Team Time Trial

I have done some work on this problem, both in theory and in practice. I believe it is the most complex mathematical modeling problem in cycling, because the optimal pacing plan is driven by the power and power/weight ratio of the rider with the least power. The power/duration of each rider on front is designed to avoid pushing the rider with the least power beyond his sustainable power. Then, of course, you have to factor in the cost of rotations (i.e., loss of one bike length per rotation). The simplest model would be to take the lowest common denominator of target power for the ride duration and divide by 0.70. This would put you in the ballpark for the power of the pulling rider (at least on the flat). For the rotation durations, a quick and dirty formula is to sum the target power for the duration for each rider and allocate pulling durations pro rata. So, let's say you have 2 riders at 250W, 2 at 275W and 2 at 300W. You could allocate time on front as 15%, 17% and 18% respectively. Target power on front under this scenario is ~357W (250W/0.70). On upgrades, the lead rider has to throttle back to maybe 300W and on the downrades the lead rider can let 'er rip. This was a bit cryptic, but hope it helps.

One more thing. Communication among the riders is huge. The team really needs to ride together in TTT format a lot to get it down. The guys with the most power must learn to throttle back and not push the riders with the least power over their limit. If you drop a guy, you need to slow down and get him hooked on again in a matter of seconds or you will lose time. Communication and pacing are everything in TTTs.

Very good job RD.

Do you know if there are teams actually managing their power durations that way?

How has it been to actually put that into practice? All the guys calculate their respective durations with the clock on their onbike cpu?

SolarEnergy said:

Very good job RD.

Do you know if there are teams actually managing their power durations that way?

How has it been to actually put that into practice? All the guys calculate their respective durations with the clock on their onbike cpu?

I don't know of any teams that manage their power durations this way (even though it is the optimal solution). To do this on a flat course with no wind isn't such a big deal because the power/duration combos are constant and one could use any sort of countdown timer. Adding hills and wind increases the complexity of the optimal pacing plan enormously and without some on-bike technology to support such a strategy it would be sort of a shot in the dark. But, who knows, such technology may be here one day. Wouldn't it be cool if the on-bike computer somehow knew when you were on front and automatically started your countdown timer and gave you your power target?

The courses I have used this strategy on have been relatively easy because there are long segments at a relatively constant grade. So, I have only had to work out the power target for the segment and assign the pull durations. The upgrade segments are easy because the pulls are relatively long and everybody rotates. The downgrade segments are way, way more difficult because the pull durations become very short (~15secs) and only ~1/2 the team rotates (the guys with >300W FT). If the downgrade is more than ~1%, the guys on front pull their guts out and we still can't put any pressure on the guys with the least sustainable power. Hell, on a 2% downgrade I can be in the middle of the team doing 35mph and ~150W while the guy on front is blasting away at 450W. Rolling courses would be infinitely more complex (solvable, just way more complex). Communication is the hardest problem, with the road and wind noise. I am planning to try some 2-way radios because we have a hell of a time relaying info up to the guy on front.

SolarEnergy said:

Very good job RD.

yeah he is like all analytical and stuff. And I was just going to say take 30 second pulls then peel off :o

Pretty cool what we can do with live power data huh? I too wonder if the Pro teams will do this, certainly some of them are. But it sure would be a competitive advantage if not all are.

wilmar13 said:

Pretty cool what we can do with live power data huh? I too wonder if the Pro teams will do this

Some are willing to do anything to win. Ask T-Mobile CSC and all 😉

RapDaddyo said:

If the downgrade is more than ~1%, the guys on front pull their guts out and we still can't put any pressure on the guys with the least sustainable power. Hell, on a 2% downgrade I can be in the middle of the team doing 35mph and ~150W while the guy on front is blasting away at 450W.

And why is that?

