System Test

[michaelg]

I'm currently testing a Place system with the neural BP algorithym. Over the past few days it has performed extremely well (also currently showing an acceptable profit with Win betting) and if it continues I'll list the rules.

Two selections today:

Canterbury R3 no.2 - Zarzeus
Moonee Valley R5 no.1 - Jacques Le Naif.

[partypooper]

Forum Curse again Michaelg!

[michaelg]

So true, but I haven't given up.

If there are any selections tomorrow I'll list them.

[partypooper]

Michaelg, you are right, 2 losers means nothing I had one plan with a LLR of 2 for 18 months till I started to bet real $, then 2nd day 6 losers in a row, beyond belief!!

[michaelg]

That's scary - hope its since been successful.

We'll try again today. The selections are:

Lismore R3 no.7 - Rose Of Ephraim
Lismore R7 no.2 - Derivative Receipt
Townsville R3 no.1 - Kanetoshi King
Townsville R5 no.1 - Sparkling Ruler
Townsville R8 no.3 - Brer Fox
Sale R5 no.4 - Orphan Hoaks.

[michaelg]

A better day yesterday - from the 6 selections there were 5 placegetters for a return of $7.40.

Two selections today:
Queanbeyan R7 no.8 - Kariz
Moonee Valley R1 no.5 - Phaze Action.

[michaelg]

Both selections yesterday placed for a return of $2.70. The method is now in profit.

Today's selections are:

Randwick R5 no.1 - Maybe I
Randwick R6 no.2 - Mentality
Newcastle R2 no.4 - Oakfield Jack
Newcastle R4 no.8 - Queen Of Hearts
Seymour R9 no.1 - Rokabaa
Gold Coast R5 no.1 - Ransom Bay
Gold Coast R6 no.4 - a Grade

[michaelg]

From yesterday's 7 selections there were 6 placegetters for a profit of $3.10.

There have now been 13 placegetters from 17 selections for a POT of 18.8%.

One selection today:
Flemington R4 no. 3 - Bird Of Fire.

[partypooper]

Wow! traveling nicely there michaelg, well done!

[michaelg]

Yes, it was traveling quite well, however yesterday's sole selection was unplaced.

There have now been 13 placegetters from 18 selections for a POT of 12.2%.

One selection today.

Moe R6 no.6 - Danni.

[michaelg]

Yesterday's sole selection placed for a return of $1.30.

There have now been 14 successes from 19 selections for a POT of 13.1%.

No selections today.

[partypooper]

Michaelg, I'm particularly interested in this thread (being an avid Place bettor)

Danni paid $1.40c Sportsbet Country Best. I haven't looked properly but I thought you sold yourself short the other day when most of your picks got up, most were available at City BEST tote, and the others at Country best (i.e. best of Supertab and NSW tab.

You've GOT to be on best tote when place betting!

[michaelg]

You're right, Party - it makes a big difference to the bottom line.

However, because I'm testing the method, although betting with real money, I use the Unitab divvies for the results.

Hoping it continues to perform...

[schonegg]

luxbet also offering best place tote, though on metropolitan racing only. Todays Randwick meeting was best tote win/place. (Not Belmont yesterday http://www.propun.com.au/racing_forums/images/smilies/frown.gif )

[michaelg]

Two selections today:

Canterbury R7 no.7 - News Alert
Bendigo R7 no.4 - Surfwatch.

[michaelg]

From yesterday's two selections there was one placegetter paying $1.50.

There have now been 15 successes from 21 selections for a POT of 9.5%

Two selections today:

Belmont R4 no.6 - Merignac
Belmont R5 no.4 - Lunar Module

[michaelg]

From yesterday's two selections there was one placegetter of $1.60.

There have now been 16 successes from 23 selections for a POT of 6.9%.

Today's selections:

Tamworth R5 no.8 - Snog
Tamworth R7 no.1 - Spock
Moruya R7 no.2 - Digger Prince
Rocky R6 no.3 - Paris Boy
Rocky R7 no.6 - Leather

[michaelg]

There were 3 placegetters from yesterday's five selections for a return of $5.50.

From today I will start over again and will apply a further rule - the selection must have placed last start. If this rule had been in force from day one then the method would be showing both a very good strike rate and profit.

Three selections today:

Flemington R4 no.2 - Baughurst
Flemington R5 no.2 - New Spice
Belmont R7 no.6 - Gloting

As a matter of interest, I am currently testing a Lay system. The selections unplaced last start are outperforming those that placed last start in both strike rate and profit.

[michaelg]

Not a bad start to the revised method - there were two placegetters from yesterday's three selections for a return of $4.00. I am now betting with NSW TAB so I'll record their divvies here.

