CycloAndy said:
I suggest you do some statistical tests to rule out chance. You could do what's called a chi-squared test to compare the proportion of top-500 riders in 2013 aged 27/28 (i.e born in 85/86) to the proportion of top-500 riders in any other year who were aged 27/28 at that time. Or you could do a t-test to compare the mean cq ranking in 2013 of top-500 riders aged 27/28 to the mean Cq ranking in other years of top-500 riders who were aged 27/28 at that time. Only then can you rule out random variation and in anyway suggest that you have evidence that the 85/86 generation are 'golden'. My guess is that you are reading too much into a minor blip that occurred mainly through chance/random variation. Technically you should also include those born in 87 as that was your original hypothesis and otherwise you're just picking an choosing data to 'prove' your hypothesis. My PhD involves a lot of stats and while I'm no expert, as there is probably more complex analysis that can be done, I don't like seeing stats being used incorrectly. Hope that helps and I hope you get the answer you were looking for!
Yes, you are right,
last year someone did a similar analysis, not like this, but similar, and 87 was weaker than 85, 86, and similar than others, that way I say is good but not deserve to be treated as golden.
But I would be interesting if I do the same for 2012 or 2011...or even for 2000 to see if the peak was as well at 27 age.
I will do it if people is interested, or any other people could do it as well...
Two years ago, I did a similar study, but gathering for generations, In my old forum, golden were 85,86, 87,... Venus 88, 89 and 90, and platinum dudes: 80,81, 82 (someone said that platinum were clearly better, and put a list of big names: Cancellara, Boonen, Contador, Gilbert, Wiggins, Cunego, Greipel,...)
I put together 84 and 83 and always did the caculation with them: x 3/2
I gather as well more years with colours... the brown generation and like that...
And the caculation was kinda complicated...it was count the number or riders of every sheet in CQ (100 every sheet), the forst 500 as well, so 5 sheets, and multiply bya factor... first 100 x5, next x4. and so on...
I painted a gauss bell as well...
And the conclusion was that golden were better than 84,83, and even little bit better to platinum being younger.. with 24-25 years old you are not as with 29-30.
They peaked as well then, and we will see if they will peak next years...
But I was wrong gathering by generations, becouse 84 is not a bad year as 83 is, is just a normal year, and 87 is not so prolific as 85 and 86, even another year the difference could be less...and for 87 there is still more improvement margin, but this golden generation has been always prominent, the same as 1990 from the begining...
So it is better make it year by year and later conclude if there are some years that could gather