I tell you naught for your comfort…

“Never tell me the odds”, Han Solo said. Must’ve been a hockey fan.


Leading hockey stat-wrangler Gabe Desjardins offers a sobering statistical treatment of Canada’s chances for hockey gold. It’s sobering because it’s not at all based on a pessimistic view of the preliminary round—which, indeed, went some way toward vindicating Steve Yzerman’s roster selection and had the useful effect of clearing away the irrelevant detritus of past trophies from the goaltending question. (When was the last time you saw a Canadian team dominate a U.S. one from goal line to goal line so decisively?) Even on the explicit, historically derived premise that Canada has the strongest team in the tournament, it would be hard to peg our chances of winning gold at much higher than 25%. On Desjardins’ pretty reasonable estimates of underlying national team strength, the figure is not close to 25%. I crunched the numbers, leaving room for the possibility of being helped somewhere along the way by an upset of a strong rival, and I get about 19%. That’s assuming we have a 100% chance of beating Germany tonight, when the real figure is probably more like 93-95%.

Even Canada supporters who don’t quantify this stuff instinctively have probably already grimaced at the terrible logic. What are the most generous possible probabilities you could assign to Canada beating Russia and beating Sweden? Even if it’s 70% for both, that leaves us, basically, with only a (.7)² shot at merely making the final. In other words, about 50-50.

The historic estimate of our chances in a single game against Russia, even taking into account some weak pre-Ovechkin Russian teams, isn’t 70%; it’s 57%. Which feels approximately right. Home ice helps, but a lot of the home-ice advantage is really a “home time-zone advantage”, and the Rooskies have had a week to adjust to the new clock. And the chance of Slovakia bumping Sweden out of our way helps too, but then, Slovakia isn’t chopped liver. The sky grows darker yet, and the sea rises higher.


I tell you naught for your comfort…

  1. I agree the numbers seem about right… but I don't see the rational for all the doom and gloom. If Canada had zero chance of winning, my interest would be zero, however if Canada was a lock to win every year my interest would also be zero. I would say winning 1/3 of the time would be near ideal.

    Of course, the above is over the long haul… this year we HAVE to win!

  2. Like everything else, the key is to avoid looking at the whole endeavour at once. Instead think of it one game at a time.

    Starting after a win against Germany (I'm assuming we get that), you have a 57% chance of beating the Russians. That's better than 50-50. Play hard and we'll probably win.

    With the Russians behind us the chances of beating the Swedes are also better than 50-50. Again, play hard and we'll probably win.

    Then you get the gold medal game. Crapshoot. I call it 50-50 if goaltending is weak, or a near certainty if goaltending has improved by then.

    It's not so bad when you look at the odds game-by-game rather than aggregated. Any team going in has long odds of winning the gold; the advantage rests with the team that plays it one game at a time.

    • That seems like a reasonable analysis, but I think it's funny that you started by saying to take everything one game at a time, then totally discounted the next upcoming game :)

      • True 'nuff – should have said "starting with Germany – better than 50-50 odds – play hard and we'll win. Then…etc."

  3. That's assuming we have a 100% chance of beating Germany tonight, when the real figure is probably more like 93-95%.

    Hasn't Germany performed extremely poorly in the tournament thus far, losing every game and only managing to score against Belarus? I'd say the odds we'll beat Germany today are more like 99%.

    • I don't think that's realistic: the best minnows win against the Great Powers more often than that. Would you actually put $1,000 on Canada to win $10?

      • Nope. But I wouldn't put $1,000 on Canada to win $50-$75 either.

        I think you have a valid point about the best minnows winning against the Great Powers more often than one would expect, but in this case I think that one out of one hundred is more often than zero. It's analogous to the Canadian women's team playing the best non-US team.

        • If you wouldn't put $1,000 on Canada to win $75, that suggests you think the real figure is under 92.5%, no?

  4. All your math is making me feel funny. Shouldn't we just hope that we can get enough Team Chemistry to put up some Big Goals, and have some Timely Saves?

  5. Well, we had to beat these teams anyway, either in the semis or the final, so I don't see where the overall odds of winning have changed. The odds of getting no medal at all have increased but the odds of winning are the same. If we beat Russia and Sweden, we will then face the Czechs, the Slovaks, the Americans or the Finns, all of which must be better odds than Russia.

    In fact, the more think about this, I have no idea of where the heck you got these probabilities from , Mr Cash. Care to elaborate?.

    • I don't, Mr. Sampson.

  6. Nope. It just shows that I'm not a rational gambler.

  7. That's fair! But statements about odds (such as "I'd say the odds we'll beat Germany today are more like 99%") ARE claims to knowledge of appropriate betting behaviour. The question is what you WOULD do as a rational gambler.

    • That's why I'd make a terrible derivatives trader. I'd never put $1,000 at risk for such a small payoff, regardless how minimal the risk seems, because I'm way too superstitious about black swans.

      Enough about me, though. I suppose that as a rational gambler, I'd be prepared to risk $1000 for a $30 payoff. So I'll change my guess from 99% to 97%. It's hard to quantify these things! I just felt, intuitively, that 93-95% was too low. ;-)

      • "I'd never put $1,000 at risk for such a small payoff, regardless of how minimal the risk seems ….. "

        That's exactly what I was thinking as I read yours and cosh's conversation. I think Canada to beat Germany is more like 99.99999999% but we don't know when the one time they lose will happen.

        I was thinking $100 – seventy five seems stingy but one hundred would be something to at least think about, for me at least.

  8. Yeah, I really just want to watch some good hockey tonight. Preferably a good whooping, but I’ll take a 1-0 win all the same.

  9. I'd say our chances of beating Russia are 60%, Sweden 80%, and the final 80%. Overall chance is therefore 38%.

    Putting Luongo in nets will make a big difference.

  10. One relevant statistical reference absent from your piece is that the chances of any other team winning the gold aren't much better. I agree that Canada's statistical chances of winning gold are probably no better than 20 – 25%, but no other teams chances of winning gold are much better, if at all. I prefer the perspective that some team will eventually win the gold and our team's chance is at least as good as any other's.

    • Odds of making the final now 100%. Odds of winning, 50-50. Play hard and we'll win.

      • And hence the win.

        Never tell me the (aggregate) odds.

  11. Interesting difference between the punter and the coach. The coach wants to change to problem from a cumulative probability problem to a quantum problem.

    It's not that we have a 90% chance of winning it's that the other guys have only .9 of our team and a 100% effort each game produces a 100% chance of winning the gold.

  12. Looks like you are WRONG!!!!! Biatch!! ( You are a bitter Oil fan like me- we traded Gretzky, and out now forever bitter and pessimistic)

    Although Slovakia looks like a giant killer (Hossa and Gaborik with a giant named Chara – this man is wrong!!! GOILERS!! GO CANADA!!!