Link to ECU
Link to Tulsa
Link to Cinci
Link to Tulane

Hey All,

Sorry for the delay, I was away getting a terrible sunburn of which i'm still recovering from

SMU (click for 2018 preview)

Your job is simple: Vote for the teams you think SMU will beat OR if you think SMU will beat none of them, only vote for the "Winless" option.

SMU fans, feel free to convince pollsters why or why not they should win every game on their schedule!

I'll post one poll a week until all 12 football teams are represented, in reverse order of conference finish. Each new post will link to the preceding ones, and polls will be open until the end of the 12 weeks.


For a more informed opinion, here's some reading material on their opponents:

North Texas

NAVY

UCF

Tulane

Cincinnati

Houston

UCONN

Memphis

Tulsa

East Carolina

As always, SMU has an incredibly tough schedule

5-7

I thought these were supposed to come out every Friday...

More Kruciff LIES!!!

SMU can just go ahead and spell Tulsa L-O-S-S.

(05-07-2018 05:47 PM)Hurricane Drummer Wrote:  SMU can just go ahead and spell Tulsa L-O-S-S.

What is Tulsa backwards?

(05-07-2018 05:47 PM)Hurricane Drummer Wrote:  SMU can just go ahead and spell Tulsa L-O-S-S.

SMU has historically played Tulsa very well. The teams SMU has not fared very well against are Navy, Houston, and Tulane.

(05-07-2018 06:00 PM)JHG722 Wrote:  

(05-07-2018 05:47 PM)Hurricane Drummer Wrote:  SMU can just go ahead and spell Tulsa L-O-S-S.

What is Tulsa backwards?

As Lut.

By far the best schedule for 2018. Nothing to sneeze at, considering that we have 8 teams with at least 2 P6 OOC opponents this year.

Not sure I understand the point of the results... we voted which game we thought SMU would win, but the results don't portray SMU's predicted win-loss record, merely the percentage of votes out of all the votes cast going to each team, without any way to tell if that percentage equates to a predicted win or not.

Ya, I know, it's the limitations of the website poll setup. But the poll results are kind of meaningless.

(05-08-2018 10:29 AM)Coog Engineer Wrote:  Not sure I understand the point of the results... we voted which game we thought SMU would win, but the results don't portray SMU's predicted win-loss record, merely the percentage of votes out of all the votes cast going to each team, without any way to tell if that percentage equates to a predicted win or not.

Ya, I know, it's the limitations of the website poll setup. But the poll results are kind of meaningless.

It's based on the limitations of the polling system, yes, but if a finite number of people can pick anywhere between 1 and 12 options, that should provide a measure for how confident participants are in X team beating Y team.


Between 0 and 44.9% I list the result as a loss

Between 45% and 55% I list the result as a tossup, providing a range for the teams record to fall in

Between 55.1% and 100% I list the result as a win.

Until a better system is established, this is the best I can do.

Either the poll is limited, this board has 10x as many trolls as legitimate posters, or the posters here are idiots if only 20% of the voters think SMU will beat HBU.

(05-08-2018 10:57 AM)mustangxc Wrote:  Either the poll is limited, this board has 10x as many trolls as legitimate posters, or the posters here are idiots if only 20% of the voters think SMU will beat HBU.

The poll is limited.

My results are a function of total unique voting members. The poll does it as a result of total votes.

For example, if you have 100 voters who each vote a win (12 votes) that means you would have 1200 votes, but each position would only have 8.33%

Using my method, in excel, you would have a 100% confidence in a win. Refer back to the other teams polls for an example. You can find the link at the top of the OP.

Motion to impeach Kruciff

Is there a way to show how many members voted? I know the website keeps track of that because it reminds me I already voted in the poll.

The best way to make the numbers show confidence on percentage scale is simply take the number of votes cast per team and divide by number of voters, voila.

So like, if EVERYONE is predicting SMU to beat HBU so far, then there have been 19 voters so far, and so 19/19 = 1.0 or 100%. Likewise only 16 of the 19 voters thought SMU will beat UNT so that's 16/19 = .84 (84%), etc. and so on.

That would show the confidence level on a normalized scale that is a little more meaningful... >50% could be considered a predicted win, for example.

(05-08-2018 11:49 AM)Coog Engineer Wrote:  Is there a way to show how many members voted? I know the website keeps track of that because it reminds me I already voted in the poll.

The best way to make the numbers show confidence on percentage scale is simply take the number of votes cast per team and divide by number of voters, voila.

So like, if EVERYONE is predicting SMU to beat HBU so far, then there have been 19 voters so far, and so 19/19 = 1.0 or 100%. Likewise only 16 of the 19 voters thought SMU will beat UNT so that's 16/19 = .84 (84%), etc. and so on.

In the bottom right of the poll. click on [Show Results | Edit poll]