Testing Lay System

[michaelg]

The method for the 12 Months ratings on the R & S website have produced a profit since April with both Win and Place betting as per lomaca's results. So the ratings would seem to be somewhat accurate.

Based on them I've looked at a Lay system from last Saturday.

1) No maidens.
2) Maximum of 9 runners.
3) Lowest points in the 12 Months ratings.
4) Last start more than 21 days.

Over the 3 days there's been 17 selections for no accidents, one second and one third. I have also looked at 10 runners - there were no accidents but have decided to limit the method to a maximum of 9 runners because the smaller the field then generally the lower the prices/liability.

There are two selections today, we'll see how they go.

Taree R1 no.1 ($2.80 pre-post fave)
Townsville R6 no.5.

[michaelg]

The Forum Curse struck yesterday with the very first selection.

The total loss was 0.95 units.

One selection today:

Canterbury R1 no.4 - $2.80 fave in the pre-post market.

[wesmip1]

Whats your maximum price ?

[wesmip1]

Here is something to help you out with laying. Here is a pure R&S laying system that has a positive expectation of approx 11%.

First the resultsfor the last 3 months (Assuming 5% commission):
Selections : 217
Strike rate : 23 / 217
Profit on laying to a risked liability of $100 : $1257.08

Month 1: $ 97.75
Month 2: $634.88
Month 3: $524.44

The rules :

1. Go to the neurals. Only look at TA and CF. Find the ranking of each one in the list and write it down. Any horse which has a combined ranking of 21 ( ie trainer = 10th rank and cf = 11h rank) is a selection.
2. Only lay up <= $10 on betfair.
4. The above is done pre-scratchings. So you can work out the selections the night before and then just worry about watching the price on raceday.

What is logical about the system rules?
The system is designed to find a trainer who is out of form/does not perform at this track with a mediocre horse in a large field at a price which is low enough to not be a risk to our bank.

Feel free to fiddle with it and add extra filters.

This weeks selections:

<TABLE><TBODY><TR><TD>Hawkesbury</TD><TD>8</TD><TD>.</TD><TD>PATIO</TD><TD>came </TD><TD>3</TD></TR><TR><TD>Ipswich</TD><TD>7</TD><TD>.</TD><TD>KOLO</TD><TD>came </TD><TD>3</TD></TR><TR><TD>Ipswich</TD><TD>7</TD><TD>.</TD><TD>WILD PROMISES</TD><TD>came </TD><TD>1</TD></TR><TR><TD>Ipswich</TD><TD>8</TD><TD>.</TD><TD>GUNNER KNOX</TD><TD>came </TD><TD>1</TD></TR><TR><TD>Ipswich</TD><TD>8</TD><TD>.</TD><TD>ZEROAD IN</TD><TD>came </TD><TD>13</TD></TR><TR><TD>Coffs Harbour</TD><TD>4</TD><TD>.</TD><TD>WAR MACHINE</TD><TD>came </TD><TD>4</TD></TR><TR><TD>Coffs Harbour</TD><TD>7</TD><TD>.</TD><TD>MR WIGGLES</TD><TD>came </TD><TD>1</TD></TR><TR><TD>Eagle Farm</TD><TD>8</TD><TD>.</TD><TD>JAZZ HEART</TD><TD>came </TD><TD>9</TD></TR><TR><TD>Eagle Farm</TD><TD>8</TD><TD>.</TD><TD>LISTEN SON</TD><TD>came </TD><TD>1</TD></TR><TR><TD>Gold Coast</TD><TD>4</TD><TD>.</TD><TD>ALETHEIA</TD><TD>came </TD><TD>3</TD></TR><TR><TD>Moonee Valley</TD><TD>9</TD><TD>.</TD><TD>KATIE LEE</TD><TD>came </TD><TD>9</TD></TR><TR><TD>Ascot</TD><TD>6</TD><TD>.</TD><TD>OUR OL' FELLA</TD><TD>came </TD><TD>10</TD></TR><TR><TD>Ascot</TD><TD>8</TD><TD>.</TD><TD>EXTRA BEAT</TD><TD>came </TD><TD>10</TD></TR><TR><TD>Sunshine Coast Cushion Track</TD><TD>7</TD><TD>.</TD><TD>MARTHA'S VINEYARD</TD><TD>came </TD><TD>10</TD></TR><TR><TD>Nowra</TD><TD>6</TD><TD>.</TD><TD>MONTE LEO</TD><TD>came </TD><TD>4</TD></TR><TR><TD>Nowra</TD><TD>7</TD><TD>.</TD><TD>BLACK ROCK</TD><TD>came </TD><TD>9</TD></TR><TR><TD>Nowra</TD><TD>7</TD><TD>.</TD><TD>ELLIOTT STREET</TD><TD>came </TD><TD>4</TD></TR><TR><TD>Sale</TD><TD>3</TD><TD>.</TD><TD>GEORGEALMIGHTY</TD><TD>came </TD><TD>8</TD></TR><TR><TD>Sale</TD><TD>9</TD><TD>.</TD><TD>KING COBWEB</TD><TD>came </TD><TD>7</TD></TR><TR><TD>Terang</TD><TD>6</TD><TD>.</TD><TD>SYLVAN O'REILLY NZ</TD><TD>came </TD><TD>7</TD></TR><TR><TD>Hobart</TD><TD>1</TD><TD>.</TD><TD>BAMAEDGE</TD><TD>came </TD><TD>2</TD></TR><TR><TD>Hobart</TD><TD>6</TD><TD>.</TD><TD>WHISKEY FELLA</TD><TD>came </TD><TD>13</TD></TR><TR><TD>Hobart</TD><TD>8</TD><TD>.</TD><TD>MUSQUI</TD><TD>came </TD><TD>9</TD></TR><TR><TD>Coonabarabran</TD><TD>5</TD><TD>.</TD><TD>UNIKIN</TD><TD>came </TD><TD>2</TD></TR></TBODY></TABLE>

