(01-14-2016 07:04 PM)quo vadis Wrote:  Updated for Bowl results.

These should be normalized by the number of bowl games played by each P5 conference. They're not equal.

(01-15-2016 11:46 AM)MplsBison Wrote:  

(01-14-2016 07:04 PM)quo vadis Wrote:  Updated for Bowl results.

These should be normalized by the number of bowl games played by each P5 conference. They're not equal.

What do you mean?

(01-16-2016 04:28 AM)quo vadis Wrote:  

(01-15-2016 11:46 AM)MplsBison Wrote:  

(01-14-2016 07:04 PM)quo vadis Wrote:  Updated for Bowl results.

These should be normalized by the number of bowl games played by each P5 conference. They're not equal.

What do you mean?

B1G only had 8 results. SEC had 11.

Point is: let's say they were equal going into the bowls and both B1G and SEC went undefeated (I know that's not possible -- just stay with me). Well, both should stay equal, then.

But your system would give the SEC 11 points and the B1G only 8. That's not fair.

In fact, that's how it should be for all five of the P5, for the whole season.

In other words, this is what you statistic should be: avg points per contest.

Because not all the P5 are playing the same number of conf games and not all the non-conf contests provide equal opportunity for points scoring.


So if the stats were like this:

B1G: 50 points in 28 contest, SEC: 65 points in 37 contests, then the metric would be this

B1G: 1.79 avg pts/contest
SEC: 1.76 avg pts/contest

(01-16-2016 01:04 PM)MplsBison Wrote:  In fact, that's how it should be for all five of the P5, for the whole season.

In other words, this is what you statistic should be: avg points per contest.

Because not all the P5 are playing the same number of conf games and not all the non-conf contests provide equal opportunity for points scoring.


So if the stats were like this:

B1G: 50 points in 28 contest, SEC: 65 points in 37 contests, then the metric would be this

B1G: 1.79 avg pts/contest
SEC: 1.76 avg pts/contest

You were 4-4 and got blown out and shut out in the most important game. We were 9-2 and won the national championship. End of story.

(01-16-2016 01:29 PM)JRsec Wrote:  

(01-16-2016 01:04 PM)MplsBison Wrote:  In fact, that's how it should be for all five of the P5, for the whole season.

In other words, this is what you statistic should be: avg points per contest.

Because not all the P5 are playing the same number of conf games and not all the non-conf contests provide equal opportunity for points scoring.


So if the stats were like this:

B1G: 50 points in 28 contest, SEC: 65 points in 37 contests, then the metric would be this

B1G: 1.79 avg pts/contest
SEC: 1.76 avg pts/contest

You were 4-4 and got blown out and shut out in the most important game. We were 9-2 and won the national championship. End of story.

Haha. Thank you.

What was all that ridiculous convoluted formula BS.

Whoever came up with that

The very fact that a conference had more bowl teams is itself a measure of strength. To deny this is another form of bias. I have no problem judging the SEC on total wins rather than "normalized" wins.

(01-16-2016 01:01 PM)MplsBison Wrote:  

(01-16-2016 04:28 AM)quo vadis Wrote:  

(01-15-2016 11:46 AM)MplsBison Wrote:  

(01-14-2016 07:04 PM)quo vadis Wrote:  Updated for Bowl results.

These should be normalized by the number of bowl games played by each P5 conference. They're not equal.

What do you mean?

B1G only had 8 results. SEC had 11.

Point is: let's say they were equal going into the bowls and both B1G and SEC went undefeated (I know that's not possible -- just stay with me). Well, both should stay equal, then.

But your system would give the SEC 11 points and the B1G only 8. That's not fair.

Well, a few factors come to mind:

1) My champ is crowned via win credit %, not total wins. So it doesn't matter if one conference has 30 chances to earn wins and another has 25, what matters is the win %. If anything, my system is slightly biased towards conferences with FEWER chances, not more, because the "regression to the mean" concept tells us that it's harder to maintain high performance over many chances than fewer (e.g., it's not hard for an NBA player to have a perfect 100% free throw percentage if he only has taken two free throw attempts, but it's harder to do that if he's taken 10 attempts, harder still to do it over 50 attempts).