Today's selections:

Sun Coast R3 no.1 - Key To Happiness
Benalla r9 no.4 - Invinciblekindness
Devenport R3 no.1 - Morsonique
Devenport R6 no.1 - Lunardi.

[michaelg]

Correction - as there are only seven starters at Devenport race 6 it is a non-qualifying race.

So now there are only three selections today.

[michaelg]

Not a good day yesterday - th re were two placegetters from the three selections for a return of $2.24.

Today's selections;
Morphetville R6 no.3 - Billy Jo
Bairnsadle R5 no.8 - Over Quota.

[Michal]

Hi Michael,


I would seriously consider not betting on selections that combined give you a profit of 24c after TWO sucessfull results.

That maths is against you. Lets say that one horse payed 1.24 take that and say there will be 10 bets where you get that 9 x 1.24 = 11.16
8 x 1.24 = 9.92

I realise that there were 2 horses one 1.2 and the other 1.04 on NSW TAB BUT

At 80% strike rate you are LOOSING, sure such low divs will or should ensure that the horses probably will run a place but somwhere in the 85% which is your break even. All you have is a good strike rate and no money, unfortunately.

Consider this

You loose 2 selections paying 1.2 that is -$2

To recover that loss you will have to win the NEXT 10 races at that Div JUST TO RECOVER.

Im talking about divs brackets, obviously your next bet might be $1.6 and next $1.4 and so on and you'll need only 4 of those to recover. By looking at the strike rate in the price ranges you will be able to make a educated decision and leave races that are risky in terms of loss recovery.

Take weekend Hustler Im sure he was 1.04 for a place ?
That loss will now take TWENTY FIVE CONSECUTIVE wins JUST to recover. Do you now see how fantactic value that bet is for the bookies, 1 in 25 for break even and 2 in 25 and they are infront ........

Im also Sure that there isnt a horse running on ANY Sunday that Justifys $1.04.

Look for profitability of results NOT strike rate of results.

Michal

[michaelg]

Hi, Michal.

I know a punter who bets mainly for the Place and will not touch any horse under $1.30, which makes sense especially due to the NSW TAB's method of rounding down.

However it is almost impossible to predict a Place divvy. One day my friend would not bet on a horse which was showing a divvy of $1.20 a minute or so before the start time - it eventually paid $1.70, and this was on a Saturday when one would expect minimal fluctuations during the last minute.

I agree it would be best not to touch the low divvies but again, one just cannot tell what a it will pay. So if it is a system selection I have to list it.

As a matter of interest, today's Bairnsdale selection is currently showing a Place divvy of $1.04.

[michaelg]

Both yesterday's selections placed for a return of $2.74.

There have now been 6 placegetters from 8 selections for a POT of 12.5%.

One selection today:

Townsville R6 no.1 - Culminante.

[partypooper]

michaelg, I've mentioned this b4, as it's topical I just thought I'd point out that my average divi is around $1.52c so nearly all favs or so, but in 4 years I've only had a handful paying $1.04c when backing BEST TOTE, it often happens that the nag pays $1.04c on one tote or even $1, and as much as $1.80c on the other, in fact I rely on this difference. keep up the good work!

[Bhagwan]

Hi Michaelg,
Looks promising so far.
Can you say at this stage what the Neural sttings are?

Cheers

[michaelg]

Hi, Bhagwan.

The rules are:

1 ) 8 to 13 starters.
2 ) All horses with at least two starts.
3 ) Default neural settings, but TIM at 0.
4 ) Must have top outright neural points.
5 ) Top or second top in the BP category.
6 ) Placed at last start.

Every now and then I also look at the top selection in other categories but have never recorded the results - might do so one day.

Off topic - anyone ever looked at the ratings on the Sky channel website? Yesterday I layed one horse per race that was in the Betfair market (under $11.50) but not selected with the top 4 Sky ratings - I had a very good day. I've never looked at their ratings so I'm curious if anyone else has.

[michaelg]

Yesterday's only selection was unplaced.

There have now been 6 placegetters from 9 selections for a break-even result.

Two selections today:
Dombeen R8 no.4 - Acquired
Canterbury R7 no.3 - Hormones.

[wesmip1]

michaelg,

I had a quick look at the sky ratings a while back with the same aim as you for laying. I only looked at it for a couple of weeks but it didn't work. Had a few good days but then hit the bad day which gave it all back.

Let me know if you find anything useful.

Good Luck.

[wesmip1]

Since you like running testing systems here is a simple one for you.

Any horse starting within 6 days (ie under 1 week) that is up in weight.

