Perfect Example Why Computer Polls Are Worthless

[FatCat]

Class 4A

Maryville remains #1, followed by defending 3A champion Ridgeway who moved up to 4A classification this year. Pearl-Cohn and Cleveland are new to top 10, replacing John Overton and Greeneville. Jackson Northside and Chattanooga Central are the biggest gainers in 4A.

 

1 Maryville (2-0)

2 Ridgeway (2-0)

3 Melrose (1-1)

4 Hillsboro (0-1)

5 Knoxville Central (2-0)

6 Memphis East (1-1)

7 Morristown West (0-1)

8 White House (1-0)

9 Pearl-Cohn (1-0)

10 Cleveland (1-0)

 

Not trying to slight enall, but something must be wrong with the computer model when a team like Pearl-Cohn easily defeats a team like Hillsboro, and is still behind them in any kind of rankings. I know that past performance weighs into the ranking, but once a team soundly defeats another, they should outrank them no matter the system.

 

I am not talking about just winning. Upsets happen. You can't move someone high just because they defeat a team. But, from what I have seen, Pearl-Cohn didn't get lucky. They controlled the game from the opening gun and strongly defeated Hillsboro. How then can Hillsboro still be so highly ranked?!? Just a sanity check on the system.

[bulldog69]

Some of these rankings are from reputation not current ability...especially in 3a.

[enall]

Class 4A

Maryville remains #1, followed by defending 3A champion Ridgeway who moved up to 4A classification this year. Pearl-Cohn and Cleveland are new to top 10, replacing John Overton and Greeneville. Jackson Northside and Chattanooga Central are the biggest gainers in 4A.

 

1 Maryville (2-0)

2 Ridgeway (2-0)

3 Melrose (1-1)

4 Hillsboro (0-1)

5 Knoxville Central (2-0)

6 Memphis East (1-1)

7 Morristown West (0-1)

8 White House (1-0)

9 Pearl-Cohn (1-0)

10 Cleveland (1-0)

 

Not trying to slight enall, but something must be wrong with the computer model when a team like Pearl-Cohn easily defeats a team like Hillsboro, and is still behind them in any kind of rankings. I know that past performance weighs into the ranking, but once a team soundly defeats another, they should outrank them no matter the system.

 

I am not talking about just winning. Upsets happen. You can't move someone high just because they defeat a team. But, from what I have seen, Pearl-Cohn didn't get lucky. They controlled the game from the opening gun and strongly defeated Hillsboro. How then can Hillsboro still be so highly ranked?!? Just a sanity check on the system.

I have always stated that computer rankings, at least mine, are really not very accurate until the 5th week of the season. After that time they are pretty good. The computer has to get to know the teams for a given year. Hillsboro/Pearl-Cohn is a great example. Given 2-3 more weeks the computer will adjust to just how good Hillsboro and Pearl-Cohn are - and if Pearl-Cohn is a better team they will replace Hillsboro in the standings.

 

The computer cannot, after one week in the season determine who are the best teams in the state. Give them a little time.

 

In the ten years of doing the Prep Performance Ratings, there has never been a team that won a state championship that was not rated in the top 10 at the week 10 rating period. I think that says that they are a pretty good measure of teams strengths. Some years back, Maryville won the state after losing their first 4 games. The Prep Ratings were the only ranking or rating system that had them in the top ten on the final rating. The same can be said for the first year Pearl-Cohn won their championship with 4 loses.

 

Take the ratings for what they are worth -- a guide as to the relative strength of each team in the state.

[FatCat]

enall,

 

Thanks for the reply. Again, its not a knock on your poll specifically. It is just a reflection of computer polls in general. I understand your point about giving it about 5 weeks to work it all out. But, being an analyst myself, I can't help but think there has got to be some method of making the numbers work out at the beginning of the season......not that it would be easy.

 

Anyway, I agree that by the end of the season, your rankings are pretty darn close to where they ought to be.

[TankTalk]

The ratings also help me fill up space on my sports pages Earl, so keep them coming!

[mapperman]

Mr. Enall,

 

Thank you for doing the poll. You are appreciated!

 

Mapp

[Teddavid]

Thank you for doing the poll. You are appreciated!

 

 

Ditto that Earl. Thanks for all you have done for Tennessee High School Sports.

[Biggamer]

the devils should be rated better than #53. we played two asweome triple a schools. give us some respect. we ready. we ready. we ready. we ready. Aaaaaaaaaaaaaaaaaaah............Yeaaaaaaaaaaaaaaaaaah!

[Opperman]

enall,

 

Thanks for the reply. Again, its not a knock on your poll specifically. It is just a reflection of computer polls in general. I understand your point about giving it about 5 weeks to work it all out. But, being an analyst myself, I can't help but think there has got to be some method of making the numbers work out at the beginning of the season......not that it would be easy.

 

Anyway, I agree that by the end of the season, your rankings are pretty darn close to where they ought to be.

There really isn't a way to do that, though. At least not objectively.

 

Give it a month and then it'll be pretty accurate.