Testing A Multiple Selection System

[michaelg]

Over the past few days I recorded the results of a system that selects half the field for races that meet certain criteria.

I used NSW TAB divvies betting at level stakes and the loss was just under 10%. I've now decided to look at S.P. divvies also with level stakes. The results of this show a healthy profit. So I'll to list today's selection and see how they go.

Albury
4/ 1, 3, 4, 5, 6, 7
5/ 1, 2, 3, 6, 7, 9
7/ 3, 4, 5, 8, 11, 16

Coffs
4/ 1, 2, 3, 4, 6, 10
5/ 1, 2, 4, 7, 8, 10, 13
6/ 1, 2, 5, 7, 10, 11, 14
7/ 1, 2, 3, 5, 9, 12, 16
8/ 1, 2, 3, 4, 5, 7

W'bool
4/ 1, 2, 3, 4, 5, 10
7/ 1, 2, 3, 4, 6, 11

A total of 63 selections.

[The Ocho]

:eek: WHOA. Good luck with that many selections :)

[michaelg]

From yesterday's 10 races there were 9 winners. The method suffered a loss of $26.55 betting level stakes of $5.

The LOT was 8% but with the TAB it was 17%. However, if the selections had been dutch-bet then there would have been a profit of 13%.

The method has had a success rate of about 82% and would have won every day of testing if the selections had been-dutch bet. Yesterday would have been the worst day even though it would have been profitable. Overall, dutch-betting is showing almost a 30% POT. But to bet this way one has to be at a computer which I'm not prepared to do. However if today shows a better result at dutch-betting than level stakes then I might have to consider it.

Today's qualifying races are:

Townsville
4/ 1, 3, 4, 6, 7, 12
6/ 2, 4, 5, 6, 8, 9, 13

Kyneton
7/ 2, 3, 4, 7, 10, 12

Queanbeyan
5/ 2, 3, 5, 6, 8, 10
6/ 2, 3, 5, 6, 7, 8
8/ 1, 3, 4, 6, 7, 8

[Shaun]

A betfair API product could handle the dutch betting side of thing easily

[Try Try Again]

Hi Michaelg,

Have you considered only Dutching the qualifying horses that open (use Betchoice) <$10?

From yesterday's results this would have eliminated 20 horses without losing any winners.

You have not indicated what prices you are Dutching but this should improve your return with or without Dutching.

[michaelg]

Try, I'm Dutching all the selections according to Betfair's prices as shown on their Exchange just prior to the jump.

My target return is $100, and if the outlay is $80 or more then it's a no-bet race. This has only ever happened once with all the many races because one or two fancied horses are often not selected, and as the minimum number of runners per race is 11 this is a strong possibility.

I'm surprised how many selections blow in the betting and often if they win, the return is pleasantly high.

[The Ocho]

Hi michaelg. I'm a bit thick but can you please explain dutch betting? Is it where you are trying to win $100 with each selection thereby betting more on the fav (or lowest priced horse) and less on the higher priced horses or is it something completely different?

[michaelg]

The Ocho, there are different variations, but you've got the basics. I'm betting on a horse to win $100 less the stake. For example, if the horse's price is $20 or thereabouts I will bet $5 on it, and if it wins, the profit on that particular horse will be around $95. I do this for all the selections.

I've discovered that the price of the fave normally shortens after placing my bet, so if it wins I normally get less than expected, and if it eases in the betting then the increased price is usually negligible. However, the opposite is generally true for the others. From tomorrow I'll be omitting the fave but adding another selection to replace it.

This will mean the total stake is not as high, and the winner (providing it's not the fave) should hopefully pay more than expected.

[The Ocho]

No worries, thanks michaelg. :)

[michaelg]

Today I'll be backing the following (providing I'm at my computer) using Dutch betting but omitting the Betfair fave.

Newcastle
7/ 1, 2, 3, 4, 6, 8, 9, 11
8/ 1, 2, 3, 4, 5, 6, 10, 11

Sandown
3/ 2, 4, 5, 6, 7, 7, 8, 9
4/ 1, 2, 3, 4, 6, 11, 14, 16, 19
5/ 4, 5, 6, 7, 8, 10, 15, 16, 17
6/ 1, 3, 4, 9, 12, 13, 15, 17
7/ 4, 6, 7, 8, 10, 11, 15

Dombeen
5/ 1, 2, 3, 4, 5, 6, 9
6/ 4, 5, 6, 8, 9, 10, 15
7/ 1, 3, 7, 11, 12, 13, 14

Murray Bridge
8/ 1, 2, 3, 4, 5, 7, 12, 14

Kalgoorlie
4/ 2, 3, 4, 6, 7, 9, 10
6/ 1, 2, 4, 5, 11, 12, 13, 14
7/ 1, 2, 4, 8, 10, 11, 13
8/ 2, 4, 5, 8, 9, 11, 13, 14, 15


My records show this method would have been successful over the past week, but betting real money can be stressful due to the outlay.