Outliers and Tony Martin

barrybridges
barrybridges Posts: 420
edited January 2014 in Pro race
In reference to the fact that Tony Martin has just destroyed the field in the World TT champs, I have a question I've always wanted to ask.

If you think of how many people in the world cycle on some kind of competitive or at least serious basis, it's a very very big number.

Now if you rank all those people in order of their time-trialling ability, from best to worst, you'll have a long line of millions of people.

I've read Outliers and I understand the concept, but what I don't get is why there could ever be such a big gap between #1 and #2 in the world on any given day.

Tony Martin was 46 seconds quicker than #2 and was around 5% quicker than many of the slowest riders.

Given that the World TT represents the top 0.000001% of the world's cyclists, how - statistically - can there be such a massive gap between #1 and #2. And how can #1 be 5% faster than #100? It's really off-the-curve.

Or am I missing something here...?

The same goes for tennis, incidentally. Statistically, it amazes me that the top 4 or 5 in the world have such a gap between them and the rest, given how fine the margins are in professional sport.

I get it, but it confuses me massively.
«13

Comments

  • By the way, I'm not suggesting that there's any sort of foul play here at all. More that I'm surprised there can be such a big gap between the top 2 riders in a world championships. You would think that - given the margins - 5 seconds might be a big gap, but 46 seconds is just massive.
  • Mccaria
    Mccaria Posts: 869
    Don't have an answer to your main question, but I am not sure the gap between 1st and the slowest in a World Championship is the same as between 1st and 100th in the world, since there will be strength in depth in the top 5-7 countries who can only send 2 riders to a Championship.

    The differential in elapsed time of Wiggins to Martin was 1.2%, the differential between Bolt and Blake in the Olympic 100 metres sprint was 1.2% !! Perhaps greatness is being 1.2% faster than everyone else....
  • RichN95.
    RichN95. Posts: 27,137
    Statistically speaking, with a large group set the distribution between consecutive data points will be separated by larger margins at the extremes of the distribution bell curve. (These may not be the proper technical terms)

    For example, look at the tallest men ever measured - http://en.wikipedia.org/wiki/Tallest_people - the top twenty are separated by a foot - the top two by 1.4%. (Contrast to the middle million who will be separated by microns)
    Twitter: @RichN95
  • @barrybridges and mccaria - I've just gone crossed eyed reading your posts :)

    I guess the answer to your question is that regardless of statistics and the law of averages, some people (in all walks of life) will always be way better than their peers, whether through nature or nurture.

    It's a pretty simplistic way to look at things I know, but it works for me.
    Reporter: "What's your prediction for the fight?"
    Clubber Lang: "Prediction?"
    Reporter: "Yes. Prediction"
    Clubber Lang: "....Pain!!!"
  • Daz555
    Daz555 Posts: 3,976
    In reference to the fact that Tony Martin has just destroyed the field in the World TT champs, I have a question I've always wanted to ask.

    If you think of how many people in the world cycle on some kind of competitive or at least serious basis, it's a very very big number.

    Now if you rank all those people in order of their time-trialling ability, from best to worst, you'll have a long line of millions of people.

    I've read Outliers and I understand the concept, but what I don't get is why there could ever be such a big gap between #1 and #2 in the world on any given day.

    Tony Martin was 46 seconds quicker than #2 and was around 5% quicker than many of the slowest riders.

    Given that the World TT represents the top 0.000001% of the world's cyclists, how - statistically - can there be such a massive gap between #1 and #2.
    The additional problem is that an athlete's performance is not static. Looking at many big TTs you wonder why it seems rare to see lots of TTers arrive at the same contest absolutely amazing shape at the same time.

    Wiggo demolished the field in London 2012 by a similar margin to today so you could have asked exactly the same question last year of a different rider. Next year it might be someone else.
    You only need two tools: WD40 and Duck Tape.
    If it doesn't move and should, use the WD40.
    If it shouldn't move and does, use the tape.
  • ddraver
    ddraver Posts: 26,348
    Don't forget also that this was a course that Tony could have designed himself. If it had been hillier, Wiggo would have done comparatively better, if it had been super technical, Canc would have done comparatively better...
    We're in danger of confusing passion with incompetence
    - @ddraver
  • Pross
    Pross Posts: 40,218
    Given that the World TT represents the top 0.000001% of the world's cyclists, how - statistically - can there be such a massive gap between #1 and #2. And how can #1 be 5% faster than #100? It's really off-the-curve.