Someone else around here might be able to run some stats for you. If you want a heap more systems email me and I will point you to a great site with hundreds of systems to try.

Good Luck.

[michaelg]

Thanks, Wesmip1.

I had another good day today with Sky - 17 successes from 18 races. Over the past two days there's been 29 successes from 32 races.

A qualifying race for my method is that at least 3 horses have to be $11 or less and that the four Sky selections are not the top 4 in the Betfair market. I then lay the horse that is closest to the $11 mark and has not been selected by Sky.

I'm keeping details of the bets in the hope of applying some filters. Today's only accident was an unraced horse where most of the horses were also unraced. In these races its not only difficult to determine a price but also be easily omitted from one's selections.

Time will tell...

[michaelg]

There was one placegetter from yesterday's two selections for a return of $2.00.

There have now been 7 placegetters from 11 selections for a return of $11.00.

One selection today:
Northam R7 no.3 - Zuboyne.

[Merriguy]

Hi Michael,

Just a question or two about your 'Sky' method:
<link rel=\"\&quot;File-List\&quot;\" href="file:///C:%5CDOCUME%7E1%5CPeter%5CLOCALS%7E1%5CTemp%5Cmsohtml1%5C01%5Cclip_filelist.xml"><!--[if gte mso 9]><xml> <w:WordDocument> <w:View>Normal</w:View> <w:Zoom>0</w:Zoom> <w:PunctuationKerning/> <w:ValidateAgainstSchemas/> <w:SaveIfXMLInvalid>false</w:SaveIfXMLInvalid> <w:IgnoreMixedContent>false</w:IgnoreMixedContent> <w:AlwaysShowPlaceholderText>false</w:AlwaysShowPlaceholderText> <w:Compatibility> <w:BreakWrappedTables/> <w:SnapToGridInCell/> <w:WrapTextWithPunct/> <w:UseAsianBreakRules/> <w:DontGrowAutofit/> </w:Compatibility> <w:BrowserLevel>MicrosoftInternetExplorer4</w:BrowserLevel> </w:WordDocument> </xml><![endif]--><!--[if gte mso 9]><xml> <w:LatentStyles DefLockedState="false" LatentStyleCount="156"> </w:LatentStyles> </xml><![endif]--><style> <!-- /* Style Definitions */ p.MsoNormal, li.MsoNormal, div.MsoNormal {mso-style-parent:""; margin:0cm; margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:12.0pt; font-family:"Times New Roman"; mso-fareast-font-family:"Times New Roman";aa @page Section1 {size:612.0pt 792.0pt; margin:72.0pt 90.0pt 72.0pt 90.0pt; mso-header-margin:36.0pt; mso-footer-margin:36.0pt; mso-paper-source:0;aa div.Section1 {page:Section1;aa --> </style><!--[if gte mso 10]> <style> /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:""; mso-padding-alt:0cm 5.4pt 0cm 5.4pt; mso-para-margin:0cm; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:"Times New Roman"; mso-ansi-language:#0400; mso-fareast-language:#0400; mso-bidi-language:#0400;aa </style> <![endif]--><!--[if gte mso 9]><xml> <o:shapedefaults v:ext="edit" spidmax="1026"/> </xml><![endif]--><!--[if gte mso 9]><xml> <o:shapelayout v:ext="edit"> <o:idmap v:ext="edit" data="1"/> </o:shapelayout></xml><![endif]--><!--[if !supportLineBreakNewLine]-->
"A qualifying race for my method is that at least 3 horses have to be $11 or less" --- in the win market on B/F, I presume. Not much chance in the lay market!

"I then lay the horse that is closest to the $11 mark and has not been selected by Sky." In which market?

You mention a few filters could help. There is an English site (guess I'd better not mention names and court trouble!) which suggests filters (among others) of not laying the fav.; no races over 1m4f (2000 for us ?); not laying horses that finished first or second in their last run, or horses that have run in the last 8 days. They also rule out maidens and a couple of other races that do not apply here; and they state that there should be at least 10 runners.

May help in your endeavours --- which are much appreciated. All the above thoughts are offered free on the site in question so I don't see any problem about mentioning them here.

[michaelg]

Hi, Merriguy.

The filter of having at least 3 selections of under $11.50 (must be the Lay Price about a minute prior to the start in the Betfair market, I ignore the TAB markets) generally rules out those races that have a hot fave. I've never been keen on laying anything in these races. As a matter of interest, today's Gosford R5 had two horses under $11.50 where the Betfair fave (Steamy) was $2. The other horse under $11.50 was no.2 which would have been the Lay selection. Steamy did not finish the race and No.2 ran second, which would have been a bit of a worry.

