Yes, if seen stuff on exponentially weighted moving average, and it looks
like a reasonable way to go.
>From what I've seen though, they all deal with time decay.

What I want is time decay AND games decay to be taken into account. Any
ideas how to combine the two decay variables in the single average?

Cheers




"Simon, Steve, PhD" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> It's always hard to answer these questions without first asking "Why?"
> The approach that makes the most sense depends on what you are trying to
> achieve.
>
> Nevertheless, here's some basic advice. First, if you don't know
> anything about filtering, you should probably pick up some basics. I did
> a web search on exponentially weighted moving average and the first site
> on the list looked quite good:
>
> Dealing with Measurement Noise (A Gentle Introduction to Noise
> Filtering) M.T. Tham, http://lorien.ncl.ac.uk/ming/filter/filter.htm
>
> Second, you have time measured two different ways (D=number of days ago
> and G=number of games ago). It seems simplest just to add the two
> numbers together T=D+G. So a performance 5 days ago and 2 games ago
> would give you a value of T=7. Perhaps you might want to weight things
> differently. In the case of weekly sporting events, T=D/7+G might make
> more sense.
>
> There's a lot of arbitrariness in your choices for the relative
> importance of D vs G and how quickly the weights decay to zero. Unless
> your results were used for something very critical like litigation or
> salary negotiations, I wouldn't waste a lot of time trying to compare
> different approaches. Most reasonable approaches will come up with
> roughly the same answer.
>
> Steve Simon, [EMAIL PROTECTED], Standard Disclaimer.
> The STATS web page has moved to
> http://www.childrens-mercy.org/stats.
>
> .
> .
> =================================================================
> Instructions for joining and leaving this list, remarks about the
> problem of INAPPROPRIATE MESSAGES, and archives are available at:
> .                  http://jse.stat.ncsu.edu/                    .
> =================================================================


.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to