SWC75
Bored Historian
- Joined
- Aug 26, 2011
- Messages
- 33,489
- Like
- 64,483
We often hear teams being evaluated at the end of the season by how they did “against ranked teams”. A couple of years ago I came up with a simple system for evaluating this, one with an acknowledged flaw but which nonetheless produced interesting results so I’ll do it again this year.
The system is very simple. First I hand out points to teams based on the ranking of their opponent. If you play the #1 ranked team, you get 25 points for playing such a powerful team. You get 24 points for playing the #2 team, 23 for playing the #3 team, etc. I’ll call them schedule points. Here’s a chart to make it easy:
1-25, 2-24, 3-23, 4-22, 5-21, 6-20, 7-19, 8-18, 9-17, 10-16, 11-15, 12-14, 13-13, 14-12, 15-11, 16-10, 17-9, 18-8, 19-7, 20-6, 21-5, 22-4, 23-3, 24-2, 25-1.
That’s step one. Now we play the games and see how these teams actually do on the field. What cued me into this is that most games will be won or lost by 25 points or less, so the points involved are comparable. You just compare the point differential to the points teams have gotten for playing a ranked opponent. If they beat the ranked opponent, their points will grow by the margin of victory. If they lose, you subtract the margin of victory from the schedule points. If that reduces them to zero or less, they get zero points. I’m not going to deal with minus points, just add up the positive points earned, or else my rankings would be cluttered up with a lot of inferior teams who were below the Mendoza line because they’d gotten blown out by their ranked opponents. If you play the #1 team and lose by 21 points, you’ll get 4 points: 25 schedule points minus the 21 points you lost by. If you play the #10 team and lose by 13 points, you’ll get three points. If you play the #15 team and lose by 14 points, you’ll get nothing, (11-14 = 0 in this system).
I’ll record the points earned each week and how the teams got them and add them up over the course of the season. That will show us how each team did “against ranked teams”. Like all cumulative standings, they will sort themselves out as the season goes on and come more and more into focus. Later listings will be more meaningful than early ones as more and more teams get through the meat of their schedules. The cumulative listing should I think be less a ranking of teams than a statistic you look at in determining where to rank the teams.
They system is simple and fun but has an obvious flaw: the ranking and quality of an opponent will change over the course of a season. Three years ago when we played Northwestern, they were ranked #19 in the country and the ease with which they whipped us suggested they might go a lot higher than that. They won their first four games and appeared to be on their way to beating Ohio State. But they let that game get away from them and seemed to lose heart after that. They lost two games by a field goal and two by a touchdown. They wound up 5-7. We actually wound up with a better record at 7-6 with a bowl game win. We didn’t get close enough to them to earn any points but if we had, wouldn’t those points have been tainted by their subsequent collapse?
I don’t think so, because when we played them, they were as good as their ranking, as we found out. They got worse afterwards. The reverse argument could also be made. When Virginia Tech beat Ohio State in their second game last year, the Buckeyes were rated #8. The Hokies won by 14 so under my system they got credited with 32 points: 18 schedule points plus the 14 points they won the game by. Considering the way Ohio was playing when they cruised to the national championship game at the end of the year, shouldn’t the Hokies have gotten more credit than that? Well, if they were playing the Ohio State team we saw at the end of the year, they wouldn’t have won by 14 points or indeed, at all. They themselves wound up 7-6. I’m giving them points based on how good Ohio State was, or was thought to be at the time they played them, (and they scored pretty well at that).
It could be more meaningful just to wait until the end of the season and use the same system based on the final rankings, (or at least the final regular season rankings). But as I’ve described that could be deceptive, too. But that can be easily done at that time. I wanted a system where we don’t have to wait until the end of the season to use it. I wanted to have fun seeing the points add up as we go along. So here goes.
