Test System

[michaelg]

Over the past week the following system has been showing good profits with Win and Place betting.

1) Placed last three starts unbroken by a spell
2) Last start finish better than previous start.
3) Single figure in the pre-post market
4) No Maidens.

We'll see how it goes today, there are quite a few qualifiers.

Flemington
2/9
3/9
6/2

Hawkesbury
1/4
4/5
5/9
7/7
8/1

Albury
3/1

Rocky
5/2

Pinjarra
1/5
7/5

[thorns]

The amount of ideas you come up with, one wonders, do you ever sleep? http://www.propun.com.au/racing_forums/images/smilies/wink.gif

[michaelg]

Thorns, the gray cells are always ticking over, maybe even when I'm sleeping, but not necessary about the horses.

From today's eleven selections there were three winners for a return of $9.60, and $6.14 for the Place.

However, what intrigued me was Pinjarrra R1 no5. It ran second, paying $1.04 on NSW TAB ($1.50 on Unitab) for the Place yet I was able to get $1.93 on Betfair. This price was more or less constant throughout betting so there was no luck or good timing in obtaining the $1.93.

This got me thinking about the TAB/Betfair comparison but only with skinny prices. Since last Saturday I've looked at all horses that were $2.30 or less in the Telegraph market. There were about 40 of them for about a 92% strike rate. They showed a small profit on the TAB, and even more surprising, Win betting showed an even better profit.

If betting for the Place on the TAB for these "supposedly" hot faves (some paid $1.04, whilst one even paid $2.00) produced a profit, then judging on today's Pinjarra scenario the profit with Betfair since last Saturday should be quite pleasing.

From tomorrow I'll Place-bet on Betfair everything (providing I'm at a computer) whose Tele price is at $2.30 and less. If it proves profitable I'll report in a few days.

[michaelg]

I've done some more research on this Place method.

I've randomnly chosen 12 day's racing and checked the results using NSW TAB divvies.

There were 107 selections for 96 placegetters for a total return of $115.36. Ten placegetters paid $1.04.

This is an overall strike rate of 89.7% and a POT of 7.8%.

[michaelg]

I'll list today's selections for the method as shown in the opening post.

Flemington
4/2

Dombeen
3/4
4/2

Warren
5/5

Goulburn
1/2

Gold Coast
7/2

Ascot
7/1

[wesmip1]

Michaelg,

Ran a quick test for you over the last 3 months on WIN betting only.

Now normally I say it shows a loss and not worth continuing but in this case the results were:

301 selections
70 winners
return 307.60 returnd (unitab)
return 351.86(betfair).

Thats a nice rreturn.

A couple of suggestions which you should follow up (just keep track of an make the decision to include these based on your own research)

It made a better profit if the horse ran 1st last start rather then 2nd.
It made most of its profit with 2 and 3 year olds.
A horse which previously won at the course/track makes a profit. Those that didn't made a loss.
Barrier 1,2 and 3 made a large loss.
All the profit came from horses stepping up in distance.


Hope that helps you identify a few areas to keep track of when following the selections.

[michaelg]

Thanks very much for that, wesmip1. It's very encouraging.

I don't want too many filters because it reduces the action, so I'll use the best of them.

1) Won last start.
2) Going up in distance.
3) Not starting from barriers 1, 2, or 3.

The following are today's selections using the above filters.

Flemington
4/2 and 4

I failed to include no.4 - it won, paying $3.30 and $1.50. So it should have been a winning day for Win betting. It drastically reduces the amount of selections, but as its profitable I''ll stick with it.

Thanks again.

[wesmip1]

michaelg,

As I said just monitor the filters ... they may end up restricting the betting too much. But was just giving you a few ideas as the base system hits a losing run on things to look at.

Remember the base system idea you came up with was showing a good profit using betfair figures so maybe no filters are needed.

[michaelg]

wesmip1,

I will monitor them as you suggest, and I will list the selections that the filters eliminate but will show them as "Lay Selections", with the reason why - this should make it interesting.

Today's Lay Selections:

Warren 5/5 - not going up in distance. Beaten.
Goulburn 1/2 - barrier 2. Beaten
Dombeen 3/4 - not going up in distance. Won, paying $1.70
Dombeen 4/2 - not going up in distance. Beaten.
Gold coast 7/2 - not going up in distance. Yet to be run.
Ascot 7/1 - barrier no.2. Yet to be run.

[michaelg]

I've had another look at my the method and it seems that if the selection is in the top three faves in the pre-post market they perform very well. I don't have the exact overall figures for my original week of testing but know they were in profit. So I checked yesterday's results.

Flemington
4/2 - unplaced
4/4 - won, paying