Another filter that I've forgotten to mention - if at least two of the four Sky selections have been scratched then it is a no-race. One of the 3 accidents so far have been in this situation. Because I bet it, I include it in the results even though it was not a valid selection.

I have had not much action today. Only 3 races so far.
Otaki R5 no.12 - success
Bendigo R1 no.2 - success
Bendigo R6 no.6 - success.

Fingers crossed...

I'm keeping detailed records of the selections, such as placed/unplaced last start, field size, and a couple more. I would not be surprised if there are accidents in the larger fields because with a max of 4 selections they should be more accurate in smaller fields. Time will tell.

Edit - just missed Bendigo R7. The selection would have been no.8 which was unplaced - bummer!!!

[Bhagwan]

I was reading somewhere that Slow & Heavy tracks producers more photo finishes than Good , Fast or Dead tracks.
So maybe its an idea to leave Slow or Heavy tracks alone for that reason.

Also for those looking at the UK market,
1 furlong = 200m
1 mile = 1600m

Cheers.

[Merriguy]

Thanks, Michael, for the advice. Appreciated. It will be good to follow how the idea pans out.

I'm pleased to hear that the opportunities were limited yesterday as that seemed to be the indication from what I was doing.

And to make things worse I spent quite a while trying to sort out my PC clock, having failed to hear of the accident on the F3 and that Gosford was running a whole race "late". Oh well....Good luck today.

[michaelg]

Hi, Bhagwan.

All three accidents so far have been on good/dead tracks but its certainly something to look at. I would not be surprised if most of the English tracks were rain affected, so if that happens to be the case then maybe the reverse might be true here - better to lay on soft/heavy tracks?

Merriguy, there were more qualifying races yesterday but I was not at my computer for most of the day. I had only one more lay bet which was Ipswich R7 no.6 - it was beaten. 4 successes from 4 selections.

Yesterday, the sole selection in the Place method paid a measly $1.20 here in NSW but $1.60 on Unitab. Having the best tote feature certainly makes a huge difference!!!

Today's selections:
Port Mac R5 no.4 - Soiled Dane
Wangaratta R6 no.2 - Danni.

[michaelg]

From yesterday's two selections there was one placegetter paying $1.40.

There have now been 9 placegetters from 14 selections for a LOT of 2.8% ($0.40).

Three selections today:

Gold Coast R5 no.1 - Left Standing
Toowoomba R3 no.4 - Spanish Sister
Belmont R6 no.3 - Dante's Banquet.

[Merriguy]

One more clarification, Michael, if I may. In writing about which horse to lay you say:


I then lay the horse that is closest to the $11 mark and has not been selected by Sky.
Is the idea that you lay the one closest to $11 (no matter if there are other horses who would qualify by being not among the Sky selections) or do you go for the lowest of those not in the Sky selections (and so have a lesser liability)?

You did mention that a smaller field means that one of the four mentioned on Sky are likely to take the money; but it is also a fact, surely, that the larger the field the more is the possibility that something else will get up (besides your pick to lose)!

[michaelg]

Hi, Merriguy.

In these cases where there is more than one selection I trust the Betfair market. I ignore the one with the lower liability and I then lay the one closest to $11.

Unfortunately I don't have the complete results of field size for last Tuesday but with the results that I've got I've split races (1) with starters of 11 and less, and (2) with starters of 12 runners or more.

My records show (omitting last Tuesday) that there have been 11 selections in fields of 12-plus runners. Of these there have been 2 accidents (I don't have records for the 2 other accidents that occurred last Tuesday), one selection that ran second, and six that ran third. Today I have layed two selections - Caulfield R1 no.10 and Port Mac race 3 no.10. Both were in fields of 12 runners and both selections ran second!!! In fact, these two successes today has put 12-plus runners into profit.

However, in fields of 11 or less there have been (again, omitting Tuesday) 25 selections for no accidents, two selections that ran second and three that ran third.

As a matter of interest, even though there were two accidents on Tuesday, the method won 5.45 units for that day.

This very small sample (a total of 36 selections) currently shows that the smaller fields might be encouraging. Only time will tell...

Since last Tuesday there have been 46 successes from 50 selections for a profit of 30.75 units.

[Merriguy]

Thanks, Michael, understood.

Will be interesting to see how things go today --- there are over 80 races!!! I know that there is no way they all will be qualifiers --- and I may get some sort of fatigue syndrome, anyway --- but things seem to be going well so far.

Incidentally, I will try (on paper) using the 4 selections on Unitab (or perhaps you have already done that)?

