Quite honestly, in the madness of this year’s tournament, many of our recommendations are not doing well. This is always a risk. We’ve run the numbers, and calculated that we can get the average user who participates in four brackets a win about 50% of the time. But this year is outside the 50% for most of those users.
The way we get this edge is by strategically “going long” a set of teams. The teams we were most exposed to this year were: Arizona, Louisville, Virginia, and Villanova. Three of these teams suffered heart-breaking losses last weekend (apparently especially so for some students in Tucson), and Villanova got derailed by a dangerous UConn team that might not be done doing damage. In contrast, the people in best position to win most pools this year are somewhere among the 33% of people that picked Florida to win it all. Our number one gambit was to advise against picking Florida and Michigan State for all but the smallest pools (or some upset pools) in favor of the four teams above. If those four teams had collectively fared better than these two, our brackets would have been in good shape. But in a bit of chaos amongst the favorites this year, Florida was the one one of these 6 teams to make the Final Four, leaving many brackets in tatters.
Given the nature of the bracket this year, a lot of our optimized brackets ended up recommending multiple small gambits as opposed to one major one. For instance, one of our fully optimized brackets for medium sized pools was built to pay off if either Arizona won the tournament or Villanova made the finals (and you got a little help). One downside of such strategies is that when they fail, they often fail spectacularly. In the 49% likelihood that neither of these teams even make the Final Four, your bracket will not fare well. And that is what happened with this bracket. And in the likely scenario where Florida wins at least one more game, my entry that followed this strategy will end up in the bottom 25% of its pool. As will my bracket that took Louisville over UCLA in a 120 person pool.
The remaining bright spots within our bracket portfolio are Florida to win, which many of our users in very small (less than 10 person) pools ended up with, and Wisconsin to win, which was our number two gambit for users in very large pools. By and large though, this year falls in that 50% where most of our users will come out behind. It also breaks a four year stretch for me personally of winning at least one pool.
But we’ll be back next year, bigger and better, with the same great advice. You need a little luck to go along with your good decisions to come out on top in the Madness, and this year Lady Luck, for the most part, has not been on our side.