I was going to say that it doesn’t seem that bad from this perspective.

By my algorithms, Broncos had the best performance of the week.

For anyone that is interested here is the results from my algorithms with explanations of what they mean below.

PR_T is the ranking number here. It is an average of RP, RD, and RT (all using standard deviation in an attempt to give them equal weight), and then add in the WAE.

RP is a mean average of how many points the team scores against the average opponents’ scores. Losing by 3 points to a team that usually wins by 10 nets 7 points to the total (before dividing by games played). Beating a team by 3 points that usually loses by 7 nets -4 to the total (before dividing by games played).

RD is very similar to RP except it is what percentage of the score was by the team. A team that beat another by 30-20 had 60% of the score. This again is compared to the opponent’s opponents average. The teams score is the mean average of that.

RP favors offense a bit where RD favors defense a bit.

RT is one that I started tracking a few year ago. It is like RP and RD, but instead of using points it uses time. The winning team gains the seconds remaining in regulation after the winning score and the losing team gets that time subtracted. It is also compared against the opponent’s opponents average and is presented in standard deviation from the mean form. The value here is divided by 36 to make it more like a percentage of a game as opposed to raw seconds. This has been pretty darn good as a power ranking method.

WAE is “Wins Above/Against Expectation”. This uses the strength of schedule and win% to give a rough estimate on how many wins the team is above or below what an average team would have with the schedule that has been played. An average team with a .500 schedule would expect to be .500. An average team with a .600 schedule would expect to be .400. If you multiply the WAE by how many games have been played, it shows how many games the team won or lost compared to what an average team would be expected to do.

Bounce (formerly “Over”) is something else that I’m monitoring. I created it last year specifically because the of the Vikings season. This column is a measurement of how teams are performing against the stats. I’m still trying to work my head around it. A zero here does not mean the team is playing exactly as the stats I made up suggest. This is not completely worked out.

Regardless, the lower the number in the Bounce column is the more a team is losing more than it probably should. The higher the number, the more the team is winning when it probably should lose. For reference, last year after week 16 (last I ran the numbers that season), the Vikings were 5.070 in this column. The next highest was the Colts at 2.399. The lowest was the 49ers at -2.887.

Below is the rated performance of each team this week using mostly the same method. It doesn’t add the WAE.

For anyone that is interested here is the results from my algorithms with explanations of what they mean below.

PR_T is the ranking number here. It is an average of RP, RD, and RT (all using standard deviation in an attempt to give them equal weight), and then add in the WAE.

RP is a mean average of how many points the team scores against the average opponents’ scores. Losing by 3 points to a team that usually wins by 10 nets 7 points to the total (before dividing by games played). Beating a team by 3 points that usually loses by 7 nets -4 to the total (before dividing by games played).

RD is very similar to RP except it is what percentage of the score was by the team. A team that beat another by 30-20 had 60% of the score. This again is compared to the opponent’s opponents average. The teams score is the mean average of that.

RP favors offense a bit where RD favors defense a bit.

RT is one that I started tracking a few year ago. It is like RP and RD, but instead of using points it uses time. The winning team gains the seconds remaining in regulation after the winning score and the losing team gets that time subtracted. It is also compared against the opponent’s opponents average and is presented in standard deviation from the mean form. The value here is divided by 36 to make it more like a percentage of a game as opposed to raw seconds. This has been pretty darn good as a power ranking method.

WAE is “Wins Above/Against Expectation”. This uses the strength of schedule and win% to give a rough estimate on how many wins the team is above or below what an average team would have with the schedule that has been played. An average team with a .500 schedule would expect to be .500. An average team with a .600 schedule would expect to be .400. If you multiply the WAE by how many games have been played, it shows how many games the team won or lost compared to what an average team would be expected to do.

Bounce (formerly “Over”) is something else that I’m monitoring. I created it last year specifically because the of the Vikings season. This column is a measurement of how teams are performing against the stats. I’m still trying to work my head around it. A zero here does not mean the team is playing exactly as the stats I made up suggest. This is not completely worked out.

Regardless, the lower the number in the Bounce column is the more a team is losing more than it probably should. The higher the number, the more the team is winning when it probably should lose. For reference, last year after week 16 (last I ran the numbers that season), the Vikings were 5.070 in this column. The next highest was the Colts at 2.399. The lowest was the 49ers at -2.887.

Below is the rated performance of each team this week using mostly the same method. It doesn’t add the WAE.