Percentage of World Records

Ben Rea · Post by **Ben Rea** » March 20th, 2006, 4:30 pm

95.3 what, i dont get it.

johnlvs2run · Post by **johnlvs2run** » March 20th, 2006, 4:35 pm

The total of 857.8 for the 9 events, divided by 9, is an average of 95.3 for the 9 events.

PaulH · Post by **PaulH** » March 20th, 2006, 4:55 pm

hjs wrote: The Nonathlon is a nice thing but if I may criticise it a bit. It seems to feavor the older athletes. High up in rankings are mostly older people to find. It can be that younger people don,t rank there times but If that is enough to explain it?

It does tend to favor older athletes - there are fewer people contesting each event on the online ranking the older the event gets, so the times are less likely to have been pushed 'to the limit'. So the raw data I'm using has an inherent bias in their favor. I don't think of that as a problem however - I hope the Nonathlon encourages older people to try more events, which helps to make those records stronger and hence improve the Nonathlon's scoring!

PaulH · Post by **PaulH** » March 20th, 2006, 5:09 pm

John Rupp wrote:I like the Nonathlon. It's a very nice system and arrangement. However, the scores are overinflated, and are not equitable across age, weight class, and gender. Some of the scores are way too high and others are too low.

Another technical note: the Nonathlon scores *can't* be overinflated, because the scores are relative measures, not absolutes.

Incidentally, you used the word 'equitable'. I don't claim that the Nonathlon is equitable, but it does try to aim for that. That explains, for example, why I don't use any arbitrary 'adjustment factors' when calculating scores - I get Excel to draw a least-squares quadratic curve through the C2 ranking data, then use that formula to calculate the Nonathlon scores. Now the data is variable enough that this can't really give equitable scores, and to the degree that it doesn't it tends to favor older rowers, and to a lesser extent lightweights. But with each year that problem is reduced because the number of decisions I impose on the data is kept to an absolute minimum*

Cheers, Paul

*Specifically I will, very rarely, discard a logged time if it is wildly inconsistent with similar results in that event (i.e. a typo). The other assumption I make is where the useful data range is; in that case my rule is simply that the age range used to calculate the curve runs from the youngest age to have logged all 10 events up to the oldest to have done so.

johnlvs2run · Post by **johnlvs2run** » March 20th, 2006, 5:11 pm

Dean Smith's 7:29.3 at age 79, is right at the top of the World Record curve that I have calculated for the Perathlon scores.

PaulH · Post by **PaulH** » March 20th, 2006, 5:19 pm

John Rupp wrote:Paul,

Are you saying Dean Smith's 7:29.3 at age 79, is a weak record compared to some of the younger ones.

His time is right at the top of the World Record curve that I calculated for the PERathlon scores.

No I'm not - I said specifically that "the times are less likely to have been pushed". It only takes one person in a category to set an awesome standard, but in general the more people there are in a category, the more likely you are to find someone able to set that standard. Clearly in Mr Smith's case we found that one person out of a relatively small pool.

The fact that he is at the top of your World Record curve is not proof, however, that his is a particularly 'strong' record, except perhaps in a trivial sense.

hjs · Post by **hjs** » March 20th, 2006, 5:43 pm

PaulH wrote:
hjs wrote: The Nonathlon is a nice thing but if I may criticise it a bit. It seems to feavor the older athletes. High up in rankings are mostly older people to find. It can be that younger people don,t rank there times but If that is enough to explain it?
It does tend to favor older athletes - there are fewer people contesting each event on the online ranking the older the event gets, so the times are less likely to have been pushed 'to the limit'. So the raw data I'm using has an inherent bias in their favor. I don't think of that as a problem however - I hope the Nonathlon encourages older people to try more events, which helps to make those records stronger and hence improve the Nonathlon's scoring!

True,

But is also keeps younger people away. Doesn,t matter but this way it is not something for the hole erger communety. But it's an understandeble choise.

Porkchop · Post by **Porkchop** » March 20th, 2006, 6:13 pm

Citroen wrote:
John Foy wrote:John, you need a perathlon calculator because the maths mess with my head.

