Metrics

With only 12 data points we have to be more specific and intentional in how we analyze a win and a loss.

Stop using "top 25 win" or "3 loss team". There's too much range, you can't compare a win over a top 5 with a win over #24.

Texas A&M lost 5 games but they were ALL to top 10 teams. This doesn't mean they are a top 10 program but you can't discount a victory over Texas A&M and count a victory over Iowa who would have the same record with the same schedule. Easy to make a case #30 Texas A&M is as good or better than #16 Iowa. You can't do that with the top 15 there's very clear differences.

Margin of Victory - Start using it. With only 12 games you have to factor it in. There's 3 categories to me (1 score 8 points or less) (9-16) and then 16+. If you lose by more than 16 it wasn't competitive. I don't care if you win by 20 or 40 those are dominant wins regardless. "close", "competitive", "convincing".

Strength of Opponent - When you factor this in with Margin of Victory it gives a much better perspective on the Win or Loss.

There's a difference in 11-1 with a close loss to a top 5 program, and 11-1 with a convincing loss to a top 25 program. It's a huge difference actually.

This also figures into wins. If you look at teams with close wins over "bad" (outside the top 25) opponents on multiple occasions it's a sign this wasn't a let down or a trap game, but that the team is probably not as good as their record indicates.

Quality of Loss - Who you lose to and by how much is absolutely important. You can't talk about quality wins and ignore who you lost to and what the situation is. Quality of loss isn't the only thing but it is certainly important when ranking teams.


When comparing teams if you use Top 5 wins, Top 10 wins, top 15 wins, combined with Number and Quality of losses, with Margin of Victory you are going to sound more impartial and less emotional.