You may have noticed that we didn't make any predictions for this week's NFL games.
There's a very good reason for that.
We apparently suck at it.
Interestingly, the system which we use at FOLLOWING FOOTBALL, kind of related to the ELO chess rating system where the level of 'surprise' in a victory or loss determines how many points the team that underperforms "gives" to the team that overperforms predictions, works beautifully for the CFL (where we went about two out of three in betting against the spread, and placed in the top 10% in the CFL.ca competition again this season). It works wonderfully for Australian footy (where we again placed in the top 10% this year, having been in the top 5% the two previous years). And in ESPN's "Capitol One Bowl Mania", we're currently in the top one percent of the 800,000 or so contestants, having picked eleven of twelve winners in the first part of the bowl season.
But in the NFL? We're losing our shirts.
We can pick winners relatively well. But our system is forecasting point spreads that, when they disagree with Las Vegas, almost always disagree the wrong way. THAT'S no bueno!
So, we're hanging up our "expert's nametag" for the season, and we'll spend the intervening time trying to refine our grading system to make it work for the NFL. Meanwhile, we'll keep sharing the college, Canadian, and Australian predictions, because those rock the house!
But the NFL? You're on you're own, pilgrim. Sorry!
No comments:
Post a Comment