John
John, I'll use one of my contacts at the European Centre for Medium Range Weather Forecasting. You clearly need a high-end Cray super computer (or something like the BlueGene Grid super computer) to begin to make any sense of the bizarre maths that JR is presenting here.

I've yet to see any evidence that these figure show anything or predict anything.

One thing missing is the fudge factor for height. (As we all know the creator of these bizarre stats is fixated on stroke length from his other musings on the two forums.)

On the other hand it may be better for you to spend the time you'd waste on calculating this stuff on rowing a decent 10K (or three) instead.

Hmm -- my youngest daughter's high school has a Cray (and a very good crew as well). Maybe one of the budding geniuses there would like to get in the action here.

PaulH · Post by **PaulH** » March 20th, 2006, 6:30 pm

hjs wrote:
True,

But is also keeps younger people away. Doesn,t matter but this way it is not something for the hole erger communety. But it's an understandeble choise.

I don't think it should keep younger people away - there's no reason a good 16 year old shouldn't be extremely competitive. It does tend to challenge those in very early middle age, however (around 27-35, roughly), but that's pretty close to inevitable I'm afraid.

whp4 · Post by **whp4** » March 20th, 2006, 6:40 pm

hjs wrote:
PaulH wrote:
hjs wrote: The Nonathlon is a nice thing but if I may criticise it a bit. It seems to feavor the older athletes. High up in rankings are mostly older people to find. It can be that younger people don,t rank there times but If that is enough to explain it?
It does tend to favor older athletes - there are fewer people contesting each event on the online ranking the older the event gets, so the times are less likely to have been pushed 'to the limit'. So the raw data I'm using has an inherent bias in their favor. I don't think of that as a problem however - I hope the Nonathlon encourages older people to try more events, which helps to make those records stronger and hence improve the Nonathlon's scoring!
True,

But is also keeps younger people away. Doesn,t matter but this way it is not something for the hole erger communety. But it's an understandeble choise.

Oh, come on - it doesn't keep anyone away, except those unwilling to take part in an activity where they can't be at the very top. If everyone felt that way, you wouldn't have many people behind you in the rankings! Think the nonathlon inappropriately gives high scores to rowers much older or lighter than you? Simply ignore them! Compare your scores with others of your age, gender, and weight class, or whatever arbitrary group you like. For that matter, compare your scores with your scores from past years - are you improving relative to the standard?

Bill

johnlvs2run · Post by **johnlvs2run** » March 20th, 2006, 7:16 pm

whp4 wrote:Compare your scores with others of your age, gender, and weight class

Bill, I like that idea too.

Bob S. · Post by **Bob S.** » March 20th, 2006, 8:18 pm

hjs wrote:
PaulH wrote:
hjs wrote:
True,

But is also keeps younger people away. Doesn,t matter but this way it is not something for the hole erger communety. But it's an understandeble choise.

There seems to be a general agreement that the system favors us codgers, but it was still neat to top Dwayne Adams recently. That was a pretty good incentive to keep at it.

Bob S.

whp4 · Post by **whp4** » March 20th, 2006, 8:19 pm

John Rupp wrote:
whp4 wrote:Compare your scores with others of your age, gender, and weight class
Bill, I like that idea too.

Yes, it's easy - just look at the nonathlon scores of other people who are about your size, shape, color, handedness and draw your own conclusions. I attribute my relative standing to a number of factors, most of which conspire (unfairly, I might add) against me:

1) older participants get more points for a given performance
2) lightweights get more points for a given performance
3) women get more points for a given performance
4) younger participants are younger and faster
5) people my age all seem to be hard-core rowers willing to train harder
6) many people have rowing backgrounds that go back years

It's a small miracle that I make it into the top 1,000 places!

I do my best to offset these things by using slides (though they make me no faster) and keeping my excess weight (still haven't figured out how to make the spare tire make me go faster - every time I lose weight by rowing a lot, I get faster, not slower!)

Bill

PaulS · Post by **PaulS** » March 20th, 2006, 8:29 pm

whp4 wrote: I do my best to offset these things by using slides (though they make me no faster) and keeping my excess weight (still haven't figured out how to make the spare tire make me go faster - every time I lose weight by rowing a lot, I get faster, not slower!)

Bill

Dang! You must jest be doing it ALL WRONG!!!!