Testing Neural Method

[michaelg]

For the past three weeks I've been incapacitated due to a pinched nerve in my back - I can't walk without experiencing terrible pain in my right shin. So I've more or less been confined to a chair, and I've had a lot of time to test diffferent systems.

The following neural one shows promise, but I've only been able to look at the past two days, there's only been a total of 15 selections.

1) Default neural setting, except CF and TIM on zero.
2) No NTD races.
3) The selection to have a minimum of 140 neural points.

On Sunday there were 14 races for 5 winners giving a return of $27.60 (NSW TAB divvies), and 9 placegetters for a return of $15.14. Three of the losing placegetters ran fourth.

Yesterday there was only one selection - Goulburn Race 5 no.13. It won, paying $6.70 and $2.10.

Out of interest I'll list today's selections and see how they go.

Geelong
4/1
5/3
7/2

Hawkesbury
5/1
6/4

Lismore
2/3

Townsville
2/4
3/4
4/4

[Try Try Again]

Hi Michaelg,

Good luck with the pinched nerve, hope you're up and about soon.

I think you have incorrectly identified Hawkesbury Race 5 selection, it should be #6 Maluckyday (145.5) not #1 Gemma's Boy (84).

Hope they are all winners!

[michaelg]

Hi, Try.

Yes, you're right. Its no 6.

Its resuming from a spell. The DLR understandly only gives it one point, but most of the other algorithyms rate it very highly and have made it a selection.

Thanks for the concern about my pinched nerve. Even though the pain in my shin can still be excrutiating its thankfully nothing like what it was. Its times like the last few weeks that make you really, really, really appreciate when you are in good health.

[Try Try Again]

Hi Michaelg,

Interestingly, the only other contender over the past 2 days that also had a DLR =1 (Hobart Race 7 #1 Conquering) did not run a place . Hopefully that does not look bad for Maluckyday!

[michaelg]

That's a good observation - and Conquering had 161 points!

We'll see how it goes.

[Try Try Again]

Hi Michaelg,

It would seem it is a bit early to get rid of DLR=1 horses as Maluckyday won and paid $3.40/$1.50.

Going well today with still 2 races to go.

[michaelg]

Try, it paid only $2.70 here in NSW.

A losing day.

However, if the selections that had run fifth or worse at their last start had been omitted, then there would have been 5 selections for 3 winners paying $5.80, and 4 placegetters (the other ran fourth) for a return of $5.04.

I checked the two previous days, and all the selections ran fourth or better at their last start, so the profitable results for those two days remain unchanged.

Tomorrow I'll apply this filter and see how they go.

[Try Try Again]

Hi Michaleg,

Not taking into account your new rule, there are a couple of races today (excluding Belmont which do not have scratchings through yet) where we have 2 qualifiers.

How do you distinguish between them or do you back them both or do you only back the top rated horse?

The races are Sandown Race 3 # 2,3 and Strathalbyn Race 7 # 3,4

It shows that by getting the best of the Totes you can make a small profit against having a losing day. If we are serious we should strive to get the best result for our money every time!

[michaelg]

Hi, Try.

I don't remember noticing if there's ever been two selections in the same race where they've got 140-plus points, because I've only been looking at the horse with the highest points. If it has 140-plus points I have recorded that horse as the selection, and have ignored all others in that race. So its possible there may have been more than one horse with 140-plus points.

My neural figures for today's races don't agree with yours. The CF, TIM and WET (I did not previously mention WET) algorithyms are set on zero, all others on 3.

Sandown R3 according to my figures, no. 3 has 139.5 points, and no.2 has 127 points.

Strathalbyn R7 - no.3 has 158.50 points, and no.4 has 136 points.

You're right about getting the best price, but at the moment I don't hold an account with any of those agencies that provide it, except for Betfair which I understand isn't really the way to go. If after a reliable period of time has elapsed and the method has done well then I might look at C'bet.

Today's selections are:

Belmont
4/1

Canterbury
4/1

Eagle Farm
6/1
7/10

Strathalbyn
7/3


There are 2 selections today that finished worse than 4th at their last starts - I won't be betting them.

