cfb AP Poll as Accurate as a Coin Toss

[UCFhonors]

Smart move from you to back out of our bet on the two teams who could win the title.

A title that only about 15 teams ever have the hope of being invited to play for. Yeah, I'm too good at math to take that UCFoolish of a bet.

#UCFacts

 

[michaeljordan_fan]

A title that only about 15 teams ever have the hope of being invited to play for. Yeah, I'm too good at math to take that UCFoolish of a bet.

#UCFacts

Which 15? I'm only taking 2. That leaves you 13/15.

 

[Guy Incognito]

As we all should recognize in today's information age, Polls are click bait. It's easy to see how bias and inaccurate they are.

This is especially bad considering that Final polls basically ignore records, and wins & losses.

There have been 1464 teams that have been "ranked" int he AP preseason polls.

Of the teams ranked in the preseason poll, 909 have finished ranked (62.1%).

Of the 701 teams that were ranked in the top 10 in the preseason, 537 have finished ranked, 375 finishing in the top 10.

Preseason #1 teams have finished unranked 6 times.

Preseason #2 teams have finished unranked 5 times.

The following chart shows how many times each position has been ranked and how many times the team at that position finished ranked:

1 69 63
2 69 64
3 69 55
4 69 52
5 69 60
6 70 54
7 68 49
8 69 47
9 70 45
10 69 41
11 63 33
12 61 32
13 62 30
14 62 37
15 62 29
16 63 34
17 63 33
18 60 26
19 63 32
20 65 32
21 29 18
22 30 15
23 30 9
24 31 11
25 29 8

The average winning percentage of teams that were ranked in the preseason is 0.714

The average winning percentage of teams that were ranked top 10 in the preseason is 0.769 which is 10 wins out of 13 games.

1 (686-119-10)--0.848
2 (691-120-14)--0.846
3 (611-183-14)--0.765
4 (614-188-12)--0.762
5 (634-166-18)--0.786
6 (619-187-14)--0.763
7 (581-202-16)--0.737
8 (596-200-13)--0.745
9 (597-211-16)--0.734
10 (554-236-14)--0.698
11 (497-217-20)--0.691
12 (479-229-13)--0.673
13 (491-224-14)--0.683
14 (508-213-11)--0.702
15 (466-240-15)--0.657
16 (485-248-7)--0.660
17 (492-234-15)--0.674
18 (457-236-10)--0.657
19 (486-236-7)--0.671
20 (484-259-13)--0.649
21 (255-106-2)--0.705
22 (242-128-0)--0.654
23 (224-143-1)--0.610
24 (248-138-3)--0.641
25 (214-143-0)--0.599


Of the preseason ranked teams, 56 have finished undefeated and untied.

The lowest starting position to end up undefeated was un-ranked UCF in 2017. Auburn is 2nd at 22 in 2010.

The worse winning percentage among preseason ranked teams was USC in 1957 who finished with a 1-9 record.

162 teams ranked in the preseason finished with a losing record.

#UCFacts

 

[UCFhonors]

My UCFavorite thing about college was leaving the willfully ignorant and tshirt bandwagoners, like you, behind

#UCFacts

 

[Guy Incognito]

My UCFavorite thing about college was leaving the willfully ignorant and tshirt bandwagoners, like you, behind

#UCFacts

 

[Gatorchip]

I see you are not a betting man.

Being correct 62% of the time, on which teams finish ranked, is significantly better results than a coin toss.

If you chose blindly (or keep flipping a coin until you narrow it down from 130 to 25 teams), you have less than a 20% chance of being successful.

 

[UCFhonors]

I see you are not a betting man.

Being correct 62% of the time, on which teams finish ranked, is significantly better results than a coin toss.

If you chose blindly (or keep flipping a coin until you narrow it down from 130 to 25 teams), you have less than a 20% chance of being successful.

Try a re-read of the OP.

#UCFacts

 

[TheReal_NU]

My UCFavorite thing about college was leaving the willfully ignorant and tshirt bandwagoners, like you, behind

#UCFacts

Then why did you go to UCF?

