CFL 2014 Ratings/Playoff Positions at 2014 11 08

[Qman]

I think likely scenerio is whether we catch SSK for 3rd. I don't think the riders will win another game this season with their current QBing.

[B.C.FAN]

Perhaps I'm not understanding how a team makes the playoffs here, but how can the Riders be shown to have clinched a playoff spot? If they lose their last two games, they'd end up at 9-9, and if the Lions win 2 of their last 3, this would put the Riders in 4th place in the West. If the Argos win their last 3, both the Argos and Riders would have identical records, and, if I'm understanding the crossover rules correctly, Toronto would get the nod because for a team to cross over it must have a better record than the third place team in the other division. Does it have something to do with the fact that not all of the three top Eastern teams can reach 9-9?

Someone please explain....

Because the Eastern teams play each other, it's impossible for three teams to reach 9-9. If Toronto wins its last three games, either Montreal or Hamilton can do no better than 8-10.

[DanoT]

Perhaps I'm not understanding how a team makes the playoffs here, but how can the Riders be shown to have clinched a playoff spot? If they lose their last two games, they'd end up at 9-9, and if the Lions win 2 of their last 3, this would put the Riders in 4th place in the West. If the Argos win their last 3, both the Argos and Riders would have identical records, and, if I'm understanding the crossover rules correctly, Toronto would get the nod because for a team to cross over it must have a better record than the third place team in the other division. Does it have something to do with the fact that not all of the three top Eastern teams can reach 9-9?

Someone please explain....

You have answered your own question. It is due to the fact that the eastern teams play each other, one of them will not get 9 wins.

[South Pender]

You have answered your own question. It is due to the fact that the eastern teams play each other, one of them will not get 9 wins.

Ahhh....Got it! Thanks, DanoT.

[sj-roc]

You have answered your own question. It is due to the fact that the eastern teams play each other, one of them will not get 9 wins.

Ahhh....Got it! Thanks, DanoT.

Yes, if Toronto goes to 9-9 they'll have by necessity leapfrogged over somebody else who drops to third and that's who would get bumped instead. So the crossover is on for sure as soon as BC gets win #9 (then all that's left to decide is whether the crossover team is BC or Ssk). Mind you, should the eastern team that ends up 3rd not reach a best-case eight wins to match the Lions' current win total then they'd already be out anyway no matter else would have happened here in the west.

[notahomer]

http://www.cfl.ca/article/2014-cfl-play ... narios2014

Lots of updates on the scenarios of the cfl playoff standings page. Looks like the BC/Bomber game will be the only one to have DIRECT impact...

IOW, BC WIN equals BC clinching playoff spot OR BC LOSS equals Eskimos clinching 2nd place.

Not sure why a Bomber loss would not eliminate them......

[sj-roc]

http://www.cfl.ca/article/2014-cfl-play ... narios2014

Lots of updates on the scenarios of the cfl playoff standings page. Looks like the BC/Bomber game will be the only one to have DIRECT impact...

IOW, BC WIN equals BC clinching playoff spot OR BC LOSS equals Eskimos clinching 2nd place.

Not sure why a Bomber loss would not eliminate them......

Well, it says "BC Lions win = BC clinches playoff berth" and this would make them the 4th western team to clinch. Which leaves Wpg no way out of the basement.

Also, I couldn't find anything about this in there, but... a Ssk win @Cgy on Oct 24 eliminates us from 2nd place.

I just realised something: Ott is the only team that will have gone winless against us this year, and even they might have done it if it weren't for that sudden storm in our game there.

There's also a simulation page that determines each team's chances of winning its division. A bit moot for us now since Cgy has already clinched but there's still numbers given for the east and I'd also like to see the simulation run to determine who would finish 2nd in the west.

They've also calculated probabilities of each team appearing in, and also winning, the Grey Cup. Cgy currently leads both categories at 76.45% and 63.15% respectively. Lions numbers are 10.51% and 3.16%. I suppose with 3.16% < half of 10.51%, this means we would likely go in as underdog if we do actually get in. I wonder if our mediocre home record factored into this?

