Neural settings

[rails run]

It's been a while since I looked at neurals and my pc holds the last setting as default.
Can you remind me? Are the default settings '3' and the market value default 114?
My questions get worse... is a strong factor '1' or '5'? i.e, if you want course suitabily to be an important rating do you set it at '1' or '5'?
Sorry for such lame questions but I can't remember for the life of me!

[Bhagwan]

1 is low
5 is high.

Default is 3
Except Wt- 0 (Wet Track)

114% default market setting.

Cheers.

[UselessBettor]

Bhagwan is right.

[rails run]

Thankyou kind gents

[lomaca]

Just in case anyone interested,

For the month of January this year the following result on the various Neural ratings I actually look at.
All settings are on 3!

Dsct baserate 1st selection Win loss $78.42 & Place loss $228.50
Morning line 2nd selection Win loss $106.20 & Place loss $145.36
Morning line 1st selection Win loss $137 & Place loss $169
Neurals 1st selection Win loss $143 & 196
Dsct baserate 3rd selection Win loss $221 & Place loss $178

Every other selection is over $220.00

All races all tracks.
Morning line is not always available but it's minimal.

Hong Kong races included, but I strongly advise against betting there unless you know more then Racing&Sports ratings gurus.

All in all betting on all races I don't think it's a bad result, I leave it to you find the races-tracks that are profitable.

Sorry I'm too busy to do any requests.

[lomaca]

Just in case anyone interested,

For the month of January this year the following result on the various Neural ratings I actually look at.
All settings are on 3!

Neurals 1st selection Win loss $143 & 196

Ps. by Neurals first selection I used CP!

The maximum draw down on the Morning Line second selection was the final result. Almost every time during the month it was hovering around
a profit.
Since I used an alphabetical order, not date, it's easy to work out which tracks were losers.

Good luck

[Bhagwan]

Thats Iomaca, for those interesting findings.

Can you say over how many races this covers ?

[lomaca]

Thats Iomaca, for those interesting findings.

Can you say over how many races this covers ?1659 and Vic Tab divies.

[lomaca]

1659 and Vic Tab divies.PPs.
I just ran January races again with the setting CF for the Neurals and DSCT rated price and the results are much worse.


Dsct Rated price 1st selection Win loss $170 & Place loss $152
Neurals 1st selection Win loss $274 & Place loss $249
Slight improvement on the Dcst third and second selections.

Dsct Rated price 3rd selection Win loss $144 & Place loss $107
Dsct Rated price 2nd selection Win loss $179 & Place loss $139

Make out of it what you wish or ignore it.
Good luck

[lomaca]

If so, here is a "fun" Neurals system.
Top rated in distance, and distance points >= 12, starting price on Supertab (important) >= $10.00
That's it.

1659 races POT $185.00 which is about 11%. that is for this January.

Be prepared for long outs, but once in profit for over about $40 then it still has the number of outs but never goes negative.

Good luck

[rails run]

Hi lomaca. I'm interested! When you say you review data alphabetically, can this be done on that website? Can you explain how l can do this please?

[lomaca]

Hi lomaca. I'm interested! When you say you review data alphabetically, can this be done on that website? Can you explain how l can do this please? rails run
Sorry, this is my own programme, if you refer to UB's site I don't use it so I wouldn't know.

When I say I do races in an alphabetical order, it's just a custom of mine, I could just as easily use date or race number order.

With the Neural distance I called it a "Fun" system not unlike Partypooper's PPM.

The outs ore fairly long but divies range from $1.80 like today to $90.00 or more. I found the range of $7 + profitable but setting the lower limit to $10 the POT goes up from $78 to about $190.

Haven't printed out the individual race results or set a variable to count them, I don't know the exact number of outs.

Just by glancing at the screen when it's ticking over I guess that betting $1.00 you would need about a $100 bank to be absolutely safe. (whatever that means in this game?)

This research is just a byproduct of running my own rating's analysis since it runs concurrently with my ratings.

I collect a lot of data but only out of academic interest I don't use them for my own betting.

Because I make my living out of punting, I have an almost pathological fair of relying on someone else's tips or rating, because they can and often do change them on a whim, and where does it leave me then?

Today I had a $1.8 $7.5, $11.30 and a $4.90 winners, the last one was Gosford. race 8.
Only bet the two over $7

Missed out on Kilmore being out most of the day, and Te Teko was not done by R&S.

If you take the chance remember it was only one month, and don't even think of betting for the place. The win return of $185 had a minus $120 for the place, this is quite normal with low strike rate high win dividend systems.

Good luck

[Bhagwan]

Hi Imaca,
I believe its a very strong idea to do your own selection method because I feel in the long run , it pans out better than just following say an individual tipster for the reasons you mention.

Well done on those results.

------------------------------
One method of selection that can work is to target the 2nd+ 3rd + 4th Fav in live market where the Fav is not odds-on then use your favourite method of reducing it to one selection.

