I took the results from past Sunday’s Women’s Pursuit, 15 km, and did a bit of analysis. (The official results can be found here)

Since pursuit means that each racer starts the current race with a delay based on performance in the previous race, it’s impossible to see, from the official results, who ran fastest, i.e. did the current race in shortest time. So, I subtracted the start delay time for each individual competitor from their official results time, and it’s pretty interesting to see that the “impossible” Therese Johaug, was not fastest, in fact, she was only third fastest in this race.

The best time in this race had Ana Maria Lampic the Slovenian sprint world cup leader, followed by Frida Karlsson, 3.1 sec behind Lampic. Therese Johaug was an additional 1.2 sec behind Frida.

Below a graph showing the number of seconds each competitor was behind Lampic:

Let’s zoom in on the Swedish skiers:

Next, let’s look at the trend over the 4 sections of the course, start-1.6 km, 1.6 km-4.9 km, 4.9 km to 8.2 km, and 8.2 km to finish at 10 km, for the top-10 fastest skiers: in the graph below, each colored bar shows the racer’s performance, expressed in seconds-behind-the-fastest-racer of that section, with the blue bars corresponding to section start-1.6 km etc. In most cases, each competitor has 4 bars, one for each section of the track, but in those cases where there are only 3 bars, the implication is that that racer was fastest on that particular sector of the course:

Another perspective on the same is given by the lineplot below, which makes it perhaps easier to see the trend, and where on the track, in which sector, each racer performed better or worse: for instance, here it’s easy to see that Frida Karlsson had burned up most of her energy when reaching the final 8.2 km sector – from having been only a few seconds from the best sector time in sector 3, she lost almost 17 seconds against the best time of the last sector.

Finally, the same trend graph, focused on the Swedish participants, but this time, instead of showing seconds-behind-the-fastest, we’ll show race positions per course sector: Here we can see that the 10+ seconds Frida lost in the last sector against the fastest racer of that sector, corresponds to dropping from 5th to 48th place in the race.

Finally, a couple of graphs,possibly meaningful only to stat’s nerds 🙂

This one shows the race time distribution by nation. Interesting to see that the swedish team on average performed better than the norwegians 🙂

Next graph shows linear regression with race time over start delay time, meaning that those under the line performed better than given by their starting time, those above performed worse.