If you’re following this site, it’s because you think, based on our brilliant analytical skills or some gut intuition, that our bets are +EV. In a previous post I discussed the benefits of “Wong” teasers, bets that cross football’s critical thresholds of 3 and 7 points. Here I’ll explain why, when the opportunity arises, you should consider taking multi-leg Wong teasers off the model, and why, if you don’t agree that we have any sort of edge, you’d be crazy to do the same.
In his March 2017 article on NFL teasers, the Wizard of Odds shows, across a fairly significant sample (+/- 2%σ), that a 6 point Wong teaser has a win rate of 72.54%. With that backdrop, the probability of hitting a given number of legs looks something like this:
Every sports book is going to pay out differently for a given number of legs on a teaser bet. I’ll be using our bookmaker’s odds here, but your mileage will vary depending on where you place your bets. Below are the “to 1” odds, or Hong Kong odds, paid by our bookmaker:
And, as a result of these odds, we can compute the edge on each number of legs by pb – (1-p), where p is our probability (success rate) and b is our odds expressed in Hong Kong format:
|Legs||Success Rate||Hong Kong||Wong Edge|
You’ll notice the edge is pretty close to 0. Based on this alone, we can conclude that we would prefer the Wong teaser over a spread bet at -110, since a spread bet at -110 has a theoretical hold of 4.55% (computed by 1 – 1/overround, or 1-1/(.5238+.5238) for a -110 wager). An explanation of theoretical hold and overround is beyond the scope of this piece, but I highly recommend Ganchrow’s post here.
If you’re concerned about variance, I commend you. The table below shows what the edges look like if you shock the Wizard’s win rate of 72.54% up or down 2σ:
|Legs||Wong low edge||Wong high edge|
And, as you might imagine, dispersion increases significantly as you add legs to the teaser.
If you believe in the model- if you believe it adds expectation to a bet, if you think it makes that 72.54% win rate just a little bit higher, you are best off taking as many legs as you can. Instead of computing edge off a 72.54% win rate, a 74.54% win rate would result in an ever-increasing edge.
|Legs||Success Rate||Wong Edge||Wong Low||Wong High|
The most critical lens I can take is that, as we progress down the model, our win rate would steadily converge on 72.54%. But even if this is the case, the exponentially increasing edge that would arise under a Wong teaser leads me to recommend you take as many Wong teaser legs as can reasonably be obtained.