    I'm not convinced there are 100,000,000 competitive cyclists in the world :wink: Also, as Rich has pointed out the 100 riders don't constitute the best 100 in the world. As an example Froome on form would be top 10 then we have the likes of Mick Rogers, Contador etc. etc. who would all be better than most of those taking part. Someone like Hutch and even a plethora of top British club TTers would have been quicker than the slowest riders at the champs I suspect.
  • There is probably a lot of psychology at play here. The top few have a distorted talent because they have full confidence in their ability.

    If you look at the tennis players, there is often a small difference in number of points won - they just have the confidence and assurance to win the decisive points. Similarly Martin, Canc & Wiggo (when dry) have the confidence to take tighter lines than other riders (and go deeper energy-wise) - if you're only ever going to be 6th-10th, why would you take the risks just before the RR?

    Further, getting updates of being ahead via ear-piece would porobbaly give a very minor benefit to performance.
  • mpie
    mpie Posts: 81
    No, it makes sense. You are confusing statistics which apply to large groups (all competitive cyclists) with a specific situation with only two samples (Martin and Wiggins). Although the latter are part of the former group, you cannot assume that all the statistic in the former apply to the latter. In general, the fewer the samples the greater the deviations from the statistical norm - what you term 'the curve'.
  • Martin's winning margin is nothing like the margins Indurain won some of his TdF tts by though - 3 minutes in an hour long ride once.
  • nic_77
    nic_77 Posts: 929
    And how can #1 be 5% faster than #100?
    That one is probably easier to explain... once you are out of the hunt for a medal, like it or not your performance will drop by some degree (thus exaggerating the difference in ability).
  • In reference to the fact that Tony Martin has just destroyed the field in the World TT champs, I have a question I've always wanted to ask.

    If you think of how many people in the world cycle on some kind of competitive or at least serious basis, it's a very very big number.

    Now if you rank all those people in order of their time-trialling ability, from best to worst, you'll have a long line of millions of people.

    I've read Outliers and I understand the concept, but what I don't get is why there could ever be such a big gap between #1 and #2 in the world on any given day.

    Tony Martin was 46 seconds quicker than #2 and was around 5% quicker than many of the slowest riders.

    Given that the World TT represents the top 0.000001% of the world's cyclists, how - statistically - can there be such a massive gap between #1 and #2. And how can #1 be 5% faster than #100? It's really off-the-curve.

    Or am I missing something here...?

    The same goes for tennis, incidentally. Statistically, it amazes me that the top 4 or 5 in the world have such a gap between them and the rest, given how fine the margins are in professional sport.

    I get it, but it confuses me massively.

    the field at a cycle race probably isn't an ideal sample if you want to get into the statistics.

    I'd say its a form issue though - there is no way of really knowing what level of form these riders are at, Martin could have been running at 100% and Wiggins at 98% we just don't know.
  • Ed-tron
    Ed-tron Posts: 165
    Daz555 wrote:
    The additional problem is that an athlete's performance is not static. Looking at many big TTs you wonder why it seems rare to see lots of TTers arrive at the same contest absolutely amazing shape at the same time.

    Wiggo demolished the field in London 2012 by a similar margin to today so you could have asked exactly the same question last year of a different rider. Next year it might be someone else.

    Agreed ^ for me this is the main aspect.

    Through my reading of your post, this statistical curve approach assumes all three (Martin, Wiggins, Cancellera) are all at their maximum ever TT ability at the same given time, which just isn't the case.

    In reality, there are still other variables involved, which accounts for most of the 40 seconds I would have thought.

    Martin, Wiggins, and Cancellera have all had different seasons. Trained, raced and prepared differently. Some of this is intended, like using the Vuelta or not, and some of it is responding to changing circumstances, like Brad and his injury and stalled start to the season after losing weight to be a GC climber.

