Game Probabilities

Probabilities for today’s games, based on Game Score. They’re the same ones I tweet out every day, but for some reason they are also here. Usually updated between 5pm to 6pm, but sometimes I have a life and am not available at those times. If you’re wondering why I hate your favourite team, you can check their projected depth charts or power rankings to see why.

How Accurate Are These?

Good Question.

The log loss for individual games is 0.667. I think that’s good, but I’m not sure.

Teams favoured to win have a 59.3 percent success rate. The theoretical limit for hockey is roughly 62 percent, so that is pretty good, but the sample size is still small (the percentage has dropped consistently throughout the year as it grows) and there’s a lot of year-to-year variance for these types of things. Due to the complexity of the model and my own incompetence, this model has not been back-tested. Consider this a live test or something. I don’t know if that’s a thing or not. Sorry.

Win Percentages Bins

Last updated: April 6, 2017

50-55%: 245 – 223, 52.4%

55-60%: 236 – 163, 59.1%

60-65%: 158 – 69, 69.6%

65-70%: 58 – 30, 65.9%

70-75%: 9 – 1, 90.0%

75-80%: 1 – 0, 100.0%

Overall: 707 – 486, 59.3%

April 6

Screen Shot 2017-04-06 at 4.20.45 PM.png


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s