What's your Eddington number ?

[MercuryKev]

Why?

Eddington had a simple system - count up the number exceeding a certain distance.  Reading the row numbers on a spreadsheet gives us that without any fancy calculations (well, other than the one converting my km into miles for the reference column).

Could. Not. Be Simpler.  Why complicate things unnecessarily?  Surely Eddington, as a scientist, would frown on anything beyond the simplest method required.

That is all the formulae are doing.  It's not fancy, just automated.

[phil d]

Since I've done more than 125 200km audaxes, without spending any time calculating, my Eddington number must be about 125

[LEE]

Since this is about miles completed in a day it's going to be bloody hard to get beyond 200 regardless of who you are.

Maybe a regular RAAM contestant may rack up 200 miles a day 200 times as part of training and RAAM itself but it gets increasingly hard to raise your number.

E=100 would be easy for me, my prep for PBP2015 will get me there.
I suppose, if I keep doing 200km Audaxes, I'll end up like Phil d with 125 but there's a significant drop off in the number of rides over 200km I've done.  Getting to E=125 would be fairly easy over the next couple of years but getting to E=130 would be a significant undertaking.

I reckon E=>150 is an exclusive club.

Best just to turn all your rides into 200 milers at an early age.

Given that this is about miles in a day rather than rides I reckon I'll break my 600s and PBP into separate days and claw back a few more E's

[Lightkeeper]

Plain Eddington is a lifetime measure, favouring those who've spent years in the saddle as well as those who cycle prodigiously in their youth. 

Having calculated plain E for getting on for 10 years or recorded rides, I'd like to calculate what it was at the end of each of those years; and what it would take to maintain it over a rolling period of n years.

The first would give me a satisfying feeling of progress; the second would give me an incentive to keep riding at least as much and as far in years to come as in my better years so far. That feels like a better target than chasing an ever-harder higher E (though it would be satisfying and achievable to get to E = 100 by the end of this year).

What I'm thinking, and the approach of  splitting 600s to rack up a few more Es, suggests that Goodhart's Law applies: as soon as you use a measure as a target, it starts to influence the behaviour it was intended to measure, with sometimes unintended consequences.

[saturn]

Best just to turn all your rides into 200 milers at an early age.

URL=http://en.m.wikipedia.org/wiki/Goodhart's_law]Goodhart's Law[/URL] applies: as soon as you use a measure as a target, it starts to influence the behaviour it was intended to measure, with sometimes unintended consequences.

Like thinking "I've only got time for E miles today, hardly worth going for a ride"

[Lightkeeper]

Having calculated plain E for getting on for 10 years or recorded rides, I'd like to calculate what it was at the end of each of those years; and what it would take to maintain it over a rolling period of n years.

Some sums later... 

• Oct 2011:  E = 93
• Oct 2010:  E = 84
• Oct 2009:  E = 68
• Earlier years:  unknown, as my individual rides list only captures rides > 100 km

If I do my planned rides this year, E will get to 100.  If I want to keep it at 100 on a rolling 10-year period, I need to do (obviously) an average of 10 century rides a year, so the equivalent of RRTY each year would do nicely.  A couple of years of that, and I'd get to an all-time E = 125 and then, because most of my long rides are 200 kms, probably not a lot further.
 

[Oaky]

Whilst we are discussing formulae etc... I'm sure there's a simpler bit of SQL, but the following works for me:-

Code: [Select]

sqlite3 -html bike.sq3 <<EOF
select cast (t.distance as integer) as eddington_number from times t where cast (t.distance as integer)<=(select count(x.distance) from times x where x.distance>=cast (t.distance as integer)) order by -t.distance limit 1;
EOF

[Ningishzidda]

I'm new here, but I reckon I've ridden 64 miles more than 64 times.
So 64 it is.

I don't think its going to get any bigger than that.


The E number would be a similar concept to the N number.
How many minutes one can keep up N mph. 26.2 anyone?

[Lightkeeper]

The E number would be a similar concept to the N number.
How many minutes one can keep up N mph. 26.2 anyone?

Oh, easily.

You did mean downhill in the Pyrenees with a tailwind, didn't you?

[Andrew]

I've only got around 6 or 7 years of rides recorded, but then I don't suppose I did anything prior to then to confuse the calculation anyway!

I'd guess my number is around 70 but if I remember next time I open up Cyclistats then I'll confirm that.

I like the K number stat too. I reckon mine would be around 5; 5 years of more than 5000km. Reckon I'd make it 6 this year, all going well.

[Oaky]

Mine is still firmly at 43, but I thought it would be interesting to see how it increased with time.  A little Python script later revealed the following (bah - just noticed the US date format on the x-axis .... #broken - fix later):-



My aim is to get it to 50 this year... 14 rides of >= 50 miles required for that though.

