Rankings: What's the base
Tim Tyler
We all know it gets harder to improve your rating as you get better -
but how much harder? How quickly does the rating slow-down progress.
Have there been any experiments to estimate the base value?
Long ago, I was told the base was 2. That doesn't seem right.
--
__________
|im |yler http://timtyler.org/ t...@tt1lock.org Remove lock to reply.
Ben Finney
> If your kyu/dan ranking is the exponent, what's the base?
What's your basis for the hypothesis that one's rank is an exponent?
> We all know it gets harder to improve your rating as you get better
> - but how much harder? How quickly does the rating slow-down
> progress.
In what units would you want such progress measured?
> Have there been any experiments to estimate the base value?
First, I'd like to see definitions of the terms the hypothesis is
using.
--
\ “Computers are useless. They can only give you answers.” —Pablo |
`\ Picasso |
_o__) |
Ben Finney
Tim Tyler
> Tim Tyler <seem...@googlemail.com> writes:
>> If your kyu/dan ranking is the exponent, what's the base?
>
> What's your basis for the hypothesis that one's rank is an exponent?
No firm basis. If there are better mathematical models, then
I'm up for those too. I'm interested in what is known about the
issue basically.
>> We all know it gets harder to improve your rating as you get better
>> - but how much harder? How quickly does the rating slow-down
>> progress.
>
> In what units would you want such progress measured?
Well, empirically, if some student spent so-many-hours per week
each week learning go they would get better at a decreasing
rate (in terms of Kyu per week). So, I'm mostly thinking
in terms of rating/time spent. However, if there are better
ways of measuring the rate of progress in go, hit me with them.
>> Have there been any experiments to estimate the base value?
>
> First, I'd like to see definitions of the terms the hypothesis is
> using.
Well, I don't want to constrain respondents too much. The
basic issue I am interested in is how best to model the
slow-down of rating improvement with progress.
Aladdin & The Gangs
news:bpRFk.8512$fU3....@newsfe19.ams2...
> If your kyu/dan ranking is the exponent, what's the base?
>
> We all know it gets harder to improve your rating as you get better -
> but how much harder? How quickly does the rating slow-down progress.
>
> Have there been any experiments to estimate the base value?
In my view, it's not the base value that determined how quickly does the rating
slow-down - the base "a" to exponent "x" gives you the false impression. But if
you take the progress as oscillating system then the damping-ratio (zeta) plays
the critical role as potential indicator for the plateau ahead (the Q factor if
you preferred).
I think different players are governed by their own unique damping-ratios
signature over the course in ranking establishment. It would be interesting to
see if a common signature implied the same needs as to progress forward - would
either a moyo or territorial oriented player choose the same progessing needs,
etc? Which would progess further first? ...
>
> Long ago, I was told the base was 2. That doesn't seem right.
It wasn't me. LOL