 

[UCFhonors]

Then why did you go to UCF?

During my Freshman, sophomore, Jr & Senior college years.

#UCFacts

 

[Gatorchip]

Try a re-read of the OP.

ok?

cfb AP Poll as Accurate as a Coin Toss

Not true. Coin flip is 50%. An unbiased random poll is 20%. You said you're good at math, so this should be pretty easy to see for you.

It's easy to see how bias and inaccurate they are.

Biased yes. Inaccurate, no.

This is especially bad considering that Final polls basically ignore records, and wins & losses.

No they don't. We don't see teams with losing records finish top 25.

Of the teams ranked in the preseason poll, 909 have finished ranked (62.1%).

62.1% isn't 50-50. having those odds on the table for betting would be a dream come true.

The average winning percentage of teams that were ranked in the preseason is 0.714

The average winning percentage of teams that were ranked top 10 in the preseason is 0.769 which is 10 wins out of 13 games.

Doesn't sound like the W/L is being ignored.

Of the preseason ranked teams, 56 have finished undefeated and untied.

Not bad. how many of the bottom 100 finished undefeated and untied? Just UCF one time? No one else?

The lowest starting position to end up undefeated was un-ranked UCF in 2017. Auburn is 2nd at 22 in 2010.

Cool stat. Validates my point.

162 teams ranked in the preseason finished with a losing record.

Poll has been around like 80 years. They only missed 162 times on predicting the top 20% of the teams? Not bad.

 

[TheReal_NU]

Then why did you go to UCF?

During my Freshman, sophomore, Jr & Senior college years.

#UCFacts

I’m so confused.

 

[UCFhonors]

I’m so confused.

I read it, "when did you go to UCF."

Why I went to UCF is obvi. Best honors college in FL. And best value proposition in the nation. And it's magical

#UCFacts

 

[TheReal_NU]

I read it, "when did you go to UCF."

Why I went to UCF is obvi. Best honors college in FL. And best value proposition in the nation. And it's magical

#UCFacts

So reading comprehension is not needed in their honors program?
Got it.

It’s not like everything that is typed by
<~~~~~~~~~~~~
Is ever in anything other than sarcasm font.

 

[UCFhonors]

ok?


Not true. Coin flip is 50%. An unbiased random poll is 20%. You said you're good at math, so this should be pretty easy to see for you.



Biased yes. Inaccurate, no.


No they don't. We don't see teams with losing records finish top 25.


62.1% isn't 50-50. having those odds on the table for betting would be a dream come true.


Doesn't sound like the W/L is being ignored.

Not bad. how many of the bottom 100 finished undefeated and untied? Just UCF one time? No one else?

Cool stat. Validates my point.

Poll has been around like 80 years. They only missed 162 times on predicting the top 20% of the teams? Not bad.

~62.5% accuracy guessing the top 20% is horrible. I UCForget that people don't understand Standard Deviation.

And that is particularly bad when the Polls basically ignore Wins and Losses and any objective measurement.

#UCFacts

 

[Rolltide94]

~62.5% accuracy guessing the top 20% is horrible. I UCForget that people don't understand Standard Deviation.

And that is particularly bad when the Polls basically ignore Wins and Losses and any objective measurement.

#UCFacts

Let me guess, your UCFhonors is not in a STEM field. English Lit? Gender Studies? Sociology?

 

[7Samurai13]

Let me guess, your UCFhonors is not in a STEM field. English Lit? Gender Studies? Sociology?

Philosophy

 

[Guy Incognito]

Let me guess, your UCFhonors is not in a STEM field. English Lit? Gender Studies? Sociology?

UCFrench and UCFrancophone Studies.

 

[7Samurai13]

UCFrench and UCFrancophone Studies.

UCFellatio

 

[Bayou Tiger]

 

[michaeljordan_fan]

~62.5% accuracy guessing the top 20% is horrible. I UCForget that people don't understand Standard Deviation.

And that is particularly bad when the Polls basically ignore Wins and Losses and any objective measurement.

#UCFacts

Congratulations on finding a bell curve. Unfortunately for you it was irrelevant to the topic.