Statistical Skier

I started writing an answer to a comment to Monday’s post, but it started getting long, so I thought I’d make it a short post, instead.

Basically, the question was how can the points be so different between OPA and Scandinavian Cup races when the penalty calculation methods is exactly the same. Recall that these were the estimated point differences between the FIS points of top ten finishers at OPA and Scandinavian Cups and their FIS points in WCs.

OPA Cups:

Scandinavian Cups:

So, the penalty calculation method doesn’t really have anything to do with this. Suppose that someone races in an OPA Cup and receives 20 FIS points. Supposedly, this should mean that they ought to get about the same result in a WC race. But in reality what we see is that they tend to do quite a bit worse than this. At least, as measured by FIS points.

Conversely, if someone gets 20 FIS points in a Scandinavian Cup race, it’s much more likely that this is close to how they’ll perform in a WC race. Again, close as measured by FIS points.

So what’s happening here is that the penalties at OPA Cups are likely artificially low, in that they are overestimating the field strength. Notice that the estimates for many of the Scandinavian Cup disciplines were actually negative, indicating that FIS points at these races are actually underestimating people’s performance in a WC race. How could this be? Well, remember that the penalty at these races can’t be any lower than 15, but Scandinavian Cups often have several Norwegian or Swedish national team members present, perhaps working their way back from an injury or something. This means that the penalty of 15 is likely too high.