This is a cool article, but 1 nitpick: I don’t think you should be including the 0.0 for teams that cancelled the 2020 season in your 5-year average. For example, UConn averaged -12.4 in the seasons they actually played a game, meaning they’re +12.6 this season. By including the 0.0 as their SRS in Y4 you artificially improve their 5Y average to 9.9 and drag their improvement down to “only” +10.2.
This is a cool article, but 1 nitpick: I don’t think you should be including the 0.0 for teams that cancelled the 2020 season in your 5-year average. For example, UConn averaged -12.4 in the seasons they actually played a game, meaning they’re +12.6 this season. By including the 0.0 as their SRS in Y4 you artificially improve their 5Y average to 9.9 and drag their improvement down to “only” +10.2.
And just to follow up, that's now fixed for UConn and ODU in the charts.
That's an excellent catch. Lemme see about fixing that in the chart in the next little bit...