[Ningishzidda]

If one consistently rides 100 miles, the first ride will result in an E number of 1 due to the number of rides and not miles. The second ride will increase the E number to 2, due to two rides ridden.
This will continue until the 100th ride, when the E number will be 100, Bingo 100 rides of 100 miles.
Another 100 mile ride will result in the E number remaining at 100 due to distance, not number of rides.
A ride of 101 miles will now increase the E number to 101.
A ride of 125 miles will stabilise the E number at 101.
Another ride of 125 miles will increase the E number to 102.

The E number does not, a/ indicate total miles ridden. b/ indicate total number of rides ridden. c/ indicate the longest ride ridden. d/ indicate average miles per ride.

It is a statistical nonsense.

[jsabine]

...
The E number does not [do anything useful]

It is a statistical nonsense.

Your point, caller?

I didn't think it pretended to be anything other than a slightly silly bit of fun (and a demonstration that anyone who's got a high E number has cycled quite a lot).

[L CC]

...
The E number does not [do anything useful]

It is a statistical nonsense.

Your point, caller?

I didn't think it pretended to be anything other than a slightly silly bit of fun (and a demonstration that anyone who's got a high E number has cycled quite a lot of quite long rides).

[Lightkeeper]

...
The E number does not [do anything useful]

It is a statistical nonsense.

Your point, caller?

I didn't think it pretended to be anything other than a slightly silly bit of fun (and a demonstration that anyone who's got a high E number has cycled quite a lot).

I think it's interesting, and not silly.  It doesn't have to be useful.  Though, interestingly, it is:  as a time series, like Oaky's, it offers a summary of a long-distance cyclist's development, quite economically and in a way that complements the other summary measures Ningishzidda mentions. 

I'm not attracted to the equation "not useful = nonsense".  That would create a lot of nonsense.

[saturn]

I'm not attracted to the equation "not useful = nonsense".  That would create a lot of nonsense.

Agree but if one person finds it interesting and is encouraged to go for one more ride or ride for one more mile, how could it be either?

[Pedal Castro]

A ride of 101 miles will now increase the E number to 101.

No it won't, it will need another 100 rides of 101+ miles to hit E=101

[Ningishzidda]

A ride of 101 miles will now increase the E number to 101.

No it won't, it will need another 100 rides of 101+ miles to hit E=101

True. That's what makes it stupid.

[marcusjb]

It's a really interesting measurement. 

I am not a recorder of miles - until this year when I will give it a go.  (I do have records of each audax I have ridden though).

Since I started audax (late 2010), I have around 47 days of 100 plus miles.  Prior to that, I'd only have half a dozen or so 100 mile days.

So - without going into too much maths and guessing (as there's quite a few 100km rides) - my number would be around the low 50s.

I'll just revisit the whole equation in a couple of years when all being well, I'll be able to prove 100 days of 100 miles.  I suspect that like most audaxers, I will, sometime in the next few years, hit 125 and there it will stay.

[clarion]

If you keep doing the same ride, you'll hit a ceiling.  If you vary your rides, your progress is likely to continue for longer.  Eddington was mainly interested in having a bit of fun playing with numbers.

[marcusjb]

If you keep doing the same ride, you'll hit a ceiling.  If you vary your rides, your progress is likely to continue for longer.  Eddington was mainly interested in having a bit of fun playing with numbers.

That's the challenge for audaxers, as discussed upthread, 125 will be where it ends for most of us.

[Ningishzidda]

If you keep doing the same ride, you'll hit a ceiling.  If you vary your rides, your progress is likely to continue for longer.  Eddington was mainly interested in having a bit of fun playing with numbers.

That's the challenge for audaxers, as discussed upthread, 125 will be where it ends for most of us.

Most 200 Randos are over 125 miles. When you ride your 126th 200 Rando, you might likely get a 126 E.
Will it be dependent upon the shortest 200 Rando you ride?
Bad luck if you rode a 200 Rando that was bang on 200 km 

[marcusjb]

If you keep doing the same ride, you'll hit a ceiling.  If you vary your rides, your progress is likely to continue for longer.  Eddington was mainly interested in having a bit of fun playing with numbers.

That's the challenge for audaxers, as discussed upthread, 125 will be where it ends for most of us.

Most 200 Randos are over 125 miles. When you ride your 126th 200 Rando, you might likely get a 126 E.
Will it be dependent upon the shortest 200 Rando you ride?
Bad luck if you rode a 200 Rando that was bang on 200 km 

Yes.

[CrazyEnglishTriathlete]

The typical 200k Audax that I've done has ended up at around 130 miles - DIY perms are usually somewhat overdistance as are most of the calendar rides.  It also helps to ride to and from the start of local Audax rides (my Kennett Valley Runs have ended up at about 150) - so I'm expecting my E to get stuck in the mid-130s...Eventually... if I keep to my current riding patterns.  Much higher than 160 seems inconceivable unless you Teethgrinder.

[clarion]

Does anyone know what Eddington's Eddington Number was?