Star the Grave Robber
 

Well, after sacking the Neurals as a starting point my selections have gone all over the place, only small numbers, but I was not happy and felt something wasn't right.

I was happy enough with my basic selection method, and if I win or loose with it so be it, because I feel at ease with it mainly because it is based on about twenty years of observation. No proof, but it reflects my philosophy in sports and racing.

What threw me was the inconsistencies in the Neurals , the shifting of the sands made me wonder what was happening.

Well. thanks to something ( garyf ) said made me have a rethink causing me to change my settings to what he suggested plus I added two others that in my opinion could not change the order dramatically, if at all. The rest I set at 0.

Things like Scratchings, Wet tracks, jockeys and maybe a few others could change the values sufficient to sway the final figures and the order.

Although I still cannot see if a scratching can affect the order , but that is for another day.

Well, after a few days of rubbish' changing back with less volatile filters, still threw up the same type of horses which was the Pre Selection method I wanted. I have watched over the Neurals for the last three or so days and the consistency is there now.

To make sure, I write down the original order of the Neurals in the selected races and only refer to that in the final elimination.

However, even checking the new Neurals, I have not seen the volatility like what I saw last week.

I suppose, I was probably lucky but I have had a few winners and my balance has jumped up to a more healthier position. it was looking very poorly for a few days.

So, I have to thank Gary and Angry Pixie and others for their Professional advice.

I was going into Information Overload, " Paralysis by Analysis '

Star

What grave have you robbed?

See my R.I.P. thread, I have recovered the body. It's a bit like Frankeinstein on Life Support.

But if I can kick start it, it might still have some life in it yet.

Star

Ha, ha.

That bought a smile to my face.

:eek:

Hi Star,

It depends on how they are developed, my ratings can sometimes change quite a bit after scratchings as I rate using a ranking method so if a horse is scratched the ranking order will change and this can have an affect on 1 or more of the factors used to rate. Sometimes its not just as simple as moving horses up in the ratings order with a scratching.

Cheers,

Mat.

Mat.

That seems to be what others have said too. But its too complicated for me to understand.

While it must be true, I still fail to see if Horses are ranked 1, 2, 3, & 4 and Horse two is scratched that regardless of the factors built in ( unless it involves some head on head data ) that Horse 3 can go anywhere else but to two and Horse four to three.

I have difficulty seeing how, Horse four can jump over Horse three and go to two if the original Horse two is scratched in some circumstances. But it must happen .

That is why I am now going by the original Neural ratings, writing them down and just moving them up the order to take the scratchings place.

It looks to complicated for this Bonehead but I yield to greater and more experienced guys on here.

Star

I understand where you are coming from, 2 examples of how a horse can leapfrog up or down could be barrier and pace. After scratchings I adjust barriers and this can have enough effect on my ratings to cause a horse to jump more than one spot.

Also pace plays a role in my ratings so if for example there were 2 recognised leaders in the race and one of them were scratched (say it was an outsider but one that could put some pace in the race), this could open up the race for the other leader to get a soft run in front and could push its ratings up while also pushing a highly rated backmarker down a bit.

Cheers,

Mat.

Thanks Mat.

I can now see that it is possible in the Neurals too. I had barriers as one of the Factors, Jockeys also.

Now I have taken them out to get some solidness in there. The variable things I am now trying to do manually.

I do not have the sophistiction to put too many factors in. Thanks for explaining in terms I think I understand. Others have tried but I am not sure I understood what they were saying.

Star

Drop me an email Star @ stix_hotcopper@aanet.com.au

I'll share what I have found after studying Neurals for the last few years.

I know certain tracks just do not perform in neurals, as well as certain other factors....hopefully help out a fellow Neural tragic :)