They are:

Canterbury 5/3
Sandown 6/1


I've now looked at today's N.Z. meeting using the method. There have been only 3 selections, and they have alredy been run. They were:

1/7 - not a system horse as it finished worse than 4th last start - it was unplaced.

2/2 - as above, it was also unplaced.

3/9 - it was a system selection as it's last start placing was 4th or better - it won, paying $2.10 and $1.40.

[Try Try Again]

Hi Michaelg,

I think you have missed Belmont Race 5# 10 Dignity Haze 141.5 points and won last start.

I discovered why my figures were wrong - somehow WET = 2, I have reset this to zero and everything is sweet!

I will take on board you comment about horse must have finished in the first 4 last start.

Good luck today!

[michaelg]

Try,

yes, I missed Belmont R5 no.10 - it is a selection.

Thanks.

[peterpan]

I might look at C'bet.

The only problem there michaelg is if the method works and shows consistent profit week in week out they will ban you or set your bet size so small it's not worth betting.

Best of luck.

[michaelg]

Going on yesterday's results I don't think Centerbet would be too concerned.

Two winners from the five selections for a loss of $0.90, and three placegetters for a loss of $1.30.

One selection today:

Geelong R3 no.11.


I'll also list N.Z. selections.

2/3 - unplaced
3/3 - won, paid $2.80 and $1.04
8/1

There are four selections that don't qualify because they finished worse than fourth at their last start or are in fields of less than 8 starters.

Maybe it might turn out to be a Lay system so I'll list them.

Gosford 1/3
Pt Augusta 1/1
Rocky 1/1
Rocky3/1

[Try Try Again]

Hi Michaelg,

I think I would hold off on the Lay system - 3 winners and a 2nd from the 4 selections (Rocky R2#1 instead of R1#1 was the selection) would certainly have dented the bank balance!

The Geelong selection won but at short quote - still a winner is better than a loser!

It all adds to the profit.

[michaelg]

Yes, it may not turn out to be a good Lay system after all, but I'll still be looking at them just in case today was an exception.

Today's winner was disappointing with its divvy, but not with its performance. And as you say, a winner is better than a loser.

I don't know how I made that typo with Rocky R2 no. 1 (the pain in my shin?) because it was a selection in another system I mentioned in a different thread on this forum. I said in that thread I would wait for about 100 races before I detailed the rules, however I'm now taking the opportunity here to say I have decided to wait longer. The system up to yesterday has had 117 selections for 52 winners for a POT of 43%, and is also in the black with Place betting. Today there are 10 selections. 9 have already been run, for 5 winners with a return of $11.10, and 8 placegetters with a return of $12.28.

Hopefully when I reveal it here, the F. Curse will not kill it.

[michaelg]

Today's selections are:

Geelong
5/12
7/8
9/8

Ipswich
4/9
7/2

Muswellbrook
9/1


The following finished worse than 4th at last start:

Geelong
1/9

Ipswich
1/8
5/4

[michaelg]

From today's 5 selections there were 2 winners for a return of $6.80, and 4 placegetters for a return of $5.40.

Since posting here on Wednesday, there have been 11 selections for 5 winners for a return of $12.30, and $10.14 for the Place.

However, if those particular selections that had finished worse than 4th at their last start had been included, then yesterday would have been a better day. Today there were 3 such selections, as named in today's earlier posting. They all won, paying a total of $13.40, and $5.50 for the Place. So from tomorrow I'll cancel that rule about last start finishing position, but I will keep the NTD filter.

In my first posting on this thread I mentioned I had tested only 15 selections. This was much too few, but in spite of that it's continuing to show promise.

[Try Try Again]

Hi Michaelg,

Interestingly, today we had 2 qualifiers for Ipswich race 7 (#2 Sequiz 145.0 and #5 Heart To Fear 140.0) and both ran unplaced.

