somewhat OT: World's ITT pacing

lanierb

Robert Chung recently put up a graph of the world's ITT times here:

http://anonymous.coward.free.fr/rbr/world05itt.png

(The course was two laps of the same course. The X-axis on the graph is first lap time and the Y-axis is second lap time, with the 45 degree line drawn in.)

Here's my question: not a single rider rode the second lap faster than the first. Only four riders (including Rogers who won) even came close, and the vast majority went way slower on the second lap than the first. Vinokourov, for example, did 26:49 then 28:06. The graph shows his splits as typical. Note that this can't possibly be due to a wind change (which would have affected people differently depending on start order) because everyone is the same.

Am I missing something, or am I right in concluding that essentially everyone went out too hard on the first lap? I guess I'm just surprised that none of these world class cyclists could get their pacing anywhere close to right. I suppose not many had a power meter with them.

Another funny observation...here's a quote from cyclingnews.com:

"Rogers starting slowly, placing only fifth at the first check, but while early leader Bradley Wiggins (Great Britain) faded, the Australian continued to build speed."

Of course the quote is dead wrong (assuming I didn't miss anything). Even Rogers died a bit on the second lap. He just came a lot closer to pacing it right than pretty much everyone else.

Lanier

biker-linz

It depends where you 'split' the course. I have another graph where the split is 1/3 to 2/3s. In this case, 12 of the top 15 rode easier in the first 3rd than the subsequent 2/3s. Wiggins is nearly off the chart in the first 3rd of the course. If what Chung posted is correct (and I'm sure it is) then most of them went out easy, went hard in the middle and then possibly faded towards the end.
Still, it just goes to show the pros aren't the 'be-all-and-end-all' as far as how things should be done.

L.

__________________
MSc (Applied Sport and Exercise Science)
RST Associate Coach
ABCC Level 3 Coach
Doctoral Student (Physiology), University of Oxford, UK.
www.cyclecoach.com
www.science4sport.com

martind

Although a preliminary analysis of the splits may indicate that the the pro's don't know what they are doing when it comes to pacing a TT, it is also worth considering that many of the pro's are riding to win. The pro's I have had a chance to work with know that to be in the hunt for a medal a solid start is required. If they are having a good day they can maintain their pace whereas if they are having a bad day the pace slows considerably. Obviously Wiggins went out way too hard but he had to attempt a very agressive pacing strategy if he was going to get a medal. As far as splits go, I am sure most would agree that the best evaluation of pacing is based on power not speed (obviously terrain influences speed). A 1/3 split puts you at the top of Citra de la playa (~14 km), the top of the second major climb. To compare this split to the final 2/3 of the race may not be entirely appropriate. I like the lap to lap comparison myself.

biker-linz

Goodness me, I don't think I said that; just that they can get it wrong and do (the only professional rider I work with is as human as you or me ).
All the riders should have a pretty good idea of what power output they can maintain for a given duration, but TBH I don't know how many riders were using PMs on the day (and if not, why not). No matter who you are, if you go out way too hard then you will pay the price.
Of course, and I don't have power files from any riders who competed last week (do you?). I didn't do the analysis myself; the split was at 10.5 km, I'm not sure why that point was chosen.
As seemingly every rider adopted a similar strategy I wonder if it was deliberate or not?

L.

__________________
MSc (Applied Sport and Exercise Science)
RST Associate Coach
ABCC Level 3 Coach
Doctoral Student (Physiology), University of Oxford, UK.
www.cyclecoach.com
www.science4sport.com

biker-linz

Sorry Martin, I should have said: you are of course absolutely right about comparing speeds. I guess the only conclusion we can draw from those are comparitive; i.e. from rider to rider.

L.

__________________
MSc (Applied Sport and Exercise Science)
RST Associate Coach
ABCC Level 3 Coach
Doctoral Student (Physiology), University of Oxford, UK.
www.cyclecoach.com
www.science4sport.com

RapDaddyo

Where can I get the raw data for these splits? I'd like to run some of these through my pacing model.

rmur17

http://www.cyclingnews.com/road/200...sults/worlds053

this should keep you happy for a while!

Robert's graphs show things quite nicely but the intermediate splits show for example that Wiggins went out way too fast. Still a darned good result.

Here's the course profile BTW (2 laps).

rmur

lanierb

The big question is could Gutierrez have won (needed 23 seconds) if he had paced himself properly?

Lanier

lanierb

I'm sort of sympathetic to that idea, except in this case it would imply that 44 out of 48 had a bad day, including the 2 out of 3 on the podium. That doesn't seem right to me.

Lanier

rmur17

Not if we assume that proper pacing would have allowed him to do two laps at the average of his 1st and 2nd lap powers.

If we assume his 1st lap was at 450W, I make the 2nd lap ~406W and the average 428W. Running the model with 428W for both laps does indeed save some time but not 20+ seconds. Actually closer to 5 seconds by my reckoning.

Same goes for Wiggins but he lost ~15% on his 2nd lap. Even Rogers dropped a little ... maybe 5%

Interesting question though .... and of course the usual disclaimers apply..

rmur

p.s before someone screams NP - run the numbers and you'll find they are within a watt - basically noise at those power levels.

lanierb

So ~5 seconds is the drag savings from having a more constant power output.

I just googled around and found numerous studies suggesting there would also be a physiological benefit from better pacing (i.e., you should be able to achieve a higher average power that way). Some studies mentioned as much as 5%, but this seems a bit large for a professional cyclist. Allowing 5 seconds for drag savings, Gutierrez only needed another 18 seconds to win, which by my calculations would require an additional 1.6% or so (7 watts or so) in average power. Possible?

Lanier

rmur17

yep I agree with those calcs ~435W average power over both laps would have done the trick for him. Maybe the next book will be entitled "Every Watt Counts"

I should also add that the 450W starting point was a guess based on the reported power output of some top cyclists. It's possible that Guterierrez was 'only' averaging ~400W but the 7W delta should be pretty darned close even in that case.

rmur

martind

Hmmm... just back from a trip and had a chance to read all the comments posted thus far about pacing at the 2005 WChamps time trial. lanierb makes a very good point and got me thinking - could all the Pro's really be having such a "bad" day?.

So I went back and checked the splits posted at www.cyclingnews.com. I recreated the graph that was posted earlier comparing lap 1 velocity to lap 2 velocity. I assumed the times at 44.1 km and the times at 21.9 km were correct and then calculated the average velocity for the second lap (22.2 km). Based on my analysis 4 cyclists rode the second lap faster (>1%) including Mick; 12 cyclists rode the second lap within 1% of the first lap velocity (even pacing) including Gutierrez and Cancellara; and 32 cyclists rode the second lap >1% slower than their first lap (including Wiggins who was ~3.5% slower).

Take a look at the attached graph. I would be interested to see if anyone else has checked these numbers and seen the same thing. Maybe most of the pro's are pacing the TT as expected - at least 50% appear to have produced "even" or "negative splits." The other 50% might have been having a "bad" day or mispaced their effort.

Attached Files

Pacing Pic.doc (76.5 KB, 14 views)

frenchyge

Great stuff, guys. This paragraph has got me confused, though. The OP's reference said that no one rode a faster elapsed time on the second lap, and now this one (quoted above) says that 4 did? Is that right, or are the comments above based solely on instantaneous velocities at 2 distinct points on the course (44.1km & 21.9km) instead of a lap-average?

Edit: or maybe the difference is that the cyclingnews.com splits are deceiving because they are given at the 21.9km point instead of the 'true' midpoint of the race?