You mean that for the power level to say in an acceptable range for the members in the middle (or the end) of the pack, the guys on front needs to pull even harder I guess (just thinking out loud 🙄 )

Thanks RD

SolarEnergy said:

And why is that? You mean that for the power level to say in an acceptable range for the members in the middle (or the end) of the pack, the guys on front needs to pull even harder I guess (just thinking out loud 🙄 )

It's because the drafting benefit goes from ~30-35% on the flat to >50% on a downgrade (obviously a strong wind can affect this either way). So, it takes a rider pulling at ~450W to even get the other riders into the 200s. Done right, the downhill segments are really hairy, with almost continuous rotations on front by 4-5 of the most powerful teammates and the rider first in line behind them has to be very good at opening/closing a bike-length gap for the rider coming off the front. You keep the pulls <15secs to exploit the physiological response half-life, so that it's equivalent to the AP for the guys on front (~260-270W assuming 4 riders sharing the work equally). If you extend the duration of the pulls, then NP skyrockets. TTTs are great events. I wish there were more of them. I am considering putting on a TTT here in LV.

RapDaddyo said:

It's because the drafting benefit goes from ~30-35% on the flat to >50% on a downgrade (obviously a strong wind can affect this either way). So, it takes a rider pulling at ~450W to even get the other riders into the 200s. Done right, the downhill segments are really hairy, with almost continuous rotations on front by 4-5 of the most powerful teammates and the rider first in line behind them has to be very good at opening/closing a bike-length gap for the rider coming off the front. You keep the pulls <15secs to exploit the physiological response half-life, so that it's equivalent to the AP for the guys on front (~260-270W assuming 4 riders sharing the work equally). If you extend the duration of the pulls, then NP skyrockets. TTTs are great events. I wish there were more of them. I am considering putting on a TTT here in LV.

That'd be pretty interesting to have all of the TTT riders using a powermeter so your could see how the four outputs interacted throughout the course. Along with modeling you could probably highlight every part where poor power management cost time.

Post analysis of TTT files would be semi time consuming but you improve a lot by laying all the files on top of each other along with the rotation plan.

With the downhills, since increasing power output on downhills will give you the least speed advantage versus uphills, how do you balance that with the fact that downhills allow only 1 person to be killing it while uphills are the opposite? Uphills at a high ouput stress the whole team more, maybe blowing up the weakers rider.

Greg

gvanwagner said:

That'd be pretty interesting to have all of the TTT riders using a powermeter so your could see how the four outputs interacted throughout the course. Along with modeling you could probably highlight every part where poor power management cost time.

Post analysis of TTT files would be semi time consuming but you improve a lot by laying all the files on top of each other along with the rotation plan.

With the downhills, since increasing power output on downhills will give you the least speed advantage versus uphills, how do you balance that with the fact that downhills allow only 1 person to be killing it while uphills are the opposite? Uphills at a high ouput stress the whole team more, maybe blowing up the weakers rider.

Greg

Hi Greg,

Yes, the speed delta/power delta is unfavorable on downhills and for a solo TT this is where one wants to back off and conserve power. In a TTT with riders of various sustainable power levels, the logic flips a bit. This is because the team members with the greatest power can't really use their power on the uphill segments because they'll blow up the team members with the least power. They can use their power on the flat by taking longer pulls. But, they can really use their power on downhill segments. It's really unbelievable how much power it takes on front to stress the rest of the team on a downgrade. If one looks at the totality of the problem (each rider's sustainable power, weight, road gradients and wind), the mathematical modeling problem is really quite complex. Then, of course, one has to have a way of transferring the pacing strategy to the course. Then, finally, one has to have a way to deal with the inevitable gaps that will open up since the riders with the least sustainable power are going to be riding right on the edge at all times. If one of them has to try and chase back on, it's disastrous for the team since these riders are least able to handle the added stress of a chase. So, it's crucial to get them hooked on again almost immediately. When I do these rides, each team member checks his rear wheel about every 15secs because we can't assume we will hear the rider behind us with the road and wind noise. I tell the guy on my wheel to tell me in advance if he is going to intentionally open a small gap (e.g., to take a drink) because if I look back and see a gap of a bike length or more I'm going to send the message to the front to slow down to close the gap.

RapDaddyo,

thanks for the help, its really a nice approach for the time trial.
But could you give me (a rule of thumb) the length of the pulls?
You only mentioned the %difference for the riders, but how do you set the length of the pulls? Or was this % of a minute?

Again, thanks a lot

Bart

bartjoosen said:

RapDaddyo,

thanks for the help, its really a nice approach for the time trial.
But could you give me (a rule of thumb) the length of the pulls?
You only mentioned the %difference for the riders, but how do you set the length of the pulls? Or was this % of a minute?

