Update on Michaelg's neural system

[lomaca]

here are the results from 27/07/09 to yesterday

JA: S/R 45.86% POT 4.5% Max outs 6
CRS: S/R 45.91% POT 3.3% Max outs 9
CP: S/R 43.84% POT 2.0% Max outs 10

These are at Supertab prices and since mostly under $5.00, BFair prices would probably increase the profit.

Not bad going for a simple system.

Note, every item in neurals set to max.

Good luck


PS OOOps must have 270 points !

[michaelg]

Thanks for the update, lomaca. The stike rates are very impressive.

Unfortunately I haven't been following the system but after the results I will begin to do so, most likely with Betfair, maybe even with the TAB.

Hope you've been betting them.

Just out of interest, have you recorded any results for Place betting?

[lomaca]

Thanks for the update, lomaca. The stike rates are very impressive.

Unfortunately I haven't been following the system but after the results I will begin to do so, most likely with Betfair, maybe even with the TAB.

Hope you've been betting them.

Just out of interest, have you recorded any results for Place betting?Yes I have the results and will run it for the place as well, the strike rate is good but the profit is a bit skinny, the highest divi was $4.10, most of them are around the even or in the 2.5 to 3.3 range, that's why I thought Betfair might be a go.
longest out is a worry but by the same token, there were six winners in a row too!
Due to the S/R I think they would be suited for the multiples maybe?
Will post the results for the place, maybe tomorrow.

PS I restrict betting to less than $5.9 so there may have been higher divis but I found the range below $5.9 to be the best.

PPS, These are all for non Metro races!!!!!

[michaelg]

Thanks, lomaca.

According to my records not only did I delete Metro but I may have also done so with Maidens. And I also looked at the outright highest in the TIM category.

[lomaca]

Thanks, lomaca.

And I also looked at the outright highest in the TIM category.Yes I have the results for the tim as well but they are losing.

here is for the Trainer: S/R 38.57% POT -18.1%
TIM: S/R 40.47% POT -1.3%
JT: S/R 44.36% POT -2.7%
Time nearly breaking even.

Will have to Zip the other file it's too big!

[lomaca]

Thanks, lomaca.

According to my records not only did I delete Metro but I may have also done so with Maidens.So you did, ************, I'm re-running it now, should make an enormous difference, but much reduced activity. (less bets)

takes a while to run it even on my super duper fast machines due to the complexity of the interconnected databases.

[michaelg]

lomaca, there were 3 Maidens today, Toowoomba R4 no.1 is still to run. The other 2 Maidens produced a winner of $1.90 and $1.40 to Place, and the other ran second paying $1.50.

There have been 3 non-Maidens for two winners paying $1.50 and $1.20 to Place, and the other paid $1.90 and $1.20. The third (Hobart R5 no.6) ran second paying $1.00 - but it had only one race start which one could easily question the accuracy of its neural points.

Is it possible to also look at selections that have had only one race start - this might eliminate some losers, particularly in Maidens?

[lomaca]

lomaca, there were 3 Maidens today, Toowoomba R4 no.1 is still to run. The other 2 Maidens produced a winner of $1.90 and $1.40 to Place, and the other ran second paying $1.50.

There have been 3 non-Maidens for two winners paying $1.50 and $1.20 to Place, and the other paid $1.90 and $1.20. The third (Hobart R5 no.6) ran second paying $1.00 - but it had only one race start which one could easily question the accuracy of its neural points.

Is it possible to also look at selections that have had only one race start - this might eliminate some losers, particularly in Maidens?Sure, I can look at all that, I'm running the no maiden no Met less than $5.00 at the moment, tried to include all prices for the first run this arvo, but it's a waste of time.
I haven't run the total analysys for a long time because when I do my ratings in the morning I have the different criteria already built in for Met and maiden races and for the restricted RB races, and the programme picks them up and treats them accordingly.

Just finished the first criteria run which is the trainer, and not having the maidens included, the profit turned around from -18% to +2.6% and the S.R is 42%.
This time I'm running it according to race date so it might be easier to follow the result.

Email you when it's finished.

Good luck

[michaelg]

lomaca, there was one non-maiden selection today: Cranbourne R7 no.1. It won paying $3.10 and $1.40.

[lomaca]

lomaca, there was one non-maiden selection today: Cranbourne R7 no.1. It won paying $3.10 and $1.40.According to my rules I found four, all at CRA
CRA Race 1 Maiden 7 Cheilani Ntrck P$2.70

CRA Race 2 Maiden 4 Skidding Ntime No good

CRA Race 3 Maiden 6 Linton NJT won
CRA Race 3 Maiden 6 Linton NTrainP

CRA Race 7 Class 1 1 Nato Ntime won
CRA Race 7 Class 1 1 Nato Nallperform
CRA Race 7 Class 1 1 Nato Ntrck


Done all the tests but not the place yet, not a lot of change but and a big BUT Maidens 270 point and CRS GO GO GO!

Sort out my thoughts about it, and let you know. One thing I found, they must have changed the algorithm with the new website, definite fall away in some parameters as far as win% and profit goes, see how it performs in the future. The new feature, "Morning Line" on offer is a definite no-no, utter loser so far at least for the top sel.

