2014 Betting Picks - Super Bowl XLIX

Pick 1: New England -1 correct NFL point spread pick

I hope people cleaned up (and made up losses) last week. I personally had a close to do or die parlay that paid big time. The Super Bowl pick is not an easy one given I personally and the predictive model are not picking the defending champs. There are a few things that worry me: Seattle's passing defense, being there last year, the inflated ball scandal, Pats losing last two Super Bowl appearances, Lynch. But there are others that I like: Brady, Belichick, 3-time champs, offense, Seattle defensive injuries. Regardless, I go with the numbers and they're pointing at NE. Shop around, you might save a point and get them at pick'em.

Since my models have home field advantage as a variable, I ran it twice as if the game was in NE or SEA. Regardless, the results were preferring New England. The confidence number was hard to calculate. It was above 58% when assuming NE at home but close to 54 if away. So basically what you see below is the average. As you see, normally in the regular season, this would be a no bet. But c'mon, this is the Super Bow. So this is how we're finishing the season.

Going with 20% of bankroll which makes a $1500 point spread bet for the Super Bowl game Feb 1st in Arizona.

Finally, I want to thank everyone for the positive and constructive comments. I will continue to learn, improve and enjoy doing this for next year. Until then..Go Patriots!

GameVegas LineEstimatePrediction-VegasConfidence
NEW ENGLAND - SEATTLE 1 8.4 7.4 56.3%

How to read the table:
  • Vegas Line: A NEGATIVE number implies the point spread favors the HOME team
  • Estimate: NFL Pickles' point spread prediction
  • Pred-Vegas: Subtraction: POSITIVE implies VISITING team will cover point spread.
  • Confidence: The probability that the point spread pick is on the correct side.

Comments

Anonymous said…
Nice job overall on the year and GL with the SB pick. I'm not among the haters, but I also didn't buy any picks. I think it's worth pointing out that your winning percentage (since you first started) does not significantly outperform a coin flip. That is, the proportion of winners vs. losers is not statistically significantly better than 50%.

That was actually true early in the year before the losing streak, the p-value will be less significant now. When you factor in the vig to actually make a profit, the results are statistically even less impressive.

So as someone commented last week, there are no guarantees when gambling and next year you may hit at 58-60% and make a profit, but people should realize, that statistically speaking, one shouldn't be highly confident of doing better than 50%. And should have no confidence that they'll do better than 55%.

Buyer beware.

Arguably, people would be better off saving their money and betting other "proven" approaches (double digit dogs, betting behind the line move). You're model is unlikely to be superior to those approaches.

I think you've built something interesting, but the truth is, the Vegas line leaves very little inefficiency for people to consistently take advantage of.

Finally, to be nit picky, I'm not sure why you've made a point of mentioning several times that your model should be good at picking the blowouts. The blow outs, most often, represent the extremes of the distribution of game results. Meaning, as you know, if the teams played 100 times, a blow out would occur a handful of times. A good model can't pick this out unless you've somehow identified unique variables that tell you when a team is about to flop or have a let down? Seems unlikely, that would be a very complex model requiring a great deal of granularity and complexity, I'm skeptical that's the type of model we're dealing with here.

In cases where a 30 point win is the average result of 100 games, then yes, hopefully models can predict that. Those don't represent the majority of NFL blow outs however. Just my 2 cents.
johnbart said…
Best if luck in the SB and moving forward. I like the blog and the conversations that fall out of it. I think you're on the right path. Keep up with it. If you're not already looking at logarithmic regressions you should definitely checked them out.

My model line has SEA by 0.5 on a neutral field which is pretty much a coin flip. I would stay away from a side pick and go with the under if it goes back up. I'm on under 49 and would go with anything at our above that.

RJW said…
Hopefully its a great game this year.
My Playoff Picks are at www.nflocksmith.com
Good luck!
Jaime said…
I doubled check my prediction with 15,000 fans' opinions on who would win. Not always the fan base is aligned with the spread. In this case it is but the margin seems wider than what the spread indicates.

The Patriots will Deflate The Seahawks’ Hopes this Super Bowl
johnbart said…
I'll take Nate Silver's opinion over 15,000 random fans any day :)
Jaime said…
@johnbart Silver was wrong. Thank you for all your comments this year. You are right, the best thing about this blog is the discussions and conversations that come from it. See you next year.
Jaime said…
Good start, mediocre season, and great finish. Went 6-2, 75% ATS during the playoffs, which brought us back to positive ground. Ending at around 10% ROI but only 53% ATS.
johnbart said…
Congrats on the NE win but I don't think Nate Silver was wrong. His line had SEA as a slight edge but basically a coin flip. Any game that comes down to a final play on the 1 yard line with 20 seconds left is indeed a coin flip.

The real time probabilities numbers that Brian Burke runs had the chance of a Seattle win at 75% at the snap of that 2nd down play.

I think the correct pick was to not pick. There wasn't enough edge either way to justify a bet unless you were willing to gamble. Nothing wrong with gambling if you're just looking for fun but not something I would bet a large chunk of my bankroll on.

johnbart said…
Is everyone getting ready for Week 1 of the Season? Is this going to be a hot spot this season? :)
Jaime said…
I'm pumped! Picks will be free this year, at least until the playoffs. So yes, let's make this a hot spot where everyone can share opinions on various games.