    Martin's season has been a big engine domestique, and the Worlds fits with this and is the goal of his season. But for Wiggins for example, he tried to be a GC climber, got injured, lost his bottle, was presented with the ToB, then throught he'd have a crack at the worlds TT, so has muddled through a lot more.

    As per above, perhaps Wiggins London 2012 TT was somewhere near his peak possible TT potential, but he wasn't riding today in Florence, and arguably it showed
  • turnerjohn
    turnerjohn Posts: 1,069
    not forgetting the fact that Wiggo beet him at the Olympics TT !!!!
    end of the day he had one of the best days riding of his life today...fair play to him
  • mfin
    mfin Posts: 6,729
    Don't forget the fact that Wiggo had to ride a shitty Pinarello
  • jawooga
    jawooga Posts: 530
    RichN95 wrote:
    Statistically speaking, with a large group set the distribution between consecutive data points will be separated by larger margins at the extremes of the distribution bell curve. (These may not be the proper technical terms)

    For example, look at the tallest men ever measured - http://en.wikipedia.org/wiki/Tallest_people - the top twenty are separated by a foot - the top two by 1.4%. (Contrast to the middle million who will be separated by microns)

    Very good post!
  • RichN95 wrote:
    Statistically speaking, with a large group set the distribution between consecutive data points will be separated by larger margins at the extremes of the distribution bell curve. (These may not be the proper technical terms)

    For example, look at the tallest men ever measured - http://en.wikipedia.org/wiki/Tallest_people - the top twenty are separated by a foot - the top two by 1.4%. (Contrast to the middle million who will be separated by microns)


    Of the top 10, 5 are from the USA. Why is that?
    “New York has the haircuts, London has the trousers, but Belfast has the reason!
  • RichN95.
    RichN95. Posts: 27,137
    RichN95 wrote:
    Statistically speaking, with a large group set the distribution between consecutive data points will be separated by larger margins at the extremes of the distribution bell curve. (These may not be the proper technical terms)

    For example, look at the tallest men ever measured - http://en.wikipedia.org/wiki/Tallest_people - the top twenty are separated by a foot - the top two by 1.4%. (Contrast to the middle million who will be separated by microns)


    Of the top 10, 5 are from the USA. Why is that?
    They probably did more measuring and kept better records in the olden days (all of those five are dead)
    Twitter: @RichN95
  • jawooga
    jawooga Posts: 530
    I remember my father once explaining an evolutionary theory:

    All human kind evolved from the African continent. Therefore, all of life's range of extremes as would be found in DNA distributed across the world would also be found in Africa. You would expect to find, the tallest, shortest, fastest, slowest, strongest, weakest etc. all within Africa - if considering nature only.

    He was sheepish in relaying the theory as though it was an un-pc book that he had read, but I would have thought it make sense. [He may have even been quoting Outliers, I've not read it :) ]
  • RichN95 wrote:
    RichN95 wrote:
    Statistically speaking, with a large group set the distribution between consecutive data points will be separated by larger margins at the extremes of the distribution bell curve. (These may not be the proper technical terms)

    For example, look at the tallest men ever measured - http://en.wikipedia.org/wiki/Tallest_people - the top twenty are separated by a foot - the top two by 1.4%. (Contrast to the middle million who will be separated by microns)


    Of the top 10, 5 are from the USA. Why is that?
    They probably did more measuring and kept better records in the olden days (all of those five are dead)

    I've often wondered why there's a similar bias in VO2 max records - many of the highest ever recorded figures are from Scandinavian athletes. Is some form of measurement bias the likely explanation there too?

    Regarding the difference between Wiggins and Martin, I'd agree with the posters who've argued that we aren't necessarily comparing like with like. Wiggins doesn't need to be particularly out of form or having a slight off day to be ~1% slower than Martin.
  • rick_chasey
    rick_chasey Posts: 72,245
    ddraver wrote:
    Don't forget also that this was a course that Tony could have designed himself. If it had been hillier, Wiggo would have done comparatively better, if it had been super technical, Canc would have done comparatively better...

