TdF average speed

SlowRower

Something for everyone there, I suspect!

Fontfroide

Would it be at all interesting to put into that graph the fastest time of the year for the 100 metres on land and in the pool? Just to see.

Bartali

Need parcours info etc to be meaningful, but a general trend of increased km/h - 6km/h in 30 years. Quite significant. Bikes weigh less and roads are a bit better. I'd like to see the graph go back to the 1940s

Nolte

Boogerd_Fan

riders are specialist professionals now too, instead of chimney sweeps...

but then if the graph did go back to 1900, surely the progress would be in the first half-century, rather than over last 10 years. That 6kmph faster trend is certainly curious...

what is different about today's peleton than 80s?
Can't just be about the hog in the earpiece can it?

Jackov

Nolte

what about between 1980 and 1981? was there anything extraordinary. it was in the pre-me era so i'm not sure

is it the average winners speed as i though 2005, 2004 and 2003 where the 3 fastest years or so

Jackov

Here's back to 1947

Biosphere

Jackov

Biosphere

headwind

Hi All, Every year I trot out this garbage, so might as well this tour as well. I was fascinated with all this too a while back and dug out all kinds of shite from the recordbooks. Heres a few figs.

This one clearly shows the inverse relationship between tour speed and distance over time. Whether this intersection point in the late 80s is meaningful Im not sure...but I bet Lemond would say it is!

...the idea its just getting faster is largely crap:

headwind

This plot shows the average speed for three races...P-R (black), LBL (blue), TdF (red). I have applied a smoothing function to the data sets to even out the variation and give an overall impression.

I think its quite interesting. I see the following: PR and LBL have been very constant in parcours over a long time, and in the early 60s, we see a dramatic leveling off of speeds, whereas the tour is not levelling off at the same rate. It tells me that the continued monkeying of the Tour, essentially making it easier, by making it SHORTER (overall, and each stage), has kept the speeds up. It also shows the advances intraining, bikes, and doping have not had a quantum effect on speed, suggesting we are at the capacity for human achievement.

The fastest PR ever was 1964!

headwind

On top of tour distance getting shorter, stage distance is getting shorter, permitting faster speeds.

.

headwind

plot showing actual data and "smoothed" data for PR

headwind

I thought Id also add my 2 cents.

Clearly the years between wars were teh most phenomenal in cementing the rigor f the major races. that 3 races show the same thing says the old days were pretty incredible...and why records were falling all the time. It might also be linked to the seriousness of the endeavor and certainly in bike construction. All these do point to huge advances in cycling, just not in our time. It strikes me the immediate post WWII years were really an apex in cycling. that bikes weigh about half now, with 2x the number of gears basically says those guys were monster hard asses on the road. Add road surfaces changes ago, and its clear we stand on the shoulders of giants today. Really doping and bikes and training etc have inched up the speeds from the 50s, but not hugely when looked at through the perspective of a lens where great races have been largely constant over time (PR, LBL). The Tour has been such a moving target that focusing on speed give a very partial picture, and one where we might easily be convinced that cyclers are getting materially much better...but I dont think thats the case!!!

your serve gang

hw

SlowRower

Thanks for the analysis, HW. Definitely something to think about!

Here's a thought for you - does making the TDF stages shorter necessarily make it easier? Riding to your limit over (for example) 250km is probably not that much different to riding to your limit over 180km, as you do the latter faster, so that in both cases, the tank is empty at the end of the effort.

Admittedly, I've never ridden anything like 180km in one go, so the above is guesswork, but I've done a lot of running and rowing over various distances and short endurance events are every bit as hard as long endurance events. (Harder in one context, actually, as your cruising pace is closer to your absolute maximum pace in the shorter distances, and "peak pain" is higher! As an athlete, I suppose which you consider harder depends on whether you like lots of pain for a short time, or lower levels of pain for sustained periods. I always found the most preferable option was the one that I wasn't currently undertaking. Smile

)

Jackov

Headwind siezes the torch of scholarship! Excellent.

I believe a huge part of it is greater participation: the pool of athletes has grown, and athletes have more exposure to different sports to discover their specialty.

Bartali

I can't see HW's graphs on this computer, but looking at Jack's last graph there seems to be a steady trend up till about 1960 and then another steady rise starting about 1970. Something odd in the 1960s. Ideas? Apologies if its clear from HW's data.

headwind

Getting back to the tour dustance and avg speed issue. I posted this elsewhere and thought it was worth reposting.

Here is total distance of the Tour plotted against Average speed. Total distance certainly does not account for all the the subtle variation, but over time it exerts a strong influence on how fast the tour is. As we revel in, faster speeds, lets not lose perspective on the overall principal cause of it...this is especially true when considering The pre-Indurain era, and especially, Merckx and Anquetil, and Coppi and...

The Indurain and Armstrong eras are showing less fluctuation in speeds and fine correlation with Tour distance, suggesting that other factors are at play, like individual race conditions and perhaps the limits of human indurance against the race itself (like Paris-Roubaix perhaps).

Interestingly, if you extrapolate the distance data to 0 km (I guess that officially is the end of the tour!), it occurs in 2160...but if you project an avg speed for a 0 km race (lets say a few km), it suggests pros could ride bikes, on avg (not sprinting) at about 59 km/h...which is exactly in the bounds of reason! It suggests this argument then is certainly resonable. Keep shortening and well see ever faster times...theres a surprise, eh?

