How can we assess performances in stage race pursuits?
As with most stage races these days, the Tour de Ski included some handicap start pursuits. Generally, the athletes will start in a staggered fashion based upon their time back in the overall standings, and the first person to the finish line wins the stage.
This creates some very unusual incentives for the athletes, depending on where they are in the overall standings. Many of the athletes are no longer really racing against each other, only those skiers who happen to be near them in the overall standings. For example, there’s no really sense in which Therese Johaug was racing against Liz Stephen in Thursday’s pursuit. Johaug only really cared about maintaining the gap behind her and trying to catch the leaders.
This means that isolating the times for just that stage is an almost useless way to gauge performance. Johaug, for instance, skied that stage around 15 seconds slower than Stephen, but we probably don’t think that that means much about Stephens’ ability compared to Johaug’s.
So I wonder about these things when I read stuff like this, or this, talking about how well Stephen and Freeman skied in the pursuit based on their times for just that stage. By my calculations, Freeman had the 11th fastest time of the day and Stephen the 13th. But that alone doesn’t mean much to me, since many other skiers were really only racing against the skiers near them, rather than the whole field.
If an athlete says they had a good day, of course, I’m inclined to believe them. They alone know how their body felt and whether it was a good effort. But, you know, I like to measure stuff, so let’s try.
Let’s compare Freeman and Stephen only to those skiers who started near them in the pursuit. Specifically, how did Freeman and Stephen’s times for the day compare to their historical performances against these skiers? Better than average? Worse? About the same?
Here’s the relevant graph for Freeman: Read more
Davos Recap
North Americans
Each weekend seems to be an interesting mix of results for the North Americans. Starting with the women (Davos result circled in blue):
Compared to this season, that was an off day for Kikkan, but compared to last season that was pretty typical. Part of me wonders if she dialed it back a bit late in the race when she knew she wasn’t feeling strong to save some energy for the sprint. But as you can see from her graph, if that’s going to be her “bad” race this season, she’s going have a strong set of results this year.
I’m very cautious about jumping on bandwagons when someone pops a good race or two, but Holly Brooks is beginning to convince me. That’s three good (and one OK) distance results in a row now. More importantly, I like the direction her trend is heading. It’s still early, so it’ll only take a few mediocre races to flatten that trend out, but so far it looks promising.
You can’t deny that Liz Stephen has had some strong results so far this season. My only concern is that they have all been roughly where we’ve seen her topping out before. Can her good days inch up towards the top ten?
As for the men: Read more
Missed Medals
Missed Medals
A reader asked on Twitter the other day if I’d recalculate the tallies for Kris Freeman’s top ten, top three, etc. results if we went back and removed a certain Estonian skier from prior results sheets.
I’m going to preface this with some standard caveats about how I can’t know for sure when Andrus Veerpalu was clean or not clean. He certainly failed a drug test recently, and the way public opinion works with doping is that basically all his results are now suspect. I’m agnostic on that question, but since this was a fairly interesting data exercise, I thought I indulge the haters out there.
But I’m going to take it one step further and tally the missed opportunities for all skiers, not just Freeman, if we remove Veerpalu. We’ll start with the least important consequences and work our way up.
There were 78 different skiers who would have moved up into the World Cup points (from 31st to 30th) if you removed Veerpalu. There wasn’t really one skier who bore the brunt of this; the most it happened to any one skier was 3 times. North Americans in this group include Kris Freeman, Carl Swenson, George Grey, and Dan Roycroft, all once.
Next, removing Veerpalu will move some skiers up from 11th to 10th. This was also fairly evenly spread out among 43 different skiers, including Carl Swenson again.
More dramatically, removing Veerpalu from the results will bump 13 different skiers up from 4th to 3rd. Here is where we finally include Freeman’s famous WSC “medal”. Also notable is that Mathias Fredricksson would gain three more podium finishes in WC races due to this change.
Finally, the big kahuna, and I was a little shocked at the results. There were five skiers who would gain a victory out of this transaction. Two Olympic gold medals and two WSC gold medals. When Veerpalu’s ten wins get apportioned out, one skier gets five of them: Frode Estil. And it kind of makes sense, since he was another classic ultra-specialist.
I didn’t bother removing any of the Finns or Russians (although a lot of the Russians who’ve been caught haven’t been nearly as accomplished). Perhaps I’ll tackle that in a follow up. Read more
USST 2011-2012 Preview
Another season is just around the corner! While the USST saw some changes over the off season, the folks we are most likely to see top results from are still the Big Three: Kikkan Randall, Andy Newell and Kris Freeman.
My expectations and questions for Freeman are summed up in this graph:
This shows his median (and 10th/90th percentile – errorbars) WC results by month, using standardized percent behind the median skier. Freeman has tended to see a big dropoff in January and has had inconsistent results mid-season. The differences between the months may not seem huge, since the y axis units are not something you’re used to thinking in. But keep in mind that a difference of 0.5 can easily mean the difference between 10th and 30th (or more).
So I won’t be surprised if Freeman comes out of the gate skiing fast in November, but I’ll be very curious what happens when we hit January.
As for Kikkan Randall, I think there’s probably a lot of excitement surrounding her, considering the breakthrough season she had last year. I’ve spoken before on this blog about her chances for contending for the WC sprint title, and I think she certainly will be a podium contender in every freestyle sprint she enters. But here’s a minor note of caution: Read more
New Zealand Continental Cups Sprint
Continuing on in our over-analysis of the recent New Zealand FIS races, we turn to the sprints. It’s much harder to do anything sensible with these races (even given that I’ve over-analyzing things!) since the fields are so small even the difference in placing between specific skiers is potentially misleading. However, just for fun let’s focus like last time on a younger sprinter, Len Valjas:
Again, this is the difference in finishing place between Valjas and a selection of folks from the New Zealand sprint race. Positive values mean Valjas won and vice versa. Note that he finished a lot closer to some of these Russians that he normally does, but again, that may be a function of the small field. It’s hard to do much analysis on these sorts of races without the heat times.
My only other note is that Kris Freeman slightly underplayed how good his sprint FIS points were from this race. Here’s a graph of all the sprint qualification results I have for him with the New Zealand race in red:
Definitely good sprint FIS points for him, but it’s more like the 5-6th best all time. Of course, a lot of those are from a long time ago in potentially very different sprint race formats, distances and courses.
US World Cup Split Times
Visualizing split times is trickier than it seems. You have data from an array of different situations that arguably aren’t terribly comparable. We all know that mass start and pursuit races play out very differently than interval start races. It may also be difficult to compare split times from a Tour de Ski prologue to a 50k. Add to all this the fact that each race will record their splits at different points in the race, even for races of the same length, and you’ve got quite a jumble of information. (Once again, I’m using split time data kindly supplied by Jan at WorldOfXC.com.)
Still, we can do some stuff with this: Read more
How’d We Do? USA/CAN Season Review 1
I think most people generally have a sense for how the past World Cup season went for the North Americans. What I’m going to do over the next few posts is to simply show some data that hopefully provides some context for what you already know. I’m going to split them into four posts for men/women and distance/sprint. Today we’ll start with the men’s distance performances.
Let’s start with the simplest of metrics, finishing place, and a style of graph that I’ve used before that shows the number of results per race at a given level:
Recent Comments