Identical Twins, Rowing, and the Luck of the Lane

Caroline and Georgina Evers-Swindell are identical twin sisters from New Zealand. But they are not identical only in the sense that they look alike; they are both very strong and very competitive athletes, and both excel at the sport of rowing.  Rather than compete against one another, however, they compete together in the sport known as Women’s Double Sculls. (To those of us less familiar with rowing terms, that’s where you have two people rowing the same boat in a race.)  Here they are, in action:​


Recently I’ve been writing on the fairness of judging in Olympic events that require subjective evaluation rather than a score, measurement, or stopwatch. But now I’d like to look at an event that seems simple—on the surface. In rowing, several boats all start at the same time on a long stretch of water and race to the finish in their respective lanes. And they race fast…even with only two rowers, they cover the nearly 1.25-mile course in just over seven minutes.

The event consists of three rounds—heats, quarterfinals, and finals—and lanes are randomly assigned for each race. Which got me thinking: does it matter which lane you are assigned?  So I analyzed the data using Lane, Rowers, and Round as factors and got the following results:


Lane is not only a significant effect, but the effect of which lane you are in is nearly as important as the effect of which rower you are!  This seemed unbelievable to me and I knew there must be an explanation. I removed the effect of rower from my model and made an Interaction Plot of the finishing time (in minutes) by Lane and Round:

Two things I pick up on right away:

  1. Speeds were obviously fastest in the quarterfinals and slowest in the finals.
  2. Regardless of round, speeds are fastest in Lane 1 and get steadily slower through Lane 5.

Looking for an explanation, I did my own gemba walk by searching for video replays of the races to see if there were any clues about what would account for this.  After a couple of views, I noticed that the water seems fairly choppy for each of the races, which would only occur in a lake in the presence of wind.  Then, presumably placed to allow me to discover this four years later, I noticed that the entire length of the viewing stands was lined with flags that, as flags tend to do, blew in the direction of the wind.

Not surprisingly, in all three races the wind blew from the Lane 6 side towards the Lane 1 side, and in the Heats and Finals came at an angle, providing headwind as well. So the lower the lane number, the more likely you are to have other boats blocking the wind for you—and paddling in even slightly smoother water may make all the difference between winning and losing!

So how did Caroline and Georgina do?  Here is a photo of the finish:

That’s them on the top, winning Gold by 0.01 seconds…

…in Lane 1.


Name: Cody Steele • Tuesday, July 31, 2012

The idea of a lane effect is really interesting, because I don't think that the lanes are completely randomized after the first race. Page 88 of the International Rowing Federation rule book says:

"For the repêchages, semi-finals and finals, the principle is to put the crews with the best placings in their previous round in the inner lanes. Similarly the crews with the lower placings in their previous round are put in the outer lanes. If the crews have the same placings in the heats, repêchages or semi-finals then there shall be a draw to determine their lanes in the next round."

Is there an interaction between lane and round? Maybe the effect is small/nonexistent in the first round, then gets progressively smaller as the best rowers are the only ones left?

Name: Joel • Wednesday, August 1, 2012


Great points, and I definitely was incorrect on stating that lanes are randomly assigned...instead it should be noted that rowers rowed from multiple lanes and didn't repeat lanes, so you had some built in "randomization". But it is not truly randomized.

There was no significant interaction between lane and round.

It would seem that the idea behind the rule is to "protect" the fastest rowers from the wind by guaranteeing someone is rowning in a lane on the windward side of them. However, since that rower should be a weaker rower than you, you would quickly pull ahead and be directly in the path of the wind. The real winner is the weaker rower who gets the most leeward lane and may get to keep up with the highest qualifiers by rowing through their wind shadow.

Not that the International Rowing Federation is asking, but in my opinion lanes should be randomized for the heats and following that, lanes should be picked by the rowers in subsequent rounds with the fastest qualifiers getting to pick first. Then they could asses the wind before choosing. Also, if you're really looking for fairness, the qualifying heats should be done one boat at a time so no one gets a wind advantage.

Thanks for the comments!

blog comments powered by Disqus