If anything, it's to a conference's credit that it scheduled (regular season) or earned (bowl season) more games against other P5 opponents or ranked opponents.

2) Also, I'd still argue that you can't ignore games. For example, the only conferences in the national title game were the ACC and SEC. So because the other three P5 conferences didn't have a chance to play that game, does that mean we should ignore the result of the ACC and SEC champs playing each other? I don't think that's reasonable. More is lost by ignoring that game than counting it.

So anyway.

(01-17-2016 07:54 AM)Hokie Mark Wrote:  The very fact that a conference had more bowl teams is itself a measure of strength. To deny this is another form of bias. I have no problem judging the SEC on total wins rather than "normalized" wins.

... and what makes Bison's critique even sillier is that my system is in fact based on win %, not total wins. For example, the ACC had more total wins than the Big 12, but they also had even more total losses, so the Big 12 is ranked ahead because the rankings are based on win %.

(01-16-2016 01:04 PM)MplsBison Wrote:  In fact, that's how it should be for all five of the P5, for the whole season.

In other words, this is what you statistic should be: avg points per contest.

Because not all the P5 are playing the same number of conf games and not all the non-conf contests provide equal opportunity for points scoring.


So if the stats were like this:

B1G: 50 points in 28 contest, SEC: 65 points in 37 contests, then the metric would be this

B1G: 1.79 avg pts/contest
SEC: 1.76 avg pts/contest

That's a Hail Mary if I have every seen one.

The league that fell on its face this bowl season was the AAC. They can thank Houston for saving some face

(01-17-2016 11:02 PM)rednblackattack Wrote:  The league that fell on its face this bowl season was the AAC. They can thank Houston for saving some face

At least the ACC was there to break the fall. The AAC won 2 out of 3 bowl games against them.

And yet, the ACC for the 2nd time in 3 years was playing for a national championship.

C-USA took care of business against the AAC this bowl season

(01-17-2016 11:10 PM)rednblackattack Wrote:  And yet, the ACC for the 2nd time in 3 years was playing for a national championship.

C-USA took care of business against the AAC this bowl season

FSU and Clemson are fantastic programs. You should thank them for carrying the conference year in and year out.

Louisville is normally good too. But you lost to an AAC team as well this year.

(01-17-2016 11:12 PM)malenko2 Wrote:  

(01-17-2016 11:10 PM)rednblackattack Wrote:  And yet, the ACC for the 2nd time in 3 years was playing for a national championship.

C-USA took care of business against the AAC this bowl season

FSU and Clemson are fantastic programs. You should thank them for carrying the conference year in and year out.

Louisville is normally good too. But you lost to an AAC team as well this year.

GA Tech won the Orange Bowl last season. Yes we lost to Houston, however they ended up being pretty good. How do you explain C-USA and the MAC having no trouble with the AAC?

(01-17-2016 11:18 PM)rednblackattack Wrote:  

(01-17-2016 11:12 PM)malenko2 Wrote:  

(01-17-2016 11:10 PM)rednblackattack Wrote:  And yet, the ACC for the 2nd time in 3 years was playing for a national championship.

C-USA took care of business against the AAC this bowl season

FSU and Clemson are fantastic programs. You should thank them for carrying the conference year in and year out.

Louisville is normally good too. But you lost to an AAC team as well this year.

GA Tech won the Orange Bowl last season. Yes we lost to Houston, however they ended up being pretty good. How do you explain C-USA and the MAC having no trouble with the AAC?

CUSA and the MAC are irrelevant to this discussion and GT was two years ago now. This past season, the AAC had a winning record against the ACC in both the regular season and the bowls. If you are going to rip on them, you should take that into account.

If the ACC destroys the AAC head to head next year, feel free to gloat.

(01-17-2016 11:07 PM)malenko2 Wrote:  

(01-17-2016 11:02 PM)rednblackattack Wrote:  The league that fell on its face this bowl season was the AAC. They can thank Houston for saving some face

At least the ACC was there to break the fall. The AAC won 2 out of 3 bowl games against them.