Mathematics & 1989 TDF

[deValtos]

My dad showed me this today which I thought was quite intriguing:

http://mathoverflow.net/questions/51299/there-are-mathematics-behind-the-1989-tour-de-france

Basically when the 1989 TDF was run they calculated the times at the end of the stages only in integers so each rider has their time rounded to the nearest second. With only 8 seconds seperating the first 2 riders it becomes interesting as it is possible to make 8 seconds up in rounding 'errors' ... well read the link it'll explain better than I can.

This got me wondering how do they calculate the times nowdays ? Is it the same so they only count whole seconds ? I say this because with current style of racing in GT's especially in the TDF it's not inconcievable that we have another extremely small time difference ... say 2" or 1" then well that could be risky.

 

[Lanark]

In time trials they measure in hundreds, in normal stages only in full seconds. Those small differences in time trials are usually unmentioned, unless two people are in the same second.

 

[biokemguy]

Problem solved by time bonuses

Can't remember the riders (Contador and Leipheimer maybe), but I think the top two in the Vuelta a few years ago had the exact same actual time, but the time bonuses throughout the race made the final difference between 1st and 2nd something like 30 sec.

Rounding errors in this case were overcome by time bonuses...a pretty good solution IMHO. And it makes the racing a bit more exciting.

 

[Armchair Cyclist]

Because it takes a peloton a long time to cross the line, with all given the same time, it is very possible that if Fignon was habitually near the front of the bunch, and Lemond relaxing farther back, that Fignon spent more time on the course than yellow jersey winner, but riders act according to the competitive criteria of the event, not the "raw" time taken.

 

[R.0.t.O]

Because it takes a peloton a long time to cross the line, with all given the same time, it is very possible that if Fignon was habitually near the front of the bunch, and Lemond relaxing farther back, that Fignon spent more time on the course than yellow jersey winner, but riders act according to the competitive criteria of the event, not the "raw" time taken.

Exactly, them's the rules and inevitably there are consequences. I've been 8th rather than 2nd on GC in a stage race once because a 1 sec gap got me placed with one group rather than another and with riders lined out on a climb this equated to 11seconds of 'race time', so a rider finished 1s ahead of me but is given 11s on GC. That's life, and that's why you have to ride at the front.

 

[Skibby the bush kangaroo]

By my reckoning there were only 9 stages where Lemond and Fignon were not at s.t. (prologue, stage 2 TTT, stage 5, stage 10, stage 15 TT, stage 16, stage 17, stage 18, stage 21 TT). Assuming each rider's time is subject to an independent uniform rounding error between -0.5 and 0.5 seconds, it is theoretically possible to account for 8 seconds by rounding error. However, the probability Fignon would have won if not for these rounding errors is somewhere around 1 in 800 million. I think that's about the same chance Lance won 7 tours clean.

 

[Jonny7c]

Is there not also a theory floating around that Fignon would have lost sufficient seconds in the ITT to win if he didn't have long hair?!

 

[will10]

If he'd not had long hair, or not used a front disc, or used tribars, easily 10 seconds+ each one of those factors