PDA

View Full Version : Neural Filters


Stix
27th May 2009, 01:35 PM
Hi All

Just wondering what filters you have or are using in your neural systems?

Why, am I asking this? Well, I'm looking at one currently and it's proving to be promising and that is with little or no filters.

I've ran a few over the results so far included in the below list, but a lot are one's I've thought about but not sure if its worth the time looking at:

No Maiden Races
No 2yo's
Apprentice/Senior rider
Must of placed in last two starts
Must not be resuming
Is it a 100 rater?
> $3 at the jump (or near enough) i.e. no even money munchers
Distance filters
Place % filters
Track Condition filters
Points/% between top two rated
Are ther any that you have found to be uselfullin your own analysis?

Cheers and thanks in advance

michaelg
28th May 2009, 06:33 AM
Stix, I could be wrong but I have doubts about the accuracy of the CF and Tim categories for horses having their second run from a spell.

Your statement of points/% between top 2 rated :- the CF and TIM categories can often be puzzling in their allocation of points. Sometims the extreme generosity of TIM to a certain horse against all others in today's race could easily be interpreted that it must have set a world record at its last start/s and is now a moral to win - but it often does not happen.

Stix
28th May 2009, 07:59 AM
Stix, I could be wrong but I have doubts about the accuracy of the CF and Tim categories for horses having their second run from a spell.

Your statement of points/% between top 2 rated :- the CF and TIM categories can often be puzzling in their allocation of points. Sometims the extreme generosity of TIM to a certain horse against all others in today's race could easily be interpreted that it must have set a world record at its last start/s and is now a moral to win - but it often does not happen.hi michaelg
I have TIM = 0 for the same rason you have stated, it's to volatile. Where CF & CP are extreme, they can often be "normalised" by adjusting back to the points awarded to the 2-4th rated. This often helps with false selections.

michaelg
28th May 2009, 08:35 AM
Does anyone know why such extremely high points in these categories are allocated, and how they are calculated?

They are so out of proportion when compared with the top points given in all the other categories that it is surprising no one at the R & S website has, to my knowledge, explained or even commented on this.

I can only guess that the neurals are saying that these 3 categories are by far the most important ones. If this is the case, surely it is extreme in most instances to allocate almost three times more points (can be interpreted as three times more important) than JA, Dist, etc.