Week 5 Rankings using Best Mathematical Fit for 22-4A and 21-3A

[WOS87]

I crunched a few numbers today and these are just how the teams rank in each of the two districts using the results of all 8 district games that have been played so far.  It's the optimum relative strength of each team that in retrospect would most closely explain the outcome of every game played so far. 

If so inclined, you can use the numbers to make predictions.  Determining the difference of the numbers below for any two teams within a district should give you the predicted margin of victory or loss when those two teams play.

The 21-3A numbers match the actual margins of victory for every district game played  +/- 1.25 points!  The error in 22-4A is a little more at +/- 5.50 points.

21-3A

(best fit using results of all district games thru 2 weeks of round-robin play)

1. 54.25 - West Orange-Stark (beat H-F by 27, beat Kvile by 24)

2. 42.25 - Silsbee (beat OF by 41, beat BC by 42)

3. 29.00 - Kirbyville (beat Jasper by 15, lost to WOS by 24)

4. 28.50 - Hamshire-Fannett (lost to WOS by 27, beat H-J by 29)

5. 12.75 - Jasper (lost to Kville by 15, beat OF by 10)

6. 1.50 - Orangefield (lost to Silsbee by 41, lost Jasper by 10)

7. 0.75 - Hardin-Jefferson (beat BC by 2, lost to H-F by 29)

8. 0.00 - Bridge City (lost to H-J by 2, lost to Silsbee by 42)

22-4A

(best fit using results of all district games thru 2 weeks of round-robin play)

1. 42.5 - Dayton (lost to Lumberton by 1, beat Ozen by 33)

2. 38.0 - Lumberton (beat Dayton by 1, beat Vidor by 19)

3. 29.0 - Nederland (beat Vidor by 21, beat Central by 11)

4. 22.5 - Port Neches-Groves (beat Ozen by 2, beat LCM by 28)

5. 15.0 - Beaumont Ozen (lost to PNG by 2, lost to Dayton by 33)

6. 13.5 - Vidor (lost to Ned by 21, lost to Lumberton by 19)

7. 12.5 - Beaumont Central  (beat LCM by 7, lost to Ned by 11)

8. 0.00 - Little Cypress-Mauriceville (lost to Central by 7, lost to PNG by 28)

Before anyone starts attacking, these rankings are not necessarily how I particularly feel teams actually rank.  It's purely a statistical fit counting the results of each game played equally.  (Yes, I realize Lumberton beat Dayton, but Dayton being ahead of Lumberton by 4.5 pts compared to Lumberton winning by 1 pt is within the error of 5.5 points.)

[bullets13]

my main problem with your system is that you have a team with a loss ahead of three teams that are undefeated in district, including the team that handed them that loss. also by using this system, there is no statistical representation of quality of opponent.

[WOS87]

my main problem with your system is that you have a team with a loss ahead of three teams that are undefeated in district, including the team that handed them that loss. also by using this system, there is no statistical representation of quality of opponent.

??  It's because there IS statistical representation of quality of opponent that that team with a loss ended up above three teams that are undefeated.  Just because a potentially 5th place team starts off the season going 2-0 by beating the 7th and 8th place teams, doesn't make them better than a 2nd place team that loses to the 1st place team and starts off 1-1. 

[Guest airitout123]

I believe these are just interesting numbers.  I do not believe that you can quantify a statistical analysis based on opponents that in some cases are not the same.

[tvc184]

I do not believe that you can quantify a statistical analysis based on opponents that in some cases are not the same.

I think that is what WOS87 said. Those were not his feelings on how good the teams may be but purely statistics.

Like you said, they are interesting numbers. Everyone knows however that you can't compare records of who you play and you can't compare scores. Teams in these types of statistics basically get punished because their opponents had lousy records.

[bullets13]

??  It's because there IS statistical representation of quality of opponent that that team with a loss ended up above three teams that are undefeated.  Just because a potentially 5th place team starts off the season going 2-0 by beating the 7th and 8th place teams, doesn't make them better than a 2nd place team that loses to the 1st place team and starts off 1-1.    

this makes no sense.  the team in first lost to a team that is upper middle of the pack, and their only win is against a team that will struggle to finish higher than 6th.  nederland beat the weakest team and a team that could make the playoffs, and lumberton beat the defending district champion and the weakest team... i have a problem with a scale that puts a team with no quality wins and a pretty bad loss (at least in years past) ahead of several undefeateds.

[ozens fastest@1906]

this makes no sense.  the team in first lost to a team that is upper middle of the pack, and their only win is against a team that will struggle to finish higher than 6th.  nederland beat the weakest team and a team that could make the playoffs, and lumberton beat the defending district champion and the weakest team... i have a problem with a scale that puts a team with no quality wins and a pretty bad loss (at least in years past) ahead of several undefeateds.

Chill Out Man, the guy wrote in his post that they're just numbers... It's ALRIGHT!!!