College Football - Early Season Fun

[Drake Tungsten]

So if this is a genuine factor in determining results, should we compensate by giving the away team 3 points to start out on the official score?

No. Home-field advantage isn't constant over time and varies from stadium to stadium.

I think that makes my point quite accurately.

You don't seem to have a point.

[MRT144]

So if this is a genuine factor in determining results, should we compensate by giving the away team 3 points to start out on the official score?

No. Home-field advantage isn't constant over time and varies from stadium to stadium.

I think that makes my point quite accurately.

You don't seem to have a point.

So you'll use the home field advantage to make a subjective point, but you won't compensate for it at all in the actual score because it isn't a reliable 3 point factor in outcomes?

[Drake Tungsten]

So you'll use the home field advantage to make a subjective point, but you won't compensate for it at all in the actual score because it isn't a reliable 3 point factor in outcomes?

You can't compensate for home-field advantage before a game if you aren't sure what the expected point differential caused by home-field advantage is going to be. You have to compensate for it later when you do know that expected point differential.

[MRT144]

So you'll use the home field advantage to make a subjective point, but you won't compensate for it at all in the actual score because it isn't a reliable 3 point factor in outcomes?

You can't compensate for home-field advantage before a game if you aren't sure what the expected point differential caused by home-field advantage is going to be. You have to compensate for it later when you do know that expected point differential.

Then why is it always 3? Or on a historical basis when has it not been 3?

[Drake Tungsten]

Then why is it always 3?

It isn't. Sagarin's average home-field advantage for NCAA football is currently 2.98, for example. Was 2.91 last week. Analyses of home-field advantage in the NFL also yield results different from 3...

http://www.twominutewarning.com/hfa.htm

[SlowwHand]

I haven't seen a spread of 2.98.

[MRT144]

Then why is it always 3?

It isn't. Sagarin's average home-field advantage for NCAA football is currently 2.98, for example. Was 2.91 last week. Analyses of home-field advantage in the NFL also yield results different from 3...

http://www.twominutewarning.com/hfa.htm

So are you an advocate of actually adjusting scores after the game in the official record to reflect that, and in case of tie forcing a overtime game at some point? What's your opinion on rounding?

[Drake Tungsten]

I haven't seen a spread of 2.98.

Shut up, Donny. You're out of your element.

So are you an advocate of actually adjusting scores after the game in the official record to reflect that, and in case of tie forcing a overtime game at some point?

No. That's ****ing retarded.

[MRT144]

I haven't seen a spread of 2.98.

Shut up, Donny. You're out of your element.

So are you an advocate of actually adjusting scores after the game in the official record to reflect that, and in case of tie forcing a overtime game at some point?

No. That's ****ing retarded.

But magical points in your mind which is one step away from adjusting the official record isn't retarded, especially because you use it as a basis to ignore the official record?

[Drake Tungsten]

No, statistical analysis is not retarded; it's a much better way to attempt to determine which team is truly the best than simply looking at wins and losses is.

[MRT144]

No, statistical analysis is not retarded; it's a much better way to attempt to determine which team is truly the best than simply looking at wins and losses is.

Except you're using the home field advantage points (as Sagarin outlines) as actual points that just aren't counted in the official record to actual games to back your opinions. The way he uses them is solely in the predictor comparison between two teams, and in looking at actual matchups, the value of the predictor with the home field advantage factored in becomes somewhat dubious;

Does anyone really believe that Notre Dame is a 5 point underdog at home to Oklahoma?
Nebraska is a 1 point favorite at home to Florida?

So, you're not only using the home field factor in the wrong way, even the way he uses it is suspect.

You're using statistical mishmash.

[Drake Tungsten]

What are you talking about? Where did I say that I was using Sagarin's home-field advantage number for anything? The only reason I mentioned it is because you seemed to think home-field advantage is always three points () and it was the only HFA number I could recall off the top of my head. Dude, you treading awfully close to tardville here...

[Ming]

Sure, you have to look at more than simply wins and losses. Miami has played all their games except one against a ranked opponent. Their one loss record is far better than most of the other one loss teams simply on strength of schedule. Heck, I think they are better than some of the no loss teams that have only played "little sisters of the poor'... So I agree that statistical analysis is important. However, the example of a team lossing 3 away games by one point is a good one. Are you going to consider them an undefeated team?

At the end of the day, it will come down to win's and losses. Sure, a one loss team from a major conference with a tough schedule is going to be considered better than a perfect team from a crappy conference that never played anybody... but a team that losses 3 games, even if by one point on the road, will NOT be considered for the national championship unless everybody else sucks

[MRT144]

What are you talking about? Where did I say that I was using Sagarin's home-field-advantage number as anything but an example? Dude, you treading awfully close to tardland here...

I'm just trying to figure out your methodology and the basis for it. So far it just seems kind arbitrary with cherry picked stats for the appearance of there being a firm foundation to which teams you think are better than others.

So pouring over every single game, how many losses were not losses and what are the resultant record outcomes of that adjustment?

[Drake Tungsten]

I'm just trying to figure out your methodology and the basis for it.

Methodology and basis for what?