PDA

View Full Version : neurals


hermes
23rd June 2002, 08:17 AM
Can anyone enlighten me as to the value or otherwise of neural assessments. I gather its just a fancy way of adding things up. What sort of success rate does neural assessment have, as opposed to more traditional ways of assessing form? There are some great neural services online. Anyone with experience of neurals?

Fryingpan
23rd June 2002, 08:48 AM
Hermes

I use it as a tool. You might have read posts about getting the top 4 or 5 prize money horses in the field. Well it's easy to do this by clicking the $ factor to say 5 and the rest to 0 and the whole field is listed in prizemoney order. Much easier than working it out on paper, unless you use the Sportsman or similar.

23rd June 2002, 12:03 PM
There are flaws in neurals. Unless you have seem the formulas - how can you place any confidence in them?

Example
Mistegic was given a low CF rating (Current Form) despite the fact that it had run 3 placings in group 1s in its past 3 starts. The argument from the people behind the neurals was that it was a 3yo carrying 56kgs against open class horses???

Isn't that class factor? - not current form. In the 3 group 1s it was racing against open class horses, so the only reasoning could be it rising in weight by about 3kgs???

I would have thought that weight rises were a separate classification from current form?
If the horse was rising 3kgs that indicates it is well placed in relation to class, otherwise the handicapper wouldn't have given that weight. Explain this?

Mistegic comes out and wins.

That's just one example - there are plenty of others.

BP - barrier position stats are also not up to date or not logical. Recent example was when a horse in barrier 6 was given a BP rating of 30 (on setting 3) while horses in barriers 5 & 7 were given 0! Where's the logic in that?

Using Algorithms to select horses has many flaws - i would stay away from them unless you know the formulas and hence can work out if they have any factual basis behind them.

hermes
23rd June 2002, 12:22 PM
Thanks. I see both points of view here. On the one hand u can use the online neural forms to rank horses pretty quickly, but on the other, what are the formulae they use? And I agree Chief you can get some odd results. I tried the barrier neurals, setting all other factors to 'No'. Bizarre results. Can't see any logic in them at all, except the middle of the field seems higher rated. So just that must cast doubts on the validity of the overall results.