FYI – I’ve explained this before, but in case you missed it and have question:
The Limper Power Rankings are the basic engine of my spreads projection model – minus game-specific matchup data, and the sorting algorithm that puts it all together. There are 4 components: W/L/P Pct., O-AVE, R-AVE, and C-AVE.
The W/L/P Pct. Is a non-traditional percentage based on a team’s win/loss record, also including ties, so that a team like MIN will have a slightly lower Pct. than WSH because of MIN’s tie, and despite both having 5-3 win/loss records.
The O-AVE is a strength-of-opponent calculation that considers both an opponent’s win/loss record, as well as their statistical strengths and weaknesses (see R-AVE). Naturally, the O-AVE for weaker teams would show a higher strength-of-opponent calculation than those of stronger teams, but it’s hardly consistently so. It’s no secret that Oakland is a pretty bad team, but their average opponent turns out to have been also pretty awful – making the Raiders easily the worst team in the NFL. On the other hand, a team like Atlanta has struggled to 4-4 precisely because they’ve played teams that statistically rank in the top 5 of the league, and at 3-5, TB has not only had QB issues, they have had the hardest schedule thus far.
The R-AVE of a team is simply a ranking calculation of several dozen offensive and defensive variables, each of which is weighted semi-subjectively. (Years ago I circulated a questionnaire among football stat-heads at a couple of service sites, which asked them – for example – to assign a significance-weight, 1 to 10, to team and player variables like: yards per game made/allowed, home/away wins and losses, yards per completion, pass completion percentage, sacks made/allowed, etc.) Most of these same variables and weights are used in my spread projections model as well.
The C-AVE is a composite calculation of the above three calculations, with a couple other factors thrown in that didn’t quite fit elsewhere.
Although these power rankings are certainly an after-the-fact calculation, I’ve included a current win/loss column as a kind of grading feature. The week just ended shows that the top 16 teams went 11-4, with those 4 loses occurring against other teams in the top 16, along with a bye. This might be useful if the ROI makes sense for money-line wagers.
The Limper Power Rankings are the basic engine of my spreads projection model – minus game-specific matchup data, and the sorting algorithm that puts it all together. There are 4 components: W/L/P Pct., O-AVE, R-AVE, and C-AVE.
The W/L/P Pct. Is a non-traditional percentage based on a team’s win/loss record, also including ties, so that a team like MIN will have a slightly lower Pct. than WSH because of MIN’s tie, and despite both having 5-3 win/loss records.
The O-AVE is a strength-of-opponent calculation that considers both an opponent’s win/loss record, as well as their statistical strengths and weaknesses (see R-AVE). Naturally, the O-AVE for weaker teams would show a higher strength-of-opponent calculation than those of stronger teams, but it’s hardly consistently so. It’s no secret that Oakland is a pretty bad team, but their average opponent turns out to have been also pretty awful – making the Raiders easily the worst team in the NFL. On the other hand, a team like Atlanta has struggled to 4-4 precisely because they’ve played teams that statistically rank in the top 5 of the league, and at 3-5, TB has not only had QB issues, they have had the hardest schedule thus far.
The R-AVE of a team is simply a ranking calculation of several dozen offensive and defensive variables, each of which is weighted semi-subjectively. (Years ago I circulated a questionnaire among football stat-heads at a couple of service sites, which asked them – for example – to assign a significance-weight, 1 to 10, to team and player variables like: yards per game made/allowed, home/away wins and losses, yards per completion, pass completion percentage, sacks made/allowed, etc.) Most of these same variables and weights are used in my spread projections model as well.
The C-AVE is a composite calculation of the above three calculations, with a couple other factors thrown in that didn’t quite fit elsewhere.
Although these power rankings are certainly an after-the-fact calculation, I’ve included a current win/loss column as a kind of grading feature. The week just ended shows that the top 16 teams went 11-4, with those 4 loses occurring against other teams in the top 16, along with a bye. This might be useful if the ROI makes sense for money-line wagers.
