Imagine for a moment that you have been hired as a pace-setter, or rabbit, for a professional track meet. Unless you are one of the very few runners who have made a career of rabbiting (thank you, Matt Scherer), or have become the personal rabbit of some other great athlete (thank you, Sammy Tangui), chances are that you’re a little inexperienced, as well as a little anxious about getting the pace right. Specifically, you have been told to run splits that are “too fast” for your ability over that distance. There might even be incentives in your contract that reward or punish you based on hitting or not hitting those splits. As you stand there on the starting line, adrenaline coursing through your body, you have one fearful thought in mind: I must not be too slow!

Now, I think there’s a human tendency to respond to performance anxiety by trying too hard. Many years ago, the director of my college’s concert chorus told us that when you had to come in on a very high note — a note in the uncertain register of your vocal range — you had to guard against stressing out so much that you “overshot” and came in sharp, i.e., too high. If you’re trying to run 400m at a pace that you know will feel too fast, I think the natural tendency is to try too hard, overdo it, and split that 400m too fast.

If the rabbit’s incentives and inclination is to err on the side of too fast, what of the legitimate contenders, what are their incentives?

I would argue that for most competitors, there is FAR more risk in going out too fast than too slow. Especially when the target pace has been selected to put the field into position for records or world-leading times, running even a half second too fast (in an 800 or 1500) can be fatal. So what do these competitors do when the rabbit runs the first 100-200m a second fast? They bide their time and let the rabbits go, that’s what they do.

Consider the women’s 1500m at the recent Diamond League Meeting in Lausanne. The field was loaded — Jenny Simpson, Sifan Hassan, Faith Kipyegon, Mercy Cherono, Dawit Seyaum, — and race organizers were hoping for fast times, certainly times well under 4:00, if they could engineer it. So they gave the pace-setters ambitious goals: the first pace-setter, Selma Kajan, was to take the race through 400m in 63 and 800m in 2:06; the second pace-maker, Beatrice Sitonik, was to hit 1200m in 3:10.

Amped up for such a fast pace, Kajan ran her first 100m in just under 15 seconds, and her first 200m in roughly 30.5 seconds. By then, it was too late. The pack had refused to run that much faster than the optimal pace, and there was already a two-second gap from the rabbits back to third place. It was all depressingly familiar. It was also predictable because all of the runners were acting in accordance with their incentives. Kajan would go on to split 61.5 for the 400, and then slow to 2:08 for 800m. It didn’t matter; far behind the rabbits, the main pack had long ago stopped worrying about hitting their splits.

I believe the way to avoid such pacing disasters is to completely rethink the incentives for the first rabbit, based on the observation that it’s not only the split that matters, but how the split is achieved.

Specifically, rabbits should be told to hold back at the start, and be penalized for going TOO FAST, rather than going too slow. The most important thing is to be running the correct PACE when each milestone is reached, and to do that, the rabbit MUST not go out too fast. If the race is a 1500m, rather than bolt to the lead, a rabbit should start with some restraint while staying safely on the outside for the first 100m, and then gradually ease into the lead around the turn. The goal is then to accelerate, if need be, and be running at exactly the target pace (63s per 400m) at the first checkpoint, 300m into the race.

So let’s use the Lausanne 1500 as an example: Instead of sprinting the first 100m, Kajan runs the first 100m closer to 15.5 or even 16 seconds. Depending on what the field is doing, she either maintains that pace or accelerates into the lead around the turn, and now she’s a) immediately in front of the main field, and b) running at approximately the right pace (15.75 per 100m/63 per 400m). Even if the first 100m was a bit slow, the first 400m will be within a half second of the goal split, and, more importantly, the rabbit will be ON goal pace, which makes it far easier for everyone involved to keep hitting fast splits.

Some will object that it’s just as hard to judge pace accurately going out more slowly, but I disagree. I think that when someone is not under pressure to “hit the high note,” the result is a more even effort, which is what’s needed. Also, I believe the risk of being too slow at the start is far less consequential than being too fast. If anything, the rabbit should have her pay docked for running the first 300m more than a half second faster than planned.

Of course, this whole discussion leaves aside the question of whether pacing is desirable in the first place, or an abomination that detracts from the true competitive essence of the sport. But let’s agree that if we’re going to have pacing at all, let’s not waste it by spooking the pace-setters into bolting for the front like frightened…um…rabbits.