SAN FRANCISCO, Calif. — After more than a week of conference tournament games and the craziness that comes with teams’ seasons coming down to one game, one half and sometimes even one shot, the real tournament is now upon us.
In my last post I wrote about the Big West Conference men’s basketball tournament and took a semi-analytical dive into who I thought would come out on top.
To put my money where my mouth is I also placed a few bets on each of the tournament games using my model. How did I do?
The Big West tourney took place at the Honda Center in Anaheim, Calif. this past Thursday, Friday and Saturday. Eight total teams made the tournament — only Cal State Northridge as ninth and last team in regular season play didn’t qualify.
My original projections at the chance each team had at winning the tournament and garnering the conference’s automatic bid to the dance were as follows:
- UC Santa Barbara (2) – 26%
- UC Davis (1) – 22%
- UC Irvine (3) – 18%
- Cal State Fullerton (4) – 12%
- Hawaii (6) – 10%
- Long Beach State (5) – 8%
- UC Riverside (8) – 2%
- Cal Poly (7) – 1%
In classic March fashion though, the unexpected happened. While the quarterfinal round saw the top four teams advance, chaos ensued from there.
UC Irvine and Cal State Fullerton, the third and fourth seeds, respectively, beat the top two seeds in their semifinal matchups to advance to the title game. The Titans then went on to defeat UC Irvine 71-55 in Saturday’s finale.
I went a step further than just calculating the odds each of these teams had of winning though. Using a pretty simple, yet crude, model I tried my hand at predicting each of the games within the tournament.
To do this I created a model not dissimilar from pro-football-reference’s win probability model.
In essence, we are assuming the final margin of victory in a given game can be approximated as a normal random variable with a mean of my previously explained SRS model, and a standard deviation of 5.38 — calculated using Big West SRS ratings since the 2009-2010 season.
The Excel function
NORMDIST(x,mean,sigma,True)gives us the probability that a normal random variable with the given mean and sigma is less than or equal to x. In our case we will say x = 0.5.
The probability Cal State Fullerton, the favorite, would beat Long Beach State in it’s opening round game for example was calculated as follows:
Probability= 1 – NORMDIST(0.5,2.1,5.38,TRUE) = 1 – 0.3831 = 0.62. The mean of 2.1 is my calculated point spread using SRS. So the Titans had a 62% chance of advancing to the next round.
If you’ve ever placed a bet in Vegas you’ll notice there are several things one can bet on: The spread, the over/under and the moneyline. I won’t go into an in depth explanation of these terms but for my purposes I focused on betting against the spread — since my SRS ratings model can be thought of as a spread — and the moneyline as this is essentially Vegas’ version of win probability.
I wound up going 6-6-1, six wins, six losses and a push on my straight up bets. Due to the idea of the vig, or juice, I lost money despite the fact that I technically went .500 on the weekend. I laid $100 on each bet thus lost $60 total.
UC Davis especially disappointed as the Aggies failed to cover their first game and lost their second round matchup despite being the favorite. I also really thought UC Santa Barbara was going to at least make the final.
At this point I want to plug the new subscription-based site I discovered ActionNetwork.com. I used their research tools and awesome user interface to track all of these games in real time.
The NCAA Tournament bracket came out on Sunday night so look out for another post on my picks for that, and get stoked because March is finally here.