Hyperlarge point spreads

[albionmoonlight]

http://64.233.161.104/search?q=cache...s&ct=clnk&cd=1

Some early lines are out for the first college football games of the season. And there are, as you would expect, some very large lines in mismatch games: Kentucky-Louisville, La. Tech-Nebraska, UAB-Oklahoma, and my favorite North Texas-Texas (40 point line).

I know that the guys who make point spreads do it in a way to try to get the betting equal on each game. And I know that they are very good at setting initial lines, and tend to be really good at breaking down matchups.

And this makes sense to me when you are saying things like UGA is a field goal better than LSU at home. Or Miami is 10 points better than Florida this season.

But how can anyone really say that Texas is 40 points better than North Texas? Isn't it totally random at that point? Texas may be able to beat North Texas by 60 if they tried to run up the score. Or they may win by "only" 21 if they decide to work out some kinks in the offense and give the backups major playing time. It seems like the margin of victory in a game like that becomes a function of random factors, most of which are just not knowable until they play the game.

So, my question/observation--how do they decide to set a line in an obvious mismatch? 25 points? 30 points? 40 points? It all seems like the same thing to me, when you get down to it.

[digamma]

I suppose you don't want the generic answer that they set the line where they think they can get the closest to 50-50 action on the game?

[st.cronin]

It's equally random at ANY point. Team A is better than Team B by 3 points - well, it's reasonable that either team could win by 10 or more points.

[albionmoonlight]

Originally Posted by digamma

I suppose you don't want the generic answer that they set the line where they think they can get the closest to 50-50 action on the game?

That's not only the generic answer, but the only one that makes sense. I just can't see how they can predict what people will do with numbers that large.

[Celeval]

Notre Dame -8 at Georgia Tech... hmm.

[stevew]

I'd be interested in seeing the cover-rate on those huge spread type games. Despite the lines being 20+ points, I would think that the favorites would still cover a large portion of the time.

[kcchief19]

Originally Posted by albionmoonlight

That's not only the generic answer, but the only one that makes sense. I just can't see how they can predict what people will do with numbers that large.

My answer is that handicappers/oddsmakers are not experts at determining whether or not Texas is 40 points better than North Texas. That doesn't matter. What they are experts at is determing what gamblers will think.

By and large, I would imagine they use the same formulas that the majority of professional bettors use to determine the initial line by using a number of splits and previous outcomes to determine the most likely result.

Plus, the line is fluid. Texas could end up a 30-point favorite or a 50-point favorite depending on which way the action goes to ensure a 50/50 split.

[Huckleberry]

Originally Posted by albionmoonlight

That's not only the generic answer, but the only one that makes sense. I just can't see how they can predict what people will do with numbers that large.

I think that at some level the large spread works in favor of 50/50 accuracy instead of against it.

A guy that's willing to bet Texas as a 35 point favorite is probably going to bet them at 40, also. A guy that's going to bet North Texas to beat a 40 point spread would probably have the same reaction to a 35 point spread. When you're analyzing how to bet the difference between 35 and 40 seems much less important than the difference between 3 and 8.

[digamma]

Originally Posted by albionmoonlight

That's not only the generic answer, but the only one that makes sense. I just can't see how they can predict what people will do with numbers that large.

There was a real sports segment on setting lines a few years ago. From what I remember, it involves people and computers running tons of numbers to come up with a first cut at lines. The odds makers will come up with a feeler line that they will run by consulting companies and/or big fish. It's massaged a bit based on feed back there and put out for public consumption. That probably way over-generalizes the process and there are probably google links that better explain it, but that's how I generally remember it.

[Vegas Vic]

Originally Posted by albionmoonlight

Some early lines are out for the first college football games of the season. And there are, as you would expect, some very large lines in mismatch games: Kentucky-Louisville, La. Tech-Nebraska, UAB-Oklahoma, and my favorite North Texas-Texas (40 point line).

Oklahoma was a 31 point favorite over TCU last year. Oops. Take UAB plus the points.