re: The SECRant Season Poll Results

quote:

---

And some fancy math, here are some estimated/updated results...

---

Not too shabby.

quote:

---

This ain't my first rodeo kid...

---

I thought you were kidding.

There's no way you could have figured out the actual results based on what I posted.

Fair enough.

quote:

---

No, but I could make an estimation. (which is likely good enough in a poll posted for "fun".)

If 52 votes represent the 20% of the total (based on what you posted) then that means there were an estimated total of 260 votes. That means you discarded approximately 208 votes.

Then you take the totals from before and break them apart based on the percentages in the first thread. (52 * x%). Then you add back in the 208 votes to the win column and you get the above percentages.

I couldn't have known, but I can make some guesses based upon information gleaned from the thread. Again, this is just for fun (right?) and doing simple math was a fun way to exercise my brain.

We'll see how close I was if and when you update the OP...

---

I forgot to mention that I threw out incomplete polls.

quote:

---

The remaining responses were compiled and analyzed on a game-per-game basis (for example, responses for Auburn vs Kentucky and responses for Kentucky vs Auburn were added) and used to determine the winner.

---

Yet another fly in the ointment when Alabama has 103 votes and Kentucky only has 51 votes.

This explains why LSU with 182 votes (far and away the highest number of poll votes) was predicted to go 12-0. If you look at just the SEC matchups and take the number of votes that came from the LSU poll vs their opponents poll you get the following:

These are the percentages of the votes in the matchups that came from just the LSU side of the poll.

At South Carolina 77%
Ole Miss 72.8%
At Arkansas 55.1%
At Texas A&M 64.7%
Alabama 64.4%
At Florida 77.1%
Vanderbilt 88%
Oklahoma 58.9%

Compared to the actual poll results you can see some of the built in bias.

At South Carolina 88% W
Ole Miss 56% W
At Arkansas 88% W
At Texas A&M 72% W
Alabama 51% W
At Florida 84% W
Vanderbilt 88% W
Oklahoma 79% W

For example, look at both Ole Miss and Bama. Both of those have higher percentage of LSU voters than actual poll results. This could indicate that a majority of unbiased voters thought these two teams would beat LSU. But those results are lost in the noise because of a greater than 2-1 advantage of LSU votes that came from the LSU poll.

quote:

---

I forgot to mention that I threw out incomplete polls.

---

Which is also bogus because it simply means people might not have had an opinion on some games because of lack of information. But it doesn't mean the votes they did submit are INVALID.

For example, if someone votes for president, their congressman, senator and governor does that mean their ballot is invalid if they do not vote for dog catcher, city or town council? No.

quote:

---

At South Carolina 77%
Ole Miss 72.8%
At Arkansas 55.1%
At Texas A&M 64.7%
Alabama 64.4%
At Florida 77.1%
Vanderbilt 88%
Oklahoma 58.9%.

---

How did you get those numbers?

quote:

---

Something is fishy. The volume of votes per team doesn’t seem to match volume of posters here. OU at 127 when Bama is at 103? No way we have 127 fans voting here so there was likely a lot of troll votes. And no way we have more fans posting here than Bama. Just an observation, but it would skew results to not have real fans vote for each team.

---

We have established that this entire survey is bogus.

Welcome to the thread.

Have fun with it or don't. It's up to you.

quote:

---

That's not how you spell ointment.

---

I'm old with bad eyes. It happens.

Obviously, this was not weighted based on fanbase participation.