An "Odd" Twist on Predicting the Season

[IrishLax]

I'll keep this as brief as possible as it's slightly numbers heavy. In a nut shell, there are early lines available for all Notre Dame games but Wake Forest and Navy. Why Wake Forest and Navy? Because we are too big of a favorite for them to put out a reasonable line this early. So we can chalk those up as wins (yes... we are chalking up Navy as a win... DEAL WITH IT). Using standard point spread to moneyline conversions, we get the expected probability of Notre Dame winning each game on its schedule. As you should all be well aware, Vegas is (typically) correct half of the time and incorrect half of the time (basically following normal standard deviation rules) and as such can be used as quite a reliable predictor of rough probability over the course of large period of games. In this case, that period will be the Notre Dame regular season:

ND -8 vs. South Florida = .7826
ND -3.5 @ Michigan = .6364
ND -8.5 vs. Michigan St = .7826
ND -10 @ Purdue = .8182
ND -4 vs. USC = .6667
Stanford -5 vs. ND = .2941
ND -4.5 @ Pitt = .6875
ND -10 vs. Air Force = .8182
ND -8.5 @ Maryland = .7826
ND –infinity vs. Wake Forest = 1.0
ND –infinity vs. Navy = 1.0
ND -8 vs. BC = .7826

Add all those up and you get an expected wins total of 9.0515 wins. Not too bad huh? Well it gets better. Brian Kelly has outperformed the spread over his career as a BCS coach. In the four years he has coached at Cincinnati and Notre Dame he is roughly 27-21 ATS. Doesn't sound awesome... but basically he outperforms expected results by 12.5% which is a big deal. Increase 9.0515 by 12.5% (this is fuzzy math...) and you get 10.1829 wins.

The bottom line? People thinking 9 to 10 wins are backed up the individual game odds. It should also make people fairly confident betting the OVER on 8.5 wins if you're that kind of person.

One last point of interest... Notre Dame is a substantial favorite (1 score+) in 8 of 12 games, an underdog in only one and greater than 3 point favorite in all others. On "expected result" analysis ND wins all 8 games where it's a heavy favorite 2 out of 3 where it is a greater than 3 point favorite and loses the one where it is a greater than 3 point underdog. What does that add up to? 10 wins.

People should find 9 wins acceptable, 10 wins expected and 11 wins exceptional. 8 wins would be a sore disappointment and 12 would be a ridiculous over-performance. At least that's what the fuzzy math says

 

[Anchorman]

At least until you consider that we're assuming a 100% chance of beating someone who is 3 of the last four against us and we haven't beaten on our turf since the last Pope.



And before we get to declaring the Navy BS is over, I present some gems from last year

ND beats Navy by at least 2 scores this year... you can take that to the bank

Navy will get stomped for reasons you can't quantify, Tulsa is a joke...

 

[IrishLax]

At least until you consider that we're assuming a 100% chance of beating someone who is 3 of the last four against us and we haven't beaten on our turf since the last Pope.



And before we get to declaring the Navy BS is over, I present some gems from last year

Tulsa is a different story... that was a perfect storm of crazy events leading to that loss... but man Navy last year was an absolute *** kicking. Did not see that coming. Sheesh. I still have no idea what exactly happened. But it happened.

The only reason I chalk that up as a 1.0 is because Vegas considers us too much of a favorite to put any odds on the game and we're -10 against AF (who is substantially better than Navy even before Navy's graduations). Don't argue with me, argue with Vegas.

 

[DomeX2 eNVy]

Great stuff LAX. This is the way to do it. I ran my odds a couple months ago and came up with 9.43 win- so I'm obviously a homer. Actually I only gave WF a 0.99 and Navy a 0.80 (maybe it isn't listed because Navy is 100% - haha) I was more optimistic on Maryland, Pitt, and Purdue to make up the difference (I haven't checked your math. . . Yet)

Regardless, gotta love the USC number and that we'll be favored in 11/12 games- as of today.

I have to agree on the Over bet being the right side. Kudos for the interesting topic. Love this stuff! You're my boy!

 

[mgriff]

Nice work. I could've cut all that work out and told it's exactly how I've been viewing the upcoming season. I really like that the numbers worked with how I'm feeling.

 

[Anchorman]

With regard to the projection, given ND's massive fanbase and its zeal, the spreads are almost always more in our favor than reality dictates, meaning that percentages are in fact lower.

 

[koonja]

Cool stuff!

 

[Mr. McGibblets]

Ejoyed that breakdown. Also, great thread title.

 

[IrishLax]

With regard to the projection, given ND's massive fanbase and its zeal, the spreads are almost always more in our favor than reality dictates, meaning that percentages are in fact lower.

I thought that too at first, but that's actually an incorrect assumption. Even with "bad breaks" last year Notre Dame was roughly 6-4 against the spread (College Football Against the Spread Rankings at VegasInsider.com, The Leader in Sportsbook and Gaming Information - College Football Against the Spread Rankings actually lists us as 7-4-2 but I went with the conservative number). During Charlies' reign, where we constantly did not meet expectations and lost to teams like Syracuse, UConn and Navy (for the first time in forever) even Charlie was ~25-26 ATS... nearly .500. So no, it does not appear that ND's fanatical fan base really affects the lines in Vegas how you are proposing it does.

 

[notredomer23]

great post! So that gives us a 2% chance of going undefeated. Sounds about right.
11 wins is 10% chance

Not bad odds. While I think we do lose to Stanford RIGHT NOW, we will have to see how the season plays out because they lost a lot last year. Even with a returning Andrew Luck they may not be as good.

 

[Old Man Mike]

To my understanding, though, Anchorman is correct on what Vegas is doing. They are not a prognosticating service but rather are attempting to judge where to place odds in order to split the betting public. Notre Dame, if viewed historically by Vegas as having "enthusiastic" betting support, will get higher favoritism in order to split some pro-ND bets down to the competition.

In Notre Dame's hey-days, the word in Vegas was that "Notre Dame always beats the spread", because we were a juggernaut and you can't put up odds TOO high because of the risk-aversion psychology of betters finally kicking in. But lately we are not [yet] a juggernaut, so Vegas can relax a bit more and just put out a favorable "ND Homer" style of odds.

Having said all that, I still believe that ND will match the Vegas 'non-evaluation'. That's because the Kelly Factor hasn't fully sunk in yet with the betting public. He will overcome the Vegas over-rate by his own excellence. And until this season is over, the betting public still won't believe that ND is back to being an a$$-kicker. The following year, I'd forget Vegas as being any kind of guide whatever, because everyone will know that we ARE back, and betting services will be over-rating us universally; but it won't matter because we'll be kicking a$$ anyway. [maybe not by quite as many points all the time --- that might take one further year].

 

[IrishLax]

To my understanding, though, Anchorman is correct on what Vegas is doing. They are not a prognosticating service but rather are attempting to judge where to place odds in order to split the betting public. Notre Dame, if viewed historically by Vegas as having "enthusiastic" betting support, will get higher favoritism in order to split some pro-ND bets down to the competition.

In Notre Dame's hey-days, the word in Vegas was that "Notre Dame always beats the spread", because we were a juggernaut and you can't put up odds TOO high because of the risk-aversion psychology of betters finally kicking in. But lately we are not [yet] a juggernaut, so Vegas can relax a bit more and just put out a favorable "ND Homer" style of odds.

Having said all that, I still believe that ND will match the Vegas 'non-evaluation'. That's because the Kelly Factor hasn't fully sunk in yet with the betting public. He will overcome the Vegas over-rate by his own excellence. And until this season is over, the betting public still won't believe that ND is back to being an a$$-kicker. The following year, I'd forget Vegas as being any kind of guide whatever, because everyone will know that we ARE back, and betting services will be over-rating us universally; but it won't matter because we'll be kicking a$$ anyway. [maybe not by quite as many points all the time --- that might take one further year].

That is 100% correct... however Vegas, because it balances the books, ends being 50% correct each way over a large sample size. That is, teams will typically beat the spread 50% of the time. Notre Dame is no exception; so while it's not Vegas' job to prognosticate... that is what they implicitly end up doing.

Also, while the ND homer theory has a large impact on ND's to-win-the-national-championship odds (skewing them way lower than what they should be) that is only because it is a one-way proposition. On a two-way prop like a point spread it's important to remember that the system self-regulates itself.. and if there were ever the perception (or reality) that homers were driving odds too far in favor of ND there are plenty of gamblers who would drop money the other way and drive the odds to where they "should" be.

 

[Old Man Mike]

All true. Both points of view are generally compatible though. The danger in considering the Los Vegas odds too seriously is because they DO have the Notre Dame fan/better favoritism factored in to their initial guesses and not just an somewhat more objective mere analysis of team strengths. Naturally there will be some degree of correlation as even betters realize that they are living in the real world, but the correlation between Vegas odds-making and simple team evaluations will not be identical because of the emotional fan factor.

But, as I said, the nation is not convinced yet of the Kelly factor, so the Vegas vs Football Evaluator difference should not be as great as it is in your usual betting year. Therefore, this line-of-hopefulness might cohere more closer to reality than it usually does. Hopefully, that plays out and ND wins ten games at least. For me, the main obstacles to that are still USC/Stanford and the screwed-up other loss to whomever.

 

[crzychris]

At least until you consider that we're assuming a 100% chance of beating someone who is 3 of the last four against us and we haven't beaten on our turf since the last Pope.



And before we get to declaring the Navy BS is over, I present some gems from last year

+1

Gone are the days when Navy served as our annual punching bag. Hopefully BK changes that trend this season.

 

[Whiskeyjack]

+1

Gone are the days when Navy served as our annual punching bag. Hopefully BK changes that trend this season.

Gone are the days of Ricky Dobbs and (possibly) Ken Niumatalolo (if UNC snaps him up).

The team we face this year will not look anything like its recent teams.

 

[JughedJones]

Gone are the days of Ricky Dobbs and (possibly) Ken Niumatalolo (if UNC snaps him up).

The team we face this year will not look anything like its recent teams.

And aren't there new cut blocking rules this year?

That would cut deep for Navy.

 

[Buster Bluth]

That would cut deep for Navy.

 

[amgarvey]

Admittedly, I'm not a big sports bettor, but why would vegas skew the odds in favor of q side that is going to be heavily bet on anyway? That seems to be a losing proposition.