How close is "close"? When might this be a problem?

February 17, 2006. In the Olympic Men's competition, Buttle and Lambiel were separated by about 3.5 points. None of the 48,400 combinations of panels would have changed the medal order. Some were extremely close, though, with one particular set of judges placing Lambiel's margin of victory at 0.007 points.

With the benefit of hindsight, Lambiel's margin of victory over Buttle was "close enough" to justify the analysis, but not close enough to possibly make a difference in the standings in this particular competition.

Ice Dancing? Less than 0.25 points separated silver from bronze. But with three segments rather than two, I couldn't run through all 220^3 possible panels. However, the consensus of all 12 judges would not have changed the medal standings. If we did look at all combinations of randomly selected panels, I am sure than many would have awarded the medals differently. Once again, good luck at work!

Ladies? I'll need to go back and look at the 2006 European Championship results to have a better idea, but I think my answer will be unchanged. If the scores of two skaters are within 5 points, I'll do the analysis. This only happens once every four years, so let's get it right.

Back to Jay's home page