From 22/8 the results (using Best of the Totes dividends) there has been 42 bets for 19 winners - S/R 45.2% - Bank $35.10 ($1 units) - POT % = 83.6%,

If you only consider horses that ran 1st to 4th last start then there has been 30 bets for 13 winners - S/R 43.3% - Bank $26.80 ($1 units) - POT% = 89.3%

So there appears to be very little in it whichever way you want to bet the selections. The main thing it is making a good profit either way. Let's hope it continues!

[michaelg]

Try, thanks for the info re the position of last start - its interesting that it didn't make any difference.

The strike rate and POT are impressive but I certaintly can't expect it to continue. IMHO the main positive factor in the system is that the two most puzzling and over-generous neural categories (CF and TIM) have no bearing in the selection process.

Ipswich race 7 no.2 got beaten in a photo-finish for third. If I remember, I'll check to see if there is more than one horse with 140-plus points in the same race. Did you notice the size of the exotic divvies?

[michaelg]

Today's selections are:

Belmont
2/2
6/4

Caulfield
1/1
2/8
5/2 (no. 6 also has 140-plus points)
6/9 (nos. 12 and 2 have 140-plus points)

Dombeen
2/2
3/1
4/1
6/1

Darwin
4/5

Gold Coast
2/7
6/1
7/10
8/6

Morphetville
5/4

Newcastle
2/5

Rosehill
1/1
2/3
3/1
4/2
5/9
7/4 (no. 1 has 140-plus points)
8/5 (no. 12 has 140-plus points)

Toowoomba
1/1
2/3
4/1
5/8
7/1

It'll be interesting to see how the selections at the four major metro tracks fare. The results of neurals for these tracks generally differ to those of the non-metro tracks.

[michaelg]

Just as was suspected with the four major metro tracks - there were 21 selections at 140-plus points, for 2 winners paying $3.70 and $2.10. Laying them would have produced a profitable day. Its something I'll look at.

The other tracks did not have the best of days even though they did not lose. There were 13 selections for 5 winners paying a total of $13.70, and 8 placegetters for a return of $13.10.

Today's selections:

Launceston
2/4
4/7
6/1

Nowra
4/3

Seymour
3/13
5/9

Sun Coast
4/3 (no.3 has 166.5 points, and no.1 has 143.5 points)

[michaelg]

Again, not the best of days. There were 2 winners from the 7 selections for a return of $7.50, and 5 placegetters for a return of $6.60.

[michaelg]

Today's selections are:

Cessnock
3/1

Cowra
1/6
2/8
4/1
6/3


Here's something interesting - I've currently got too much time on my hands.

Keeping the same neural settings used for the above method, if from Friday one had bet in races that had:

1) 11-plus runners.
2) Dutch betting every horse under $20 in the neural market for a target of $100

the results would have been using NSW TAB divvies:

Friday
11 races for 11 winners
Outlay of $975
Return of $1397
Profit of $422

Saturday - excluding the four major states:
20 winners for 15 winners
Outlay of $1682
Return of 1702
Profit of $20

Sunday
23 races for 18 winners
Outlay of $1947
Return of $2282
Profit of $335

By my reckoning, there have been 54 races for 44 winners. The outlay has been $4604 for a return of $5381 which is a profit of $777, and a POT of 16%.

There have been no filters, apart from excluding Saturday's four major states.

Almost all the winners that were not struck in the method were at single-figure odds on the TAB. Conversly, many winning outsiders were snared, some even at single-figure prices in the neural market which produced a good result. I haven't checked the exotics but I remember there was a First Four that paid about $12,000, and there were only 7 selections in that race.

The average number of selections per race was about 7.5, and the average outlay was about $85.

[michaelg]

Again, not the best of days yesterday. There were 5 selections for one winner of $3.50, and 3 placegetters totalling $6.90. Place betting seems to be propping up the system.

Today's selections:

Townsville
4/1
5/7
6/3

[tedbluegum]

Hi michael.. thanks for your very interesting and thought provoking posts in this forum.

With reference to the post 23 you looked at dutch betting based on your neural setttings of 3 with the exception of CF,TIM and Wet which you set at 0.
What % did you use please? The normal 114%?