Again, thanks a lot

Hi Bart,

The lengths of the pulls is a two-part question. One part is the allocation of total time on front to each team member. The second part is the absolute duration of the pulls. The first part can be arrived at for flat segments of the course by summing the sustainable power of each team member and then allocating the pull time pro-rata. So, let's say (as in my example above) that the riders' sustainable power for the duration of the ride are 2 x 250W, 2 x 275W and 2 x 300W. The sum of the sustainable durations is 1650 and the allocations are 1.00, 1.10 and 1.20 respectively. So, let's say the riders with 250W take turns on front of 60 secs each. Then, the riders with 275W would take turns for ~66 secs and the riders with 300W would take turns of ~72 secs. This may seem to be a trivial difference, but if you look at it over the course of an hour, the riders with 250W would be on front for ~9 mins and the riders with 300W would be on front for ~11 mins. The target power on front is intended to push the riders with 250W of power right to their limit. Assuming a drafting advantage of 30%, this would be 250/0.70=357W. Now, the drafting advantage will average closer to 35% with a team of 6, so that is why I haven't adjusted for the fact that (5x250 + 1x357)/6 = 267, which is greater than 250, much less NP which will be higher yet.

The absolute pull durations is a tradeoff between the physiological response half-life and the loss of one bike length per rotation. To exploit the physiological response half-life, you would keep the pulls to 15secs or less and that is exactly what I would do on the downgrade segments because the power is so high on front (e.g., 450W). I would also move the 2 riders with 250W to the back and they wouldn't pull at all on the downgrades. They are the primary constraint and you want to save them when you can. They will have much more difficulty with 450W, even for short durations, whereas for the riders with 300W of power this is just a routine, short AWC interval. On the flat, I would probably extend the pull durations to ~60/66/72 secs respectively. On the upgrades, I would change the power target to <=300W (because the drafting advantage drops to maybe 15% at 2-3%, or near zero at 6%) and would change the pull durations to ~30 secs, 1 min and 2 mins respectively. On the upgrades, the riders with 300W are actually riding below their sustainable power even when they are on front. The upgrades are where you will blow up the riders with 250W if you aren't careful, so power management and communication here is crucial. If they begin to feel they are getting in trouble, they must call for a slower pace immediately. They must be told that this is not the time to "tough it out," because they are jeopardizing the entire team by keeping their pain to themselves.

I dont for one moment want to take anything away from what has already been said, but in my experience in TTT's over the past 2-3 years, where there's been no power metering to mention(until I got mine late last year), the average time spent by the leader is between 30 seconds and 1 minute. The TTT's I've done are either 10 or 25 mile distances, some on flat terrain and some on more rolling courses, and they are very similar to interval training, where you're over your limit whilst on the front, and recovering when behind.
The hardest part to get right is after coming off the front, when you have to push hard to get back onto the last rider's rear wheel, which essentially means saving a bit for this purpose.
Most of the events I've done have been with team riders faster than myself, and I have always come away from these events having pushed myself harder than I thought able. More recently, I have been the more powerful rider, and as RD mentions, it is so easy to take the front position and ride away from the team, so you have to get all the team to shout hard as soon as they, as individuals, start to drop away.
The other factor to consider is that a lot of club level cyclists(many without power meters) ride TT's using their own pacing strategy, which may not be the same as the other team riders, so it is worth getting some practice in as a team, in order to get some kind of synergy and flow to the team.

AndROOb said:

the average time spent by the leader is between 30 seconds and 1 minute. The TTT's I've done are either 10 or 25 mile distances, some on flat terrain and some on more rolling courses

It's true that 30-60 second rotations work well for flat and rolling courses. For courses with extended downgrade segments, the power required on front is so dramatically higher that shorter rotations are virtually required. I have been working with a team on a 25 mile course that has two extended downgrade segments, one of 7.5 miles and one of 9 miles. If even our strongest riders stayed on front for 30 seconds per rotation (at ~450W), they would blow up very quickly because their NP would skyrocket. And, if they do 30sec rotations at ~120%FT, the rest of the team would be coasting 80% of the time.