Maybe use something like your favourite Neural setting , other than the std setting , to work out the strongest contender of the 3 horses.

In races with 10 & less runners.

At least you know that you are targeting some good prices and not the fav all the time.


---------------------------

If you cant be bothered doing that then try this approach...

Betting 3 Horses a race.
Bet 3 + 2 + 1 = $6 per race.
On the 2nd + 3rd + 4th Favs

Fields 5-11 runners.

Just make sure that the 2nd fav is 4.00+

And the 1stFav is 2.20+ or maybe 2.60 or 3.00 or 3.20+ (Optional) have a play around with this setting rule

You shouldn't experience too many runs of outs.
Use $100 bank for this exercise.
That allows for 16 outs in a row.

STOP for the day once $6 profit is made. (Stopping is the hard bit)
We dont want to give it all back again once we have got it.
That's 6% on bank.

If all goes to plan that's 42% a week.

[lomaca]

Betting 3 Horses a race.
Bet 3 + 2 + 1 = $6 per race.
On the 2nd + 3rd + 4th Favs

Fields 5-11 runners.

Just make sure that the 2nd fav is 4.00+

And the 1stFav is 2.20+ or maybe 2.60 or 3.00 or 3.20+ (Optional) have a play around with this setting rule

You shouldn't experience too many runs of outs.
Use $100 bank for this exercise.
That allows for 16 outs in a row.

STOP for the day once $6 profit is made. (Stopping is the hard bit)
We dont want to give it all back again once we have got it.
That's 6% on bank.

If all goes to plan that's 42% a week.I'm testing something like this for a friend, but with his settings we drew a blank.
I will check your idea when I have a chance.

Thanks

[rails run]

rails run
Sorry, this is my own programme, if you refer to UB's site I don't use it so I wouldn't know.Thanks lomaca. I thought perhaps this was a 'back data' feature on the R&S site I hadn't yet discovered.

[The Ocho]

Hi Imaca,
I believe its a very strong idea to do your own selection method because I feel in the long run , it pans out better than just following say an individual tipster for the reasons you mention.

Well done on those results.

------------------------------
One method of selection that can work is to target the 2nd+ 3rd + 4th Fav in live market where the Fav is not odds-on then use your favourite method of reducing it to one selection.

Maybe use something like your favourite Neural setting , other than the std setting , to work out the strongest contender of the 3 horses.

In races with 10 & less runners.

At least you know that you are targeting some good prices and not the fav all the time.


---------------------------

If you cant be bothered doing that then try this approach...

Betting 3 Horses a race.
Bet 3 + 2 + 1 = $6 per race.
On the 2nd + 3rd + 4th Favs

Fields 5-11 runners.

Just make sure that the 2nd fav is 4.00+

And the 1stFav is 2.20+ or maybe 2.60 or 3.00 or 3.20+ (Optional) have a play around with this setting rule

You shouldn't experience too many runs of outs.
Use $100 bank for this exercise.
That allows for 16 outs in a row.

STOP for the day once $6 profit is made. (Stopping is the hard bit)
We dont want to give it all back again once we have got it.
That's 6% on bank.

If all goes to plan that's 42% a week.
Hi Bhagwan. What's the stop loss on that last one? The profit stop is $6 but what would the stop loss be with a $100 bank (as you say).

[The Ocho]

Actually Bhagwan. I just ran that over a number of different days at random (stopping at +$6) and it didn't turn out too good at all. :(

[Bhagwan]

Stop loss can be approx 30%
or have a bank of 200.00 split half & half.
2 operating banks of 100 +100
At end of week , add 2 banks together.
Then split half & half again.
Keep doing this each week or month.

You smallest bet will be 1% of operating bank ($100) & bet accordingly on the others.

-----------------
Hi Ocho
If it did not pan out - It tells me different rules were used, what rules were you using , can you give us a list of the race numbers used & venue.

Here is a sample of that plan used ...

RULES I used.
5-11 runners
1st Fav 2.60+ (This is the option I chose to use)
2nd Fav 4.00+
Target 2nd-4th Favs
Bet 3+2+1 = 6 a race.


Result Mon 30th
UniTAB Divs used.
Goulb
.......Div.....O/L
R1. 4.00 x 3.00 = 12.00
R.4..8.20 x 1.00 = 16.40
R.5..5.30 x 2 = 10.60

Scone
R.6..6.70 x 3 = 20.10

Kilmore
R.1..4.40 x 3 = 13.20
R.4.. Loss -6.00

Total 5 wins from 6 races.

Ret 72.30 (UniTAB)
O/L 36.00
Prof +36.30
100% POT

This of course is only one day will vary day to day but it shows the potential on a good day.

There will be profitable & not so profitable days , just like all methods.

Please do your own due diligence , by going over past results before betting any real money.