Simulation page: http://www.cfl.ca/article/cfl-simulation-2014

[notahomer]

................Simulation page: http://www.cfl.ca/article/cfl-simulation-2014

Yes, I've been following that simulation page for a few weeks now. Not too closely though because there is a reason why the games have to be played. I get that analytics etc... can help but its still the players/coaches on the teams that impact who wins/losses, IMO.

[sj-roc]

................Simulation page: http://www.cfl.ca/article/cfl-simulation-2014

Yes, I've been following that simulation page for a few weeks now. Not too closely though because there is a reason why the games have to be played. I get that analytics etc... can help but its still the players/coaches on the teams that impact who wins/losses, IMO.

My impression is that they're using input data similar to what "Power Rankings" pundits use to generate their results, and applying it to project the rest of the season. So to whatever extent you believe PRs probably should also apply to this sim.

[WestCoastJoe]

Win/Loss records say a lot. Games left.

1. Calgary. Still the class of the league. No guarantees in the playoffs. Brand new season. H Peg then @ B.C.

2. Edmonton. Tough team under Jones. A good challenge for us. Host B.C., then @ Sas.

3. B.C. Game in hand over Saskatchewan. We go to Edm and we host Cal. We could lose both. We could win both. Or in between. ???
........

Now it gets interesting in the East. Of Mon, Tor and Ham, 2 make the playoffs. Let the games decide.

4. Montreal. I am happy for Tom Higgins / Ned Flanders. Good guy. Good credentials. H Tor. @ Ham. Could have 8, 9 or 10 wins.

5. Toronto. @ Mon. H Ott.

6. Hamilton. @ Ott. H Mon.

7. Saskatchewan. In the playoffs. One of the Eastern teams gets knocked out, just because they cannot all make it to 9 wins.

8. Winnipeg. I expect O'Shea will have them competitive next year.

9. Ottawa. Dunno about Rick Campbell yet. Expansion team, of course. Decent talent though.

[notahomer]

http://cfl.ca/article/stephen-rest-vs-r ... in-calgary

This so called RUST versus REST debate USED to be an annual dialogue here on Lionbackers when the Lions were the ones put in the position of wondering whether to rest/play people. Oh well.... IMO, its obvious Calgary knows the players need to keep the routine going. Getting some action, keeping those wheels greased is important but at the same time this is a good time to get those other guys some snaps AND DO NOT LOSE YOUR D-LINE on the last game of the year, like the Stamps did versus the LIons in 2013. Unfortunately, it was the Riders who benefited most from that problem.....
----------------------------
Even the Lions may have some of this debate this season. You have to go all out this week considering a home playoff game is still in the balance. And if they win this week, they want to win over the Stamps too. But once (or IF?) second place is a foregone conclusion (regardless of details), obviously you'd want some selective use of players. Will be interesting to see how the final weeks play out.....

_________________________________________________
And the simluation has me a little puzzled.....
http://www.cfl.ca/article/cfl-simulation-2014

Thankfully the simulation MEANS nothing. The games themselves HAVE TO BE PLAYED. According to this BC Lions have an 8.61% of making it to the Grey Cup and if they DO a 2.9% chance of winning it. Now I realize these numbers ARE BASED ON TODAY (trust me last weeks were worse). That being said, IF the Lions make the Grey CUp, as the host city, I think the Lions chances are a little better than 2.9% of winning it. I guess part of the problem of making these calculations is still simply where are the Lions playing. As a crossover team or the two possiblities (second/third) in the west.

[sj-roc]

And the simluation has me a little puzzled.....
http://www.cfl.ca/article/cfl-simulation-2014

Thankfully the simulation MEANS nothing. The games themselves HAVE TO BE PLAYED. According to this BC Lions have an 8.61% of making it to the Grey Cup and if they DO a 2.9% chance of winning it. Now I realize these numbers ARE BASED ON TODAY (trust me last weeks were worse). That being said, IF the Lions make the Grey CUp, as the host city, I think the Lions chances are a little better than 2.9% of winning it. I guess part of the problem of making these calculations is still simply where are the Lions playing. As a crossover team or the two possiblities (second/third) in the west.

Yeah, it's basically like a mutual fund prospectus. They tell you stuff based on past performance but there's always that caveat that "past performance may not be repeated". They ran the simulation 10,000 times on the remaining schedule and then allocated favourable outcomes to each team so I think there is some statistical validity to their work, to the extent that their input data is valid. This may or may not be the case depending on, for example, whether they take into account sudden, long-term injuries. It's still an interesting exercise, though and offers some discussion fodder. In this sense it's not unlike a mock draft drawn up in the spring.

I think the 2.9% takes into account *all games* that have yet to play out from now until the trophy is hoisted and allows for the fact that we may not even make it into the game. It doesn't mean, for instance that the other team would have a 97.1% chance of beating us if we made it; more like, it means there's a 97.1% chance at this moment that someone else besides us will end up winning it, whether they face us or some other team. It should be noted that the appearance probabilities add to 200% and those of winning add to 100% as they should under this interpretation.

Should we indeed make it, they could rerun the simulation with all the new water that's flown under the bridge between now and Grey Cup eve and it would have to be a heck of a lot more than 2.9%.

If you wanted a team's chances of winning the Grey Cup were they to play the GC game today, I think you would get that by dividing overall chance of winning (2.9% for BC) by overall chance of appearing (8.61%), which in BC's case would be 2.9/8.61 = 33.7%.

It seems a little unusual that all three eastern contenders, none of whom have yet clinched a berth and one of whom definitely will NOT, are all ranked higher than us on their current championship prospects, but this simply reflects the fact that all three teams are very much mathematically alive for the two remaining berths. I don't imagine all of them will stay ahead of us much longer.

I'd like to see them disclose more results from their sim — numbers that I'm sure they could dig out if they wanted to, like what are the probabilities of each team finishing in second, etc. A while back I pegged the Lions' chances for 2nd place at around 5% but this was before a couple of needed results came to pass, so I'd say it's a little higher now, but still no more than 10%. If each of three remaining games that will determine our playoff seed could be viewed as a random coin flip then our chances would be (1/2) x (1/2) x (1/2) or 12.5%. But we need to win by 7+ against Edm and this would probably be under a 50% chance all else being equal, and the Edm@Ssk game prob has under 50% chance of going our way (ie Ssk win) given the Riders' play of late. So a 10% chance of us finishing 2nd would be the highest I would peg it right now; even this might be a little generous.

Something else they could have done with their sim is to run it with last season's data — with the hindsight of already knowing how it played out, but only using the results known after each week of action and then seeing how the numbers evolve week-to-week. Perhaps they've indeed done this, or perhaps they used last season's data (and possibly even multiple previous years) as a guide in building their model to analyse *this* year's data at various points of the season. I haven't seen any discussion of this anywhere.

[South Pender]

If you wanted a team's chances of winning the Grey Cup were they to play the GC game today, I think you would get that by dividing overall chance of winning (2.9% for BC) by overall chance of appearing (8.61%), which in BC's case would be 2.9/8.61 = 33.7%.

This should be right by Bayes's Theorem. You're looking for a conditional probability--the conditional probability that the Lions will win the GC given that they get to it--and that would be 33.7%, given the probabilities provided.

I haven't looked at the simulation and don't know whether the probabilities given (that form part of the calculation of a conditional probability) are reasonable or not, so can't say whether the result is credible. Of course, a lot can happen between now and the playoffs. The Lions could catch fire and be the stronger team if they get to the GC.

A while back I pegged the Lions' chances for 2nd place at around 5% but this was before a couple of needed results came to pass, so I'd say it's a little higher now, but still no more than 10%. If each of three remaining games that will determine our playoff seed could be viewed as a random coin flip then our chances would be (1/2) x (1/2) x (1/2) or 12.5%. But we need to win by 7+ against Edm and this would probably be under a 50% chance all else being equal, and the Edm@Ssk game prob has under 50% chance of going our way (ie Ssk win) given the Riders' play of late. So a 10% chance of us finishing 2nd would be the highest I would peg it right now; even this might be a little generous.

I agree. I'd put the Lions' chances of winning in Edmonton at about .40, along with the same for Edmonton losing to Saskatchewan. So my estimate of the probability that Lions will end up in 2nd place might be a little lower--at about 8%. But you have indicated that 10% is an upper bound, and I agree.

And, of course, nota is right. This all goes right out the window once the games are played!

[B.C.FAN]

If you wanted a team's chances of winning the Grey Cup were they to play the GC game today, I think you would get that by dividing overall chance of winning (2.9% for BC) by overall chance of appearing (8.61%), which in BC's case would be 2.9/8.61 = 33.7%.

This should be right by Bayes's Theorem. You're looking for a conditional probability--the conditional probability that the Lions will win the GC given that they get to it--and that would be 33.7%, given the probabilities provided.

I haven't looked at the simulation and don't know whether the probabilities given (that form part of the calculation of a conditional probability) are reasonable or not, so can't say whether the result is credible. Of course, a lot can happen between now and the playoffs. The Lions could catch fire and be the stronger team if they get to the GC.

A while back I pegged the Lions' chances for 2nd place at around 5% but this was before a couple of needed results came to pass, so I'd say it's a little higher now, but still no more than 10%. If each of three remaining games that will determine our playoff seed could be viewed as a random coin flip then our chances would be (1/2) x (1/2) x (1/2) or 12.5%. But we need to win by 7+ against Edm and this would probably be under a 50% chance all else being equal, and the Edm@Ssk game prob has under 50% chance of going our way (ie Ssk win) given the Riders' play of late. So a 10% chance of us finishing 2nd would be the highest I would peg it right now; even this might be a little generous.

I agree. I'd put the Lions' chances of winning in Edmonton at about .40, along with the same for Edmonton losing to Saskatchewan. So my estimate of the probability that Lions will end up in 2nd place might be a little lower--at about 8%. But you have indicated that 10% is an upper bound, and I agree.

And, of course, nota is right. This all goes right out the window once the games are played!

If you want to put money on the Grey Cup, here are this week's odds from Caesars casinos in Nevada (with the season-opening line in brackets):

Calgary: Even (9/2)
Edmonton: 4/1 (9/1)
Hamilton: 7/1 (7/1)
Montreal: 7/1 (8/1)
B.C.: 8/1 (3/1)
Toronto: 8/1 (7/1)
Saskatchewan 10/1 (4/1)

[sj-roc]

If you wanted a team's chances of winning the Grey Cup were they to play the GC game today, I think you would get that by dividing overall chance of winning (2.9% for BC) by overall chance of appearing (8.61%), which in BC's case would be 2.9/8.61 = 33.7%.

This should be right by Bayes's Theorem. You're looking for a conditional probability--the conditional probability that the Lions will win the GC given that they get to it--and that would be 33.7%, given the probabilities provided.

I haven't looked at the simulation and don't know whether the probabilities given (that form part of the calculation of a conditional probability) are reasonable or not, so can't say whether the result is credible. Of course, a lot can happen between now and the playoffs. The Lions could catch fire and be the stronger team if they get to the GC.

This isn't meant to be inconsistent with or to contradict your comments, but I suspect the 2.9% and 8.61% results they came up with are not independent. Rather, I suspect they first arrived at 8.61% for our chances of winning a divisional final and then a 33.7% chance of us winning a Grey Cup should we advance to play in it (as a separate matter decoupled from the other one of getting there), and then did basically the same final step of arithmetic I did above (only in reverse of course!) to arrive at a current 2.9% chance of winning it all.