Good luck

[michaelg]

lomaca, I too had those maidens, and also Kyneton R4 no.2 (maiden) which was top points only in the CP algorithym - it was unplaced.

Cranbourne R2 no.4 and R3 no.6 only had one previous start. The first selection was unplaced, and the second selection won paying $1.30 and $1.20.

[lomaca]

lomacaQuicker than email, yes can do!
what is the "default settings"?

Cheers

[michaelg]

Hi, lomaca.

The default settings, as I call them, are the settings automatically done by the neural website. That is, all on a setting of 3 except WT which is on zero, just as in the 270 points method. I've read here it is also referred as "factory settings."

There are 14 selections today. On Wed even the Metro selections showed an each-way profit.

Can you also look at the max points for each selection - there might be a worthwhile cut-off point just as in the 270 points method where any horse with a total below that number is automatically eliminated.

[lomaca]

Hi, lomaca.

The default settings, as I call them, are the settings automatically done by the neural website. That is, all on a setting of 3 except WT which is on zero, just as in the 270 points method. I've read here it is also referred as "factory settings."

There are 14 selections today. On Wed even the Metro selections showed an each-way profit.

Can you also look at the max points for each selection - there might be a worthwhile cut-off point just as in the 270 points method where any horse with a total below that number is automatically eliminated.Thanks.

I only have 4 selections for win and 2 for place.
(the win selections do not automatically qualify for place!!!)

Strictly following my results, by betting where there is a positive win% or place%, my brother in law, not a gambler, has turned a $100 starting bank into $896.00 in less than 2 months, starting off with 2%. He is on 15% now, with a S/R of over 45% you can do that! So, IT CAN BE DONE.

Good luck

[michaelg]

Your brother-in-law could be called Oz's answer to Maria.

Referring to the method I emailed you this a.m. - so far there's been 3 selections already raced: 2 winners (Pakenham R2 no.7 and Sun Coast R1 no.5) paying $2.70 and $8.20, and $1.40 and $2.40 for the Place. The third selection was unplaced. So far so good...

[lomaca]

Your brother-in-law could be called Oz's answer to Maria.
he is an accountant, a no nonsense numbers man, he wanted 6 months worth of data before doing anything, and he was right, because as you well know some of the early, promising algorithms have fallen by the wayside.

Also not being interested in jockeys, trainers, or horses in general, he is not swayed by opinion and gossip, plus level stakes only!!!

I will do your test tomorrow latest, and let you know, if it works, you may be swimming in money, good luck!

Cheers

[michaelg]

lomaca, if its not too late can you differentiate between metro and non metro tracks?

5 winners today for a return of $23.70, but a small loss on Place betting. Canterbury did not fare too well.

[lomaca]

lomaca, if its not too late can you differentiate between metro and non metro tracks?

5 winners today for a return of $23.70, but a small loss on Place betting. Canterbury did not fare too well.Sent!
Good luck

[lomaca]

Sent!
Good luckWell Michael, I think if you just stick to what you got so far, you couldn't go wrong.

My BiL just rang, out of the selections today, two early races paid off already and he is quitting for the day.
Good one at Moe race 2 Cheilani at $5.5 Supertab and a place somewhere else, I'm not betting today, so I'm sure where.

Good luck

Ps, keep looking but, I'm always looking for new angles myself!

[Bhagwan]

Does anyone know how to find the Neural set up, like the way it used to be ,where one could adjust the individual settings.settings?

[thorns]

Its still there Bhags,

Click on the neural gear cog for the meeting you want, than in the box that pops up, just under the field list for that race, there is a link to click to 'set prefernce scaling', you should be able to adjust them there.

[Bhagwan]

Thanks for that Thorns.

[lomaca]

Hi Michael, those number 1s still doing OK but today's results proves why I never bet on odds on favs.
When horses at $1.40 and $1.20 with good jocks on board fall over more often than not, what's the point backing them?
Still, if you backed all four qualifiers today (3 of them No.1), you are in profit.

Cheers

[michaelg]

I've been only backing them to Place and am showing a double-figure POT with the TAB.

And if they're very short I often back them on Betfair where the prices are superior to the TAB. The highest $1.04 TAB divvy I have obtained on Betfair is $1.29.

[lomaca]

I've been only backing them to Place and am showing a double-figure POT with the TAB.

And if they're very short I often back them on Betfair where the prices are superior to the TAB. The highest $1.04 TAB divvy I have obtained on Betfair is $1.29.Good stuff Michael, never thought of backing them for the place unless they specifically qualified for place only.
As to betfair it's a constant frustration for me, I can never get the amount matched, or miss out before they close betting.
What I found is that using fixed odds on Supertab, which is a bookie anyway, about five to ten minutes before jump gives you acceptable odds.

Good luck

[lomaca]

Good stuff Michael, never thought of backing them for the place unless they specifically qualified for place only.

Yes just checked yesterday's results, and from four selections two won for a return of $44.00, but all placed for a return of $48.50. (that being for $10.00 bets)

makes you think of the value in place betting eh?

[aferw]

Michaelg,
just wondering how you are going with this?