[The Ocho]

Yes, you are right Bhagwan. It wasn't run exactly as you say but the first fav was over 2.20 although I'm not sure about the amount of runners.

In your example you have a profit of $36.30 however you did say to stop at +$6.

Just on a side note, I tried that last night on the first hour and a half of UK greyhound racing before I went to sleep :oops: (in SIM mode with my bot) and it made +$50. I'm not sure it will work every time though.

Maybe instead of stopping at +6 or -30 you have a trailing stop of -30? That way a bad day is a 30% loss of bank but on a good day the sky's the limit. In the UK greyhound example if it had of hit hard times after the +50 it still would of made +20 after stopping due to the -30 trailing stop being hit.

However, this goes against what you're saying about stopping at +6. I think any system where it would take 5 days just to get back to square following one loss is too long.

[kiwi]

Setting the default for all distances is not very effective.

I have three settings, one up to 1350m another for 1400m and another for 1600m+.

I find the neurals tend to run hot and cold profit wise and can't be relied on as an exclusive tool.

[Bhagwan]

Yes Ocho its another slant thats worth perusing .

Well done and thanks for sharing your findings.

Like all systems on the forum, they are not set in concrete and are there to be pulled apart and improved upon.

I can see it working well on the UK Dogs because there are only 6 runners & the first 4 tend to dominate proceedings.

Their market is not usually formed until approx 20 secs till jump.

There prices are usually pretty good for this exercise.
For instance its very rare one finds a 2nd fav under 4.00
The Favs for instance are nearly always over 3.00 ,try getting that on an Aust Dog races and many Horse races for that matter.

Cheers.

[lomaca]

Setting the default for all distances is not very effective.

I have three settings, one up to 1350m another for 1400m and another for 1600m+.

I find the neurals tend to run hot and cold profit wise and can't be relied on as an exclusive tool.Depends what you do, if you only change one factor then you are wasting your time. It will only change the "NR" value which is the total points added up, of course it will change the values in the changed column too but ranking will stay the same.

Being called "Neural" one would think that changing one factor would change all, but that is not the case.
I worked out how the individual setting are calculated and that is why I save ratings set at a uniform value and if I want to experiment I can change the setting.

Good luck

[The Ocho]

Bhagwan, I had a SIM loss of -$18.30 today using your criteria with 5 races and only one winning race.

There was a race there where the 2nd fav was picked 10 seconds out which turned into the first fave. Of course the first fave at the time (which ended second fave) won at $4.50. :rolleyes: Swings and roundabouts there I guess.

I'm not so sure about this one (on the Aussie gallops at least). I'll let it run on the trots and Oz dogs tonight and see how it goes and, of course, I will try the UK dogs again tonight (it may have just been a fluke though last night).

[The Ocho]

Just tried those settings with the night time dogs and trots and lost 6 in a row making a loss of -$21.64 (got to a high of $14.36). Not looking good. I will test the UK dogs again to see if last night was a fluke or not.

[Bhagwan]

So what your saying is it made a substantial profit one day & a small loss the next day & your in profit if both days added together 50-22= +28 profit
It don't work too good.
Hello! velcome to racing my friend.

Keep in mind that 1 unit is 6.00 as in amount per race so if down by $22.00
Thats only down by -3.7 units that's nothing.
Have a bank of 600.00 & go for it, have 100 shots at it.

You say it got to a high +14.00 then went backwards .
One could maybe stop once 2 lots of $6.00 is made = +12.00



So if the races are not being won by the 2nd-4th favs , what is winning all the races , surely its not the Fav every time or the rough heads.

If so, make the price filter for the 1st Fav higher.


Why don't you just fix it?

Try this race filter.
1st Fav 3.30+
2nd Fav 4.00+

Dogs
Horse races 1-11 runners .

Set Bot for 20 races.
Then STOP.

[The Ocho]

Of course you are right Bhagwan, a few days doth not make a summer. As a matter of fact a few months results may not mean anything.

I stopped the bot just before on the UK dogs on -$37.62. As you would say that is +12 up on the 2 days of that however it would of been stopped out overnight on -30 anyway (that's 20 up :) ).

I'm not sure who was winning all the losing races, all I know is it wasn't the 2nd, 3rd and 4th favs. But this goes along with what I was saying previously when tested on some of my (very limited) data - it's a bit up and down.

I'll leave it for someone else to test if they wish.

[rails run]

Being called "Neural" one would think that changing one factor would change all, but that is not the case.
I worked out how the individual setting are calculated and that is why I save ratings set at a uniform value and if I want to experiment I can change the setting.

lamaca, can you tell me how they calculate the TIM factor?

Do I understand this right? To get a true value on a particular factor you set all factors at '3', then increase or decrease the sole factor you are wanting to isolate?

Sorry for all these questions but you seem to have a great handle on these neurals.