WEEK ONE (9/1-9/5/2016)
Alabama beat #20 Southern California 52-6 52 points
Houston beat #3 Oklahoma 33-23 33 points
Florida State beat #11 Mississippi 45-34 26 points
Wisconsin beat #5 Louisiana State 16-14 23 points
Texas beat #3 Notre Dame 50-47 19 points
Auburn lost to #2 Clemson 13-19 18 points
Texas A&M beat #16 UCLA 31-24 17 points
Georgia beat #22 North Carolina 33-24 13 points
Mississippi lost to #4 Florida State 34-45 1` 11 points
Appalachian State lost to #9 Tennessee 13-20 10 points
Kansas State lost to #8 Stanford 13-26 5 points
Oklahoma lost to #15 Houston 23-33 1 point
The system is very simple. First I hand out points to teams based on the ranking of their opponent. If you play the #1 ranked team, you get 25 points for playing such a powerful team. You get 24 points for playing the #2 team, 23 for playing the #3 team, etc. I’ll call them schedule points. Here’s a chart to make it easy:
1-25, 2-24, 3-23, 4-22, 5-21, 6-20, 7-19, 8-18, 9-17, 10-16, 11-15, 12-14, 13-13, 14-12, 15-11, 16-10, 17-9, 18-8, 19-7, 20-6, 21-5, 22-4, 23-3, 24-2, 25-1.
That’s step one. Now we play the games and see how these teams actually do on the field. What cued me into this is that most games will be won or lost by 25 points or less, so the points involved are comparable. You just compare the point differential to the points teams have gotten for playing a ranked opponent. If they beat the ranked opponent, their points will grow by the margin of victory. If they lose, you subtract the margin of victory from the schedule points. If that reduces them to zero or less, they get zero points. I’m not going to deal with minus points, just add up the positive points earned, or else my rankings would be cluttered up with a lot of inferior teams who were below the Mendoza line because they’d gotten blown out by their ranked opponents. If you play the #1 team and lose by 21 points, you’ll get 4 points: 25 schedule points minus the 21 points you lost by. If you play the #10 team and lose by 13 points, you’ll get three points. If you play the #15 team and lose by 14 points, you’ll get nothing, (11-14 = 0 in this system).
I’ll record the points earned each week and how the teams got them and add them up over the course of the season. That will show us how each team did “against ranked teams”. Like all cumulative standings, they will sort themselves out as the season goes on and come more and more into focus. Later listings will be more meaningful than early ones as more and more teams get through the meat of their schedules. The cumulative listing should I think be less a ranking of teams than a statistic you look at in determining where to rank the teams.
They system is simple and fun but has an obvious flaw: the ranking and quality of an opponent will change over the course of a season. Three years ago when we played Northwestern, they were ranked #19 in the country and the ease with which they whipped us suggested they might go a lot higher than that. They won their first four games and appeared to be on their way to beating Ohio State. But they let that game get away from them and seemed to lose heart after that. They lost two games by a field goal and two by a touchdown. They wound up 5-7. We actually wound up with a better record at 7-6 with a bowl game win. We didn’t get close enough to them to earn any points but if we had, wouldn’t those points have been tainted by their subsequent collapse?
I don’t think so, because when we played them, they were as good as their ranking, as we found out. They got worse afterwards. The reverse argument could also be made. When Virginia Tech beat Ohio State in their second game last year, the Buckeyes were rated #8. The Hokies won by 14 so under my system they got credited with 32 points: 18 schedule points plus the 14 points they won the game by. Considering the way Ohio was playing when they cruised to the national championship game at the end of the year, shouldn’t the Hokies have gotten more credit than that? Well, if they were playing the Ohio State team we saw at the end of the year, they wouldn’t have won by 14 points or indeed, at all. They themselves wound up 7-6. I’m giving them points based on how good Ohio State was, or was thought to be at the time they played them, (and they scored pretty well at that).
It could be more meaningful just to wait until the end of the season and use the same system based on the final rankings, (or at least the final regular season rankings). But as I’ve described that could be deceptive, too. But that can be easily done at that time. I wanted a system where we don’t have to wait until the end of the season to use it. I wanted to have fun seeing the points add up as we go along. So here goes.
WEEK ONE (9/1-9/5/2016)
Alabama beat #20 Southern California 52-6 52 points
Houston beat #3 Oklahoma 33-23 33 points
Florida State beat #11 Mississippi 45-34 26 points
Wisconsin beat #5 Louisiana State 16-14 23 points
Texas beat #3 Notre Dame 50-47 19 points
Auburn lost to #2 Clemson 13-19 18 points
Texas A&M beat #16 UCLA 31-24 17 points
Georgia beat #22 North Carolina 33-24 13 points
Mississippi lost to #4 Florida State 34-45 1` 11 points
Appalachian State lost to #9 Tennessee 13-20 10 points
Kansas State lost to #8 Stanford 13-26 5 points
Oklahoma lost to #15 Houston 23-33 1 point