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.