[michaelg]

Merriguy, I forgot to mention that if a $11 selection is layed then one of Sky's selections must be over $11. So if there are two selections then its possible that the lesser price might end up being the one that is layed.

Ive looked at Unitab using the same method but it failed miserably, so too did the DFS ratings.

The only one that showed promise was the neural BP barrier (can't remember the exact rules) but I had a couple of bad accidents. I remember that the damaging accidents were in fields of 12-plus. I might see if I can find the rules.

I've just had my first "11 runners" accident - Port Mac R5 no.7 (liability $6.60) but am still in profit for the day.

[michaelg]

Merriguy, I did not have the best of days yesterday. I layed 31 selections for 4 accidents including a $10 liability. However, I finished the day with a profit of 1.85 units. Did you give the method a go?

From last Tuesday I've layed 81 selections for 73 successes giving a profit of 32.60 units with a success rate of 90%. Two of the accidents were the race fave - it's surprising that Sky didn't select them.

So far so good...

[Merriguy]

Well, I came out ahead --- about the same amount as yourself, just over 2 units.

I only layed in 18 races as I had to go out for a few hours on a 'mercy' mission; but was unsuccessful in three races --- two by a nose!!! One of them was also a favourite not selected by Sky (probably one of yours) which I too found strange.

I was probably not as strict as yourself with a couple of your rules so I doubt we can directly compare results --- but the 'trend' is there I think (hope). I cannot quite get myself away from the taking of the lowest possible horse to lay when there is more than one available. Must be the Scrooge in me --- but, to be honest, I doubt it makes much difference from what I have seen. The savings in liability add up!

Perhaps the advice I mentioned above from that English site is a necessary filter: "Never laybet the favourite".

Thanks for the advice --- and Good Luck.

[michaelg]

Merriguy, I've added another filter to the method - if the selection is the fave then it becomes a no-bet race.

I've now been using the method for exactly one week.

I've layed 110 selection for 100 successes giving a profit of 42.88. This includes accidents of two faves and also an unraced horse. The current rules would have eliminated these three accidents.

So far so good...

[michaelg]

Merriguy, I've added another rule - if the Betfair fave is odds-on then it's a no race. Are you still using the Sky selections to lay?

It is now the second week - it has not been as promising as the first but still profitable. There were 128 selections for 112 successes for a profit of 11.10 units.

As per the method I have now layed 238 selections for 212 successes (89% strike rate) for a profit of 53.98 units.

So far so good.

[Merriguy]

Hi Michael,

Have been very busy over the last few days --- and have not been able to concentrate on the idea. I struck a very bad patch on Friday, however, and lost about 5/9. Still I am pretty sure that I am still ahead with it even if only a little. Hopefully will be able to concentrate a little better later this week.

Good Luck

[michaelg]

Merriguy, I think the rule that if there are more than one selections then laying the highest liability has paid dividends (pardon the pun) over the past two weeks.

There have been a few instances that laying the lowest liability would have been quite damaging.

But as it has been only over the past two weeks it's too early to say if it has long term benefit.

Friday was my worst day - 3 accidents from 10 selections for a loss of 9.15 units.

[michaelg]

Merriguy, or anyone else, have you been laying the Sky system?

It's now the end of the third week. During this period I have layed 127 selections for 116 successes, a strike rate of 91.33%, for a profit of 35.70 units.

After three weeks I have layed 365 selections for 328 successes, a strike rate of 89.86%, for a profit of 89.68 units. As a matter of interest, the strike rate to date is higher than Maria's.

Because of the strike rate I have been considering a stakings plan - start with 1 unit and when an accident happens it then increases to 3 units until the loss from the accident has been regained. However, if another accident happens before recouping the loss its then increased to 6 units until the loss has been recovered. However, continue at 6 units even if the loss has not been regained even if another accident happens. If I had used this plan then the method would be 216.00 units in profit.

So far so good, and there's a lot of action.

[Wombat2]

Michaelg,

If you would like me to put your 365 odd lays into the TSM staking software, I can but we need to be able to email each other.

Cheers Wombat2

[michaelg]

Hi, Wombat2.

That would be great but I would have to e-mail you the info manually race-by-race because I don't have it on a spreadsheet - I know nothing about spreadsheets.

My address is mglfyl at primusonline.com.au.

[wesmip1]

Michaelg,

Looks like you are going very well. Congratulations.

Hope it continues..

Good Luck

[michaelg]

Thanks, Wesmip1. Yes, let's hope it continues.

Today was a losing day, the fourth in 22 days. It lost 3.80 units.

However, the stakings plan won 3.85 units. From the 22 days its only had one losing day - but that could change dramatically from tomorrow.