TORONTO — Heading into the Division Finals, the BC Lions are the favourites to win the 112th Grey Cup.
If Hamilton and Saskatchewan are favored to meet in grey cup game, should not BC has lower probability behind them?
cfl math is fun
Sask though best record didnt really have very explosive offense,won some very close games, will they be explosive this week?
Ticats who knows what team shows up. If was wagering with money would take Montreal
Saskatchewan of course.
Looks like they’re favoured, but just barely. 52 vs 48. I dunno how Montreal is so low, considering most people think they’ll take out Hamilton.
I’d take these stated probabilities (to the hundredth of a percent, no less) far more seriously if they actually told us their methodology. As it is, it is basically mystical hand-waving.
But it is possible for the probabilities to be as they say. It looks to me as though they are giving some serious home field advantage boost for the probabilities in the divisional finals. Enough so, that they are, for example, favouring SSK to get through to the Grey Cup, but, at the same time, predicting that BC would have a better chance of winning the Grey Cup game than SSK if BC manages to win the division. I think that they also give MTL a better chance of winning the GC game, if they get there, than HAM, but not as dramatically.
It is possible for the probabilities to come out this way. One way is if the GC game head to head probabilities are:
| P | HAM | MTL |
|---|---|---|
| SSK | 61.15% SSK over HAM | 54.70% SSK over MTL |
| BC | 78.20% BC over HAM | 72.40% BC over MTL |
Combine these made-up GC game h2h probabilities with the division final probabilities from the article, West: 51.18% SSK, East: 63.63% HAM and you get the GC winner probabilities from the article. BCs higher number comes from predicted better chances against the East winner.
TL;DR
For the record, before someone jumps on me, I don’t agree with the cfl.ca probability prediction. Certainly the correct numbers are SSK 100% in the Western Final and SSK 100% in the GC. 
Not necessarily. The simulator is saying that although Saskatchewan is a very slight favourite to win the West Final, BC has a better chance of beating either eastern opponent should they make it that far. This results in more scenarios where BC wins it all than scenarios where Saskatchewan wins it all. The sim also very little faith in the east, giving them a less than 1-in-3 combined chance of either of them winning the Cup.
Makes you wonder if the simulator takes the probability of Saskatchewan having too many men on the field at a critical moment into account as well 
You guys have left out one very distinct possibility……that you are all completely nuts. After looking into the crystal ball , consulting the tea leaves , and reading the tarot cards , the answer is obvious…..El Nino is cancelled….which has caused a huge influx of hot air to be expelled onto these message boards.
After scientifically determining the east and west division winners ( coin flips…best eleven out of twenty-one ) it calls for both home teams to advance. And there you have it. Now you won’t have to watch two very exciting playoff games , but you will anyway.
And……..while we’re talking sour grapes………did the Dodger pitchers have a bigger strike zone than what the Jays pitchers received. ( Let it go…..walk away………let it go….walk away…there is no conspiracy ) lol.
Yea you make one critical counting players mistake in a championship game 50 years years ago and nobody lets you forget it😄
Methodology….
1 - Develop a linear optimization model (also known as a “Massey Model”) using previous game scores to determine each team’s power index. Incorporate an exponential decay function to highlight that recent games provide a clearer indication of team strength that games played a few months ago. (Recall the Saskatchewan Roughriders debacle in 2014 when Darian Durant was injured in the Banjo Bowl).
2 - Plug these power indexes into a logistic regression equation to determine probabilities of victory for each team in upcoming games. The parameters for the logistic regression equation were generated through an analysis of seven years worth of CFL regular season games.
3 - Input the probabilities into a spreadsheet simulation model and run the model for 100,000 iterations of the remaining regular season games and concomitant playoff brackets. The model will generate the likelihoods that each team makes the playoffs, earns a home playoff game, finishes first overall in their respective division, qualifies for the Grey Cup, and wins the Grey Cup championship.
Three easy steps. Massey optimization model. Logistic regression. Spreadsheet simulation.
…. and definitely the components of the Massey matrix (computation) and Massey ratings (output).
Although there won’t be any easy games it is certain that the computer doesn’t take into consideration that no team has ever played our starting lineup. Unfortunately that will still be the case this weekend but we are closer than ever.
Hmmmm, what, no Blue Bombers. Oh well, it was a good run while it lasted. Now, I will be cheering for a Ticats (its been a long time coming for them) Riders Grey Cup, and then, may the best team win on Grey Cup Sunday win.