Mrs John Murphy

HW - how do you take into account time trialing and climbing within the parcours, split stages, etc. Surely the amount of each discipline may result in a distorting of speed ie one or other is top heavy in the race.

headwind

mayofan

in jackovs 1947 to now graph there seems to be a lot of "shark teeth" in the first half of the graph especially, that is, up one year down the next, up the next year.

could there be a correlation between the direction the tdf goes and average speed? interesting, does doing the pyrenees first make the race slower or faster.

would be interesting to see average speed for years it went through pyrenees first, and average speed for alps first years.

Jackov

I saw those sharkteeth earlier and had no idea. Interesting theory about the direction.

We would need a graph with line for clockwise tours and one for the counter-clockwise

headwind

From 55 to 65 the saw tooths are opposite each other for distance and speed. that cold explain some of it. interestingly the later years have much more noise and would say that the distance is not the sole factor in individual years speed.

berck

Great analysis HW. Very interesting results. There does seem to be a correlation between the distance and speed.

headwind

Thanks Berck.

Id also add that on the tour speed through time plot we can glean the following. If we were to claim that advances in the last decade, like training, carbon fiber, wind tunnel testing, specific designer dope products, or whatever were the cause of the rising avg speeds, this would require there to be a steeper slope of the line for tour speed in the last decade. In fact the data of the last decade (up to 2005 when I did this) lay just above the best fit line on the same slope that extends back to the 50s. This says the cause of the progressive rise in speed is long term in its making, not just in our times.

just a thought!

CapeRoadie

That's my take on it as well, hw. Have always enjoyed your analyses.

headwind

Thanks doc.

Here is the detailed data for PR and LBL since WWII. I decided to post this because of the "sawtooth" comment raised by mayo. This turns up obviously in these races too, where parcours is not substantively different in direction. I would guess this means the annual fluctuations are more the effects of weather, specific individuals, parcours changes etc. Whats interesting is that these fluctuations still occur indicating that bikes and training are not likely to be as significant as local matters of the race day.

The systematic frequency int he PR data very much interest me. there is this clear 5-7 year frequency that seems to beat though much of the last 50 years...I wonder what thats about????!!!

hw

mayofan

thats a weird one hw! any ideas?

the only thing i can think of is some kind of weather pattern, or maybe length of dominance of particular star riders: say 10 years ago museeuw was at his peak, then as he was declining and boonen emerging the speed fell, and in the past few years boonen has improved drastically, edging speed up, and as he declines and the next star emerges it will drop?

doesnt make sense but its all i can think of?

its a pretty defined trend...weird!

Mrs John Murphy

headwind

headwind

SlowRower

HW,

Agree that shorter is faster, if undertaken at the same degree of "flatoutness". In road running and endurance rowing, the general rule of thumb is that if your max average speed for a distance is x, then if you double the distance, you would average 0.95x.

For rowing, where there's no impact damage, the the same basic endurance training programme will prepare you equally well for a 5k, 10k or half marathon - you just need to do one of them flat out to be able to estimate a likely average pace for the other distances.

(Road running is different due to needing to specifically train up to deal with the impact damage, but once this has been done, the relationship holds very well on average.)

Jackov

I think increased TV exposure is driving up the speed.

A way for a sponsor to gain tv exposure is to be in the break for the broadcast. So there is more pressure and reward to be in the break and that shoots up the competitiveness of the break. It is the break that defines the speed for most days

Mrs John Murphy

Nolte

here's something that might be another indication that the peloton is getting cleaner

the amount of retirements,

i think with 24 retirements this year, that significantly down on previous years even if you think into account 2 years ago when several teams were withdrawn from the race for having dirty cheats on their team

Bartali

Getting cleaner or the parcours was relatively easy?

SlowRower

Bart,

Why do you describe the parcours as relatively easy?

The "ease" from a physical viewpoint depends on the speed it's ridden at. I'll grant you that the parcours promoted a style of racing that led to less flat out efforts up hill, so it probably was easier for the autobus, riding solely to avoid the time limit in the hills.

Conversely, for the top guys, the course was "hard". There were limited places where they could gain significant time over weaker riders, so where the route did lend itself to attacks, the attacks were vicious, as this was the only way to gain serious time. Such significant stages may have been fewer in number than in normal years, but their intensity would have been higher, requiring a VO2 max of nearly 100 if Greg LeMond is to be believed. Smile

Finally, the reduction in opportunuties for gains in the hills made the tactical aspect more important than normal. The top riders had to be constantly alert even on seemingly innocuous flat stages, knowing that one false move could cost valuable seconds, with far fewer opportunities to get this time back than normally. This extra mental strain would eventually take its toll.

FWIW, I think an equally significant reason for this year's "boring" TDF was the fact that most of the GTGC big hitters didn't race it or were badly off the pace from the word go for Giro reasons or unspecified Evans-reasons. Given that A Schleck was the only realistic rival to Bertie and is crap at ITTs, a more "exciting" course would probably still have been pretty dull. Most of the Armstrong wins are described as "dull" (or worse) and they took place on more traditionally "exciting" courses.

So in summary... Dull? Yes (although not without a fair deal of interest, if not excitement.) Easy? Very much doubt it for the top guys, albeit in a different way to normal years.

Bartali

SR - The context was Nolte's last post and by 'easy' you hit the nail on its head in your first paragraph. Easier on the gruppeto hence fewer drop outs than most years.

SlowRower

Bart - Indeed. The autobus could go slower uphill than in normal years, as there was less scope for time-limit threatening gaps to be generated.

I wonder how many of the top guys modified their training to account for the different demands that the course would present this year compared to normal.

thunderthighs

1980 ---1983.. cuz epo began to imerge..

thats the greg lemond era...

oh well

kathy

For those who can read French, you might find this interesting (and the links to previous races as well).

http://www.cyclismag.com/article.php?sid=5207

Biosphere