    Canc likes a few hills thrown in too. He's said as much in the past. Think Bejing and Mendriso.

  • If you look at the tennis players, there is often a small difference in number of points won - they just have the confidence and assurance to win the decisive points.

    I quite like the fact that you can lose a tennis match despite winning more points.
  • Did we see all 3 at their best?

    EDIT

    No. Scratch that. Did we see all 3 in form?
    “New York has the haircuts, London has the trousers, but Belfast has the reason!
  • andyrac
    andyrac Posts: 1,123
    It might be a bit simplistic, but it's what Tony Martin bases his whole season around. Winning TT's, and especially the World Championship.
    Wiggins and Cancellara have other goals......
    All Road/ Gravel: tbcWinter: tbcMTB: tbcRoad: tbc"Look at the time...." "he's fallen like an old lady on a cruise ship..."
  • rick_chasey
    rick_chasey Posts: 72,245
    Did we see all 3 at their best?

    EDIT

    No. Scratch that. Did we see all 3 in form?

    One prepped specifically and started the Vuelta, one prepped specifically and didn't, and one didn't prep specifically but started the Vuelta..... :wink:


    So no.
  • le patron wrote:

    If you look at the tennis players, there is often a small difference in number of points won - they just have the confidence and assurance to win the decisive points.

    I quite like the fact that you can lose a tennis match despite winning more points.

    And also win a match despite winning fewer games (theoretically - according to Wiki this happened in the 2009 Wimbledon final).
  • rick_chasey
    rick_chasey Posts: 72,245
    gpreeves wrote:
    le patron wrote:

    If you look at the tennis players, there is often a small difference in number of points won - they just have the confidence and assurance to win the decisive points.

    I quite like the fact that you can lose a tennis match despite winning more points.

    And also win a match despite winning fewer games (theoretically - according to Wiki this happened in the 2009 Wimbledon final).

    Tennis scoring system is genius.

    No loser can ever claim they woz robbed.
  • Did we see all 3 at their best?

    EDIT

    No. Scratch that. Did we see all 3 in form?

    One prepped specifically and started the Vuelta, one prepped specifically and didn't, and one didn't prep specifically but started the Vuelta..... :wink:


    So no.

    oops wrong thread
    “New York has the haircuts, London has the trousers, but Belfast has the reason!
  • cesco
    cesco Posts: 252
    Let's some up some differences between Martin and the other medalist:
    - Martin rode 58/11; the others 56.
    - Remember his solo "training session" in the Vuelta?
    - Besides that, he has been focussing on the world cup since the Tour de France, or at least more than the others. Cancellara stayed in the Vuelta longer than intended, to help Horner. Who did Martin have to ride for? Exactly. I didn't bother with the ToB, but even when it's not the hardest race, it's still a race. Wiggins may still have had that one in the legs, and not in the good preparation kind of way.
    - Camelbacks are lame, but even when they are not allowed to have an aero design, perhaps they are quicker than aero-bottles like Cancellara's (didn't catch what Wiggins used).
    - Even though I just read he'd been moaning about it, I thought Cancellara didn't want too much information from the car.
    - This stage simple fit Martin best.

    This could easily account for say 5% of the difference. The rest is perfectly explained by this post:
    RichN95 wrote:
    Statistically speaking, with a large group set the distribution between consecutive data points will be separated by larger margins at the extremes of the distribution bell curve. (These may not be the proper technical terms)

    For example, look at the tallest men ever measured - http://en.wikipedia.org/wiki/Tallest_people - the top twenty are separated by a foot - the top two by 1.4%. (Contrast to the middle million who will be separated by microns)
  • Daz555
    Daz555 Posts: 3,976
    cesco wrote:
    - Camelbacks are lame, but even when they are not allowed to have an aero design, perhaps they are quicker than aero-bottles like Cancellara's (didn't catch what Wiggins used).
    Lame? How so?
    You only need two tools: WD40 and Duck Tape.
    If it doesn't move and should, use the WD40.
    If it shouldn't move and does, use the tape.