CFL 2015 Predictions

Welcome to my CFL 2015 predictions page!

This year I've decided to change up my model a bit, and I'm now using an adapted version of the ELO rating system that FiveThirtyEight uses. I've backtested it over a few seasons and tweaked it a bit, and it seems to work fairly well!

One of the big variables of their system that I've left unchanged is how to calculate a team's starting rating following the previous season's results. As a result, the starting position for this year shows Calgary and Hamilton as frontrunners, and until more games have been played this isn't a bad position for the model to be in.

Enjoy!






Current record: 7-8-1 (13-3-0 against Vegas spread)

Week 5 Matchups

Home Team	Win %	Visiting Team	Win %	Exp. Point Spread	Vegas Spread
Ottawa	30%	Calgary	70%	Cal -6.0	Cal -5.5
BC	54%	Toronto	46%	BC -1.0	BC -3.0
Edmonton	74%	Winnipeg	26%	Edm -7.0	Edm - 5.5
Saskatchewan	40%	Hamilton	60%	Ham -3.0	Sas -1.0

Week 4 included an exact 50/50 tie in my model, so I've called that one a draw for myself. Otherwise, the model is holding up quite well against the Vegas spread, but so far is miserably average in predicting individual games.

Week 4 Matchups

Home Team	Win %	Visiting Team	Win %	Exp. Point Spread	Vegas Spread
Montreal	50%	Hamilton	50%	n/a	Ham -3.0
Ottawa	39%	Edmonton	61%	Edm -3.0	Edm -3.5
Saskatchewan	58%	BC	42%	Sas -2.0	Sas -3.5
Calgary	76%	Winnipeg	24%	Cal -8.0	Cal -9.0

Week 3 went much better for the model. For the first time, my model ended up in scenarios where its favorite team won, and was on the right side of the Vegas spread, ending up with a win in both columns.

Week 3 Matchups

Home Team	Win %	Visiting Team	Win %	Exp. Point Spread	Vegas Spread
Edmonton	73%	Ottawa	27%	Edm -7.0	Edm -6.0
Winnipeg	47%	Montreal	53%	Mon -1.0	Win -3.5
BC	57%	Saskatchewan	43%	BC -2.0	BC -3.5
Calgary	68%	Toronto	32%	Cal -5.0	Cal -6.5

Week 2 was abysmal. Let's hope for a better week 3. Part of the reason the model might still be inaccurate is because it's drawing solely last year's performance and the results from week 1. I suspect (hope?) in a month or so it'll be much more accurate.


Week 2 Matchups

Home Team	Win %	Visiting Team	Win %	Exp. Point Spread	Vegas Spread
Winnipeg	49%	Hamilton	51%	Ham -0.5	Ham -2.0
Montreal	41%	Calgary	59%	Cal -2.5	Cal -10.5
Ottawa	49%	BC	51%	BC -0.5	BC -3.0
Saskatchewan	54%	Toronto	46%	Sas -1.0	Sas -3.5

Week 1 was a mixed result for my model. On the one hand, a 1-3-0 record isn't too hot, however my model did tend to err on the side of caution from a spread point of view. It wasn't too great at picking overall winners, but it would have done well against the Vegas point spreads. Hopefully week 2 turns out a bit better, although the spreadsheet is giving two games essentially even odds.


Week 1 Matchups

Home Team	Win %	Visiting Team	Win %	Exp. Point Spread	Vegas Spread
Montreal	78%	Ottawa	22%	Mon -8.5	Mon -8.0
Calgary	68%	Hamilton	32%	Cal -5.0	Cal -6.5
Toronto	47%	Edmonton	53%	Edm -1.0	Edm -7.5
Saskatchewan	64%	Winnipeg	36%	Sas-4.0	Sas -7.5