[michaelg]

Thanks for the method, wesmip1.

Shouldn't the positive expectation be 89%, which is very high seeing that all the selections are at single figures, and the results from the past 3 months are very impressive. Just for comparison, is it possible to look at the results assuming the selections had been layed for the same stake ($20?), and not to a liability of $100?

The combined ranking of 21 - I assume it is a minimum of 21.

To have a very low rating in the CF algorithym would more or less mean the selection is resuming from a spell. I have always been uncomfortable laying these, but maybe the poor record of the trainer compensates for this. Have you looked at replacing the CF category with CP or another category?

From today I'll have a look at your method, and if there are any selections I'll lay them.

[Raven]

could anyone enlighten me as to what the achronyms represent? CF, CP, etc

[michaelg]

Raven, these are algorithym categories on the neurals which is to be found on the Racing and Sports website.

CF is current form.
CP is career performance.

[michaelg]

Yesterday's sole selection was beaten for a profit of 0.95 units.

Today's selections are:

Bunbury
4/4
7/4

Gosford
2/5

Launceston
7/8

[Bhagwan]

Heres a Lay plan that is showing strong results.

RULES
Target TAB Nos. 4,5,6,7

Target the first one going down the list of these four Nos. that won LS

Ignore resumers.

Must not be Fav.

Must be <=9.80

Bet to 5% Fixed Liability, good for 20 outs.
E.g. $5 Liab per $100 bank.

[wesmip1]

Michaelg,

Shouldn't the positive expectation be 89%, which is very high seeing that all the selections are at single figures, and the results from the past 3 months are very impressive. Just for comparison, is it possible to look at the results assuming the selections had been layed for the same stake ($20?), and not to a liability of $100? The positive expectation is worked out using this formula:

(1-(StrikeRate)*0.95 )-(StrikeRate)*(Average odds of all Runner)

Where strike rate is your lay strike rate ie. 194/217 = 89.4% (ie there were 23 winners).

It tells you your edge ... which is 6% using this system.

Laying to the same stake makes the following return : $20 stake returned $1225.60 after 5% commission.

The better way to lay is to a fixed risk though as that is true level stakes for laying.


The combined ranking of 21 - I assume it is a minimum of 21.
Yes that is right. 21 is accepted as a bet.

To have a very low rating in the CF algorithym would more or less mean the selection is resuming from a spell. I have always been uncomfortable laying these, but maybe the poor record of the trainer compensates for this. Have you looked at replacing the CF category with CP or another category?
You could be right. In fact of the (now 218) selections there were 140 had their last start 28 days or more before todays race. If you assume using unitab X as resuming then there were 112 selections resuming.

There was 1 selection today:

Seymour R8 Chilled by Choice ... it won though. Ranked 8th CF and 15th for trainer.

Feel free to add more filters when you look at the selections. Adding 1 more filter takes the results from a 6% edge to a 12% edge. Adding 2 filters takes it to a 21% edge. Its just about finding the right filters to look through. The filters i use are unitab rating != 100 and not a Distance winner.

Have you looked at replacing the CF category with CP or another category?CP dropped the selection count to 146 and gave a positive 0.01% expectation and showed a $ profit of $598.50

[michaelg]

Thanks, wesmip1.

Its certainly something to look at.

[wesmip1]

There were 2 selections yesterady under $10:

ALITTLEVOLCANIC 2nd
ARCLACE 8th

[wesmip1]

todays only selection was JUSTA BUSTA. Came 4th.

[Raven]

I just had a brief look at the 12m ratings at Doomben. Looking for lay opportunities, I feel we should only look at horses whose 12m rating is at least 4 kg below the best in the race & ranked outside the top 4.

Why 4kg? Mark Reads old adage that 2.25 lengths equals a 'grade', so 3.5kg roughly equals a grade when considering the class & weight ratings.

I was looking at horses that you could lay at less than $20. I found 13 in Brissy for 1 winner around the $5 mark, so a profit today. In race 7 at Syd there was Domitian, Godfather & Verminator, all unplaced.

Another filter might be needed. As for price, I dont mind laying at $20-$30 just so long as your not offering a premium over the TABs. Someone suggested that you multiply Tab price by 1.2 & if you cant lay at that price or less, move on. Thats the sort of rule i will incorporate.

I cant get fussed with the neurals though.