NO Metro Sat meetings for all States considered?

Fields of 11+ runners, Dutch bet everyhorse under $20 in the neural market -
for a target of $100.

Is this every horse in the market (maybe up to 6) under the $20? Or do you filter the number to say 3 or 4?

I had a sore neck last week..an old sports wound ..maybe its catching..the Doc gave me a packet of Inflamitory tabs and they did the trick..

Hope yours is getting better

Regards
Ted

[michaelg]

Hi, Ted.

The neural market is set at 100%, this can be done on the PR facility. Sorry for not mentioning that.

Every horse under $20 in the neural market is dutch-bet as per their neural price, So if there are, say 10 selections, then all are bet. I had a look at omitting races where more than $79 is bet, they performed quite well but it drastically reduced the number of qualifying races so I did not persue it. I also had a look, not too comprehensively, at horses that were neural-priced at single-figures. The large wins were all from these, but because it reduced the overall action I did not keep an accurate record.

Yesterday would have been a losing day because most of the faves won the qualifying races, and I haven't checked today. However, I've now looked at today's Grafton R4 which was a qualifying race - the winner paid $31 on NSW TAB, and the neurals priced it at $8.00. So maybe focussing at horses under $10 might be something to consider.

Only the four major NSW, QLD, VIC and SA meetings are omitted - imho these seem to historically have very weak results with neural Win betting.

Tonight I returned to the scene of the crime of my problem shin - the gym. I had a limited workout, but at least I can walk even though I still have to limp. I expect to be back to normal in a few days.

Thanks for the concern.

Michael.

[michaelg]

A good day yesterday. There was one winner from the 3 selections for a return of $4.50, and all three placed for a return of $5.00.

Today's selections:
Ballarat 3/6
Armidale 5/3.

[michaelg]

Tedbluegum,

a qualifying race today for the system was Armidale R2. The winner paid $44 on NSW TAB - the neurals priced it at $8.50.

If every neural single-priced selection today was backed for $1, then the total outlay would have been $42.

[michaelg]

There was one winner from yesterday's two selections, paying $2.30 and $1.10.

Today's selections are:

Northam
2/7
4/7
6/3

Sale
1/9

Strathalbyn
7/1

Wyong
2/7
8/6

Sun Coast
2/1
4/1
6/1

[Try Try Again]

Hi Michaelg,

Are you sure Northern Race 6#3 is a qualifier, I get as only scoring 90 points.

Also I have Sale Race 3#5 Chapin (155.0 points) as a qualifier.

[michaelg]

Try, I made a typing error with Northam R6 - it should be No2. Even though it's got 139.5 I rated it up to the nearest point. When testing earlier there were 2 selections with 139.5 points, one of them won and the other placed. Hopefully I've done the right thing here.

And Sale R3 no.5 is also a selection. Don't know how I missed it.

Thanks.

[michaelg]

From yesterday's 11 selections there were 5 winners for a return of $12.30, and 8 placegetters for a return of $10.84.

Today's selections are:

Cairns
2/1
4/2
5/2
6/3

Canberra
4/5

Geelong
6/4
9/6

Wyong
2/5
3/7
4/3
5/1
6/1
7/1

[michaelg]

Not a bad day.

From the thirteen selections there were 3 winners returning $20.60, and 8 placegetters returning $14.10.

[Try Try Again]

Hi Michaelg,

I now only bet on the qualifiers that have finished 1st, 2nd, 3rd or 4th last start and are not running on a City track (Melb, Syd, Bris, Adel, Perth and Hobart). If there are 2 qualifiers from the one race, I back both. Today had 2 races where that occurred and both winners were found ($7.70 and $4.10).

Since you put up the original rules (22/8) there has been 13 betting days. Betting only on the above rules there was one day where there was no qualifiers. Of the other 12 days, only 2 were losing days with the bank now reaching a new high today and showing a profit of 35.5 units.

There has been 57 bets for 24 winners (S/R 42.1%) and places (S/R 71.9%), the place dividends are also showing a 7.84 unit profit.

Well done on providing such a great little money earner!

[michaelg]

Try,

it seems you've ramped the system up to another level - well done. Hopefully it will continue.

As a comparison to your results I've recorded mine from 25 August. They are:-

24 winners from 65 selections for a POT of 16%. Of the 10 days there have been 8 winning days.

44 placegetters from 65 selections for a POT of 4%. There have been 6 winning days.

I have looked at Perth and Hobart. There have been 7 Perth selections for 1 winner of $2.50 and 5 placegetters for $7.34. And with Hobart there have been 3 selections for 1 winner of $3.20, and 2 placegetters of $2.90. I will now record the results from these 2 states.

As a matter of interest I can only check from Wednesday, but if the selection has the outright most points in the JA category, there would have been only 7 selections, but 3 winners totalling $20.60, and 5 placegetters totalling $9.30.

[Stix]

Hi Guys

Have you looked at selections that are carrying >=56.0kg?

I know with one method I've used, that it has helped. Also is you bet only 99 or 100 Unitab raters, does that have a positive effect?

Just a couple filters I've used before...and it works for me.

Good Luck...

[michaelg]

Hi, Stix.

I've checked the past two days results using your suggestions.

There were 24 selections, and the weight allocated to them did not make any difference. And for the 99 and 100 Unitab pointers, there were 12 of them for 3 winners returning $7.00, and 6 placegetters for a return of $10.38. It would appear they were overbet. It's still much too early to tell but unfortunately it does not appear encouraging.

Today's selections:

Belmont
4/3
8/14*

Cairns
6/3*
7/2*
8/1*

Darwin
1/2*
2/3*
4/1
5/3

Pt Mac
5/3

Tamworth
2/4
3/5*

Toowoomba
1/7*
2/3
3/6
4/3
5/3
7/7*

* = outright top JA points.


The following are the selections for the 4 major states. Judging from last Saturday they might be good Lay bets. We'll see how they go.

Eagle Farm
3/1
5/2
6/1
8/1

Flemington
1/1
2/3
4/3
7/4
8/2

Morphetville
4/6

Randwick
2/1
4/2
5/1

[lomaca]

The neural market is set at 100%, this can be done on the PR facility. Sorry for not mentioning that.

Hi Michael, while testing somet I noticed that only the rated price is changing a little by altering the %.

To the other numbers it makes absolutely no difference, which is as it ought to be, so unless you use the rated prices it's not much point mucking about with.

Specially since from 90% to 114% it only changes by a few $ the order of selections doesn't change.

Good luck

[michaelg]

From yesterday's 18 selections there were 5 winners paying a total of $17.30, and 10 placegetters for a total of $15.84. A Win profit would have been achieved if Perth had been omitted. I'm recording them seperately.

There were 9 selections that were top in the JA points, for a Win return of $8.60, and $8.34 on Place. Three of them ran fourth so it looks worthwhile to continue recording them.

The results from the 4 main metro tracks would've been profitable if they had been bet, not layed. The outlay of $13 would have returned $20.40. Interestingly, only two selections were top in the JA category - they both won paying $4.20 and $1.40, and $3.30 and $1.50.

Today's selections:

Sun Coast
2/5
4/2
6/2

Kalgoorlie
2/1
3/2
6/1

None are top in the JA neural category.

[michaelg]

Not a bad day yesterday. From the 6 selections there were 2 winners for a return of $7.90, and 4 placegetters for a return of $7.00.

Today's selections:

Coffs
2/4
6/1

Geelong
2/6 (Top JA)
6/8 (biggest outsider in pre-post market)

[michaelg]

A bad day yesterday - no winners from the 3 selections, and 2 placegetters for a return of $2.60. Thye JA selection ran second paying $1.10.

Today's selections:

Geelong
4/1
7/5

Goulburn
5/8

Mackay
5/1 (top JA)

[michaelg]

A good day yesterday. From the four selections there were 3 winners for a return of $8.60, and all placed for a return of $5.20.

Today's selections:

Belmont (a very poor record so far)
1/1
5/1

Hobart
5/1 (top JA)

Murray Bridge
1/1
6/4


Even though the 4 major states/tracks are not in the method, the following have the highest JA so I'll see how they go.

Canterbury 2/8
Sandown 3/5

[michaelg]

From today I'm following Try in omitting Belmont, Ascot and Hobart from the system as they are major cities in W.A. and Tasmania.

Today's selections are:

Ballina
1/10

Geelong
1/7
2/4

Northam
2/5 (top JA)
6/3

Rocky
3/3
4/2 (top JA)
5/1
6/4

[michaelg]

From yesterday's 9 selections there were 2 winners for a return of $9.70 and 4 placegetters for a return of $8.30. One of the top JA selections won paying $6.00 and $3.30.

Today's selections:

Geelong
3/7 (top JA)
9/12

Ipswich
2/9
3/1

Port Mac
5/1
6/3 (top JA)

[tedbluegum]

G'day Michael

I have been following this thread for some time now and would appreciate if you could spell out your neural settings etc. I seem to get lost keeping up with them.

Thanks for the time you are able to put in.

Regards
Ted

[michaelg]

Hi, Ted.

The neurals are all on a setting of 3, except for CF, TIM and WT which are on 0. The method's selection must have at least a total of 140 points. If there is more than one horse with 140-plus points then I bet the one with the most points. This seldom happens though.

The filters are.

1) Minimum of 8 starters.
2) Omit the major tracks in the 4 main states.
3) Omit Belmont, Ascot and Hobart tracks. This is a very recent filter.

I am also looking at the selections that have the most outright points in the JA neural category. They are performing quite well so far even though there's only been a handfull of them.

You may be following Try's thread titled Neural Favourites where he is having success based on his variations to the above method.

[tedbluegum]

Thanks Michael for your prompt response.

What % (percentage) do you set your neurals to>

Regards
Ted

[michaelg]

I set the neurals to a market of 100%.

One selection has already been run - Geelong R3 no.7, which was also a TA selection. It won paying $1.80 and $1.10. Very disappointing divvies, but to quote someone on the forum "a winner is better than a loser".

[michaelg]

Not the best of days yesterday. From the 6 selections there were 2 winners paying a total of $3.70, and 4 placegetters for a return of $5.90.

There was one winner from the 2 JA selections paying $1.80, and both placed for a return of $2.70.

Today's selections:

Gold Coast
1/4
6/3

Kembla
3/1
5/6

Toowoomba
1/1 (JA)
2/1
3/4


Out of curiousity there are 2 metro selections that are top in the JA category:

Morphetville 6/2
Rosehill 3/6

[michaelg]

A bad day yesterday. 1 winner from 7 selections for a return of $4.00, and 3 placegetters for a return of $5.00.

Today's selections:

Scone
1/9
2/2
7/3

Queanbeyan
5/2

Sale
2/7 (JA)

Bendigo
2/2
3/14
4/5 (JA)
7/2

Sun Coast
5/3
7/4

Kalgoorlie
2/5 (JA)
4/1 (JA)
6/2

[Try Try Again]

Just to give an update on the results achieved with the adaption to Michaelg's original work.

Results are from 22th August up until today (excluding 14th to 23rd September when I was away O/S), hence 35 betting days.

There has been 141 bets for 49 winners (S/R 34.8%) and 101 places (S/R 71.6%).

Win Bank = +24.0 units, POT % = 17.0%
Place Bank = +17.48 units, POT % = 12.4%

The maximum the win bank has been was +41.90 units, whereas the place bank is at a new high. There has been one run of 15 places in a row.

Obviously getting the best possible dividends is the utmost importance to achieving a profit.

[michaelg]

Try, those are excellent results.

Win betting from most systems that provide a single selection can always be a worry because of its topsy-turvy nature. Your highest point was a profit of 41.90 units which has now almost halved to 24, but one or two good days could quickly change that. Place betting is sailing along very nicely, and with a strike rate of 71.6 from 141 races giving a POT of double-figures makes the Place bet look quite promising.

Keep up the good work.