January 25, 2006

Amateur Medal Prediction for Torino: 21

I'm getting a little bit of web traffic from people looking for 2006 winter Olympic predictions, so I decided I should do some. (I also get a little bit of web traffic looking for nude photos of Romanian gymnasts, but that one I've decided is a little bit outside my scope here.) Now, these may not be the kind of predictions that people are looking for, but I don't really follow all the winter sports closely enough to make event-by-event forecasts.

I am interested, though, in Canada's Own the Podium program and its implementation. The Own the Podium program has a goal of finishing first in the medal table, with 35 medals, at the 2010 winter Olympics in Vancouver. The interim goal for 2006 is to win 25 medals, and finish third in the medal table.

I should say up front that what follows is not meant to be a criticism of the Own the Podium program in any way. There is plenty of evidence that the Canadian winter sports teams are doing better than they have ever done before, and moving up compared to the rest of the world. That's the real goal of the targeted funding, and I am convinced that it is working.

Now, having said that, here are my two fearless predictions for 2006. First, Canada will not win 25 medals in Turin. And second, 25 medals would not be enough for third place in the medal table, anyway.

I don't really have any solid basis for the second prediction; I just think that at least three countries are going to win more than 25 medals this time around. But I did run some numbers on the first hypothesis. Here's the methodology. For each of the 84 Olympic events, I've made an estimate of the probability that Canada's 'A' entry will win a medal. For events where we have more than one entry, I've also guessed at the probability that the 'B' entry and (if applicable) the 'C' entry will win medals.

These estimates are what I would call "wild-assed guesses" or WAGs. That is, I have not put very much thought into any of the estimates, and I have not tried to be very precise. I've allowed each probability to be one of 0, 0.10, 0.25, 0.50, 0.75, 0.90, or 1.0, and I haven't done any detailed research into the current World Cup rankings, or done any best time analysis. I've given one entry a 100% (1.0) probability of winning a medal, which is to say that I think their chances are closer to 100% than they are to 90%; and I've given a great number of athletes no chance at all, which is to say that I think their chances are closer to zero than they are to 10%. As I said at the outset, I am not an expert on any winter Olympic sport, so I am sure to have made some mistakes. My heroic assumption is that although the individual guesses may be (quite badly) wrong, I am making a lot of guesses, and those individual errors will tend to cancel each other out.

Once I have the probabilities for each event, I can calculate the total probability that Canada will win zero medals in the event, one medal in the event, two medals in the event, or three medals in the event. These are calculated by considering all the possible combinations of outcomes for the A, B, and C entries.

These four probability scores for each event can be used to give an expected or most likely number of medals that Canada will win any single event. Table 1 below shows my estimates for every event, and the expected medal score for each event. The table also shows the expected Canadian medal total for each winter Olympic sport, and the expected Canadian medal total for all sports.

Table 1 — Medal win probabilities for Canadians at the 2006 winter Olympics
Sport Event PA PB PC Expected Medals Athletes
Speedskating - Long Track Men's 5000 m 0% 0% 0% 0.0
Women's 3000 m 75% 25% 10% 1.1 Hughes, Groves, Klassen
Men's 500 m 50% 0% 0% 0.5 Wotherspoon, Ireland
Women's 500 m 10% 0% 0% 0.1 Rempel
Women's Team Pursuit 90% 0% 0% 0.9
Men's Team Pursuit 10% 0% 0% 0.1
Men's 1000 m 25% 10% 0% 0.4 Wotherspoon, Morrison
Women's 1000 m 10% 10% 0% 0.2 Klassen, Rempel
Men's 1500 m 10% 0% 0% 0.1 Morrison
Women's 1500 m 90% 10% 0% 1.0 Klassen, Groves
Men's 10000 m 0% 0% 0% 0.0 Dankers
Women's 5000 m 50% 10% 0% 0.6 Hughes, Groves
TOTAL 5.0
Speedskating - Short Track Men's 1500 m 10% 10% 0% 0.2 Hamelin, Turcotte
Women's 500 m 50% 25% 10% 0.9 Roberge, Leblanc-Boucher, Kraus
Women's 1500 m 10% 0% 0% 0.1 Leblanc-Boucher, Overland
Men's 1000 m 25% 0% 0% 0.3 Bedard, Tremblay
Women's 3000 m Relay 90% 0% 0% 0.9
Men's 500 m 50% 25% 0% 0.8 Bedard, Tremblay
Women's 1000 m 25% 10% 0% 0.4 Overland, Vicent
Men's 5000 m Relay 75% 0% 0% 0.8  
TOTAL 4.2
Hockey Men's 75% 0% 0% 0.8  
Women's 100% 0% 0% 1.0  
TOTAL 1.8
Bobsleigh Men's Two-Man 50% 0% 0% 0.5 Leuders
Men's Four-Man 50% 0% 0% 0.5 Leuders
Women's 75% 0% 0% 0.8 Upperton
TOTAL 1.8
Freestyle Skiing Men's Moguls 10% 0% 0% 0.1 Bilodeau
Women's Moguls 90% 10% 0% 1.0 Heil, Richards
Men's Aerials 25% 10% 0% 0.4 Nissen, Bean
Women's Aerials 10% 0% 0% 0.1 Bauer
TOTAL 1.6
Curling Men's 75% 0% 0% 0.8 Gushue rink
Women's 75% 0% 0% 0.8 Kleibrink rink
TOTAL 1.5
Skeleton Men's 75% 10% 0% 0.9 Pain, Boehm
Women's 50% 0% 0% 0.5 Hollingsworth-Richards
TOTAL 1.4
Figure Skating Women's 10% 0% 0% 0.1  
Men's 75% 10% 0% 0.9 Buttle, Sandhu
Pairs 0% 0% 0% 0.0
Ice Dancing 25% 0% 0% 0.3 Dubreuil, Lauzon
TOTAL 1.2
Cross Country Ski Women's 15 km Pursuit 25% 10% 0% 0.4 Scott, Renner
Men's 30 km Pursuit 0% 0% 0% 0.0
Women's Team Sprint 25% 0% 0% 0.3 Scott, Renner
Men's Team Sprint 0% 0% 0% 0.0
Women's 10 km Classical 0% 0% 0% 0.0
Men's 15 km Classical 0% 0% 0% 0.0
Women's 4x5 km Relay 0% 0% 0% 0.0
Men's 4x10 km Relay 0% 0% 0% 0.0
Women's Sprint 50% 10% 0% 0.6 Scott, Renner
Men's Sprint 0% 0% 0% 0.0  
Women's 30 km Free 0% 0% 0% 0.0  
Men's 50 km Free 0% 0% 0% 0.0  
TOTAL 1.2
Alpine Ski Men's Combined 0% 0% 0% 0.0 Bourque
Men's Downhill 25% 0% 0% 0.3 Guay
Men's Super G 25% 0% 0% 0.3 Guay
Men's Giant Slalom 10% 0% 0% 0.1 Grandi
Men's Slalom 10% 0% 0% 0.1 Grandi
Women's Combined 0% 0% 0% 0.0 Brydon, Forsythe
Women's Downhill 10% 0% 0% 0.1 Brydon, Vanderbeek
Women's Super G 10% 0% 0% 0.1 Brydon, Simard
Women's Giant Slalom 0% 0% 0% 0.0
Women's Slalom 0% 0% 0% 0.0
TOTAL 0.9
Snowboard Men's Halfpipe 0% 0% 0% 0.0  
Men's Snowboard Cross 50% 0% 0% 0.5 Anderson
Men's Parallel Giant Slalom 10% 0% 0% 0.1
Women's Halfpipe 0% 0% 0% 0.0
Women's Snowboard Cross 10% 0% 0% 0.1 Maltais
Women's Parallel Giant Slalom 10% 0% 0% 0.1 Loo
TOTAL 0.8
Luge Men's Singles 0% 0% 0% 0.0  
Women's Singles 0% 0% 0% 0.0
Doubles 0% 0% 0% 0.0  
TOTAL 0.0
Biathlon Men's 20 km Individual 0% 0% 0% 0.0  
Women's 15 km Individual 0% 0% 0% 0.0
Men's 10 km Sprint 0% 0% 0% 0.0
Women's 7.5 km Sprint 0% 0% 0% 0.0
Women's 10 km Pursuit 0% 0% 0% 0.0
Men's 12.5 km Pursuit 0% 0% 0% 0.0
Men's 4x7.5 km Relay 0% 0% 0% 0.0
Women's 4x6 km Relay 0% 0% 0% 0.0
Men's 15 km Mass Start 0% 0% 0% 0.0
Women's 12.5 km Mass Start 0% 0% 0% 0.0  
TOTAL 0.0
Ski Jumping NH Individual 0% 0% 0% 0.0  
LH Individual 0% 0% 0% 0.0
LH Team 0% 0% 0% 0.0  
TOTAL 0.0
Nordic Combined Individual 0% 0% 0% 0.0  
Team 0% 0% 0% 0.0
Sprint 0% 0% 0% 0.0  
TOTAL 0.0
ALL TOTAL 21.1

So according to my WAGs, Canada should expect to win 21.1 medals in Turin. I can say more than that, though. If we consider all the possible combinations of the probabilities in Table 1, the estimates can be combined globally to generate a probability distribution for Canada's total medal haul across all events. That probability distribution is shown in Figure 1 (inset).

Figure 1

Canadian medal prediction, 2006

Figure 1 — Probability distribution for Canada's 2006 winter Olympic medal total (click to enlarge).

There are a couple of things worth noting here. First of all, the probability distribution looks a lot like a normal distribution, even though I started with just a set of discrete probabilities. I think that's pretty cool, but then I'm geeky that way.

Second, the picture confirms that the most likely number of medals for Canada is 21; the one-sigma uncertainty on that estimate, again according to my collection of WAGs, is ±3. The probability of Canada winning 25 or more medals is about 13%.

As far as the reliability of the estimate goes, that depends entirely on whether or not my initial WAGs are biased. Again, it doesn't really matter if the guesses are poor, as long as I am not generally too optimistic or too pessimistic. My guess would be that I tend to be optimistic. I tend to hear more news about Canadian successes than I do about Canadian failures, so I probably have a tendency to overestimate the chances of the Canadian stars. On the other hand, I have probably underestimated a number of athletes simply because they haven't been in the news as much.

Getting back to the Own the Podium program, it appears to me that the interim goal of 25 medals in Turin is just a little bit unrealistic. Of course, there is a lot of motivational value in setting "stretch" goals for yourself, and athletes do it routinely. I know that the COC needed an inspiring goal just to get the extra funding in the first place, and I understand just how difficult that ultimate goal is going to be. If you want to convince people that your plan is working, however, there's nothing more effective than a goal that is actually reached.

No comments: