NFL Confidence Pool Against the Spread Strategy

Picking a confidence pool against the spread is basically the same as the confidence pool picks we know and love, with one minor difference: instead of picking who wins the game outright, you’re picking who wins the game against the spread. So how would the strategy work for this little nuance?

It actually makes things a little simpler. With a normal confidence pool, we need to figure out how likely one team will win. On this site, we use an implied probability given by Vegas Moneyline values to figure out the win probability of a team. With a point spread confidence pool, the key difference is you can almost assume that the spread makes each game a 50/50 proposition (we can test this assumption and tweak off it later). There may be some subtle reasons why it’s not quite 50/50, but it makes for a good start. Basically, this is a normal confidence pool except all the games are 50% win probability.

So how do you pick assuming every game is 50/50, essentially a coin flip? The answer lies in going contrarian. If 100% of the people go with heads, and it’s a 50/50 proposition, you should go with tails. 50% of the time you’ll take first place, and 50% of the time you’ll take last place. Over the long run, you’ll have the same score because the probabilities are the same, but in the short run you’ll win your share of weeks.

How do you assign confidence? In this format, you don’t use win probabilities: they’re all close to 50/50. So all that’s left is going contrarian. Pick the teams that are lowest percentage picked and go high on confidence. As for how much you should pick against the crowd, you’ll need to get a feel for your league to see how contrarian it is. If a lot of people pick against the crowd then you have to as well. If a lot of people pick with the crowd, you can go with the crowd more. Also, if you’re up in the standings, pick against the crowd less. If you’re down in the standings, pick against the crowd a lot.

Overall, this type of confidence pool league is a lot more random. If you don’t know much, you have a much better chance of winning in this format than in the classic format because Vegas gives you the spread.

There are a few sites that I’ve found that attempt to model going against the spread. The best one I’ve found is a model that uses a Monte Carlo simulation from

Anyone out there playing confidence leagues picking against the spread? Any thoughts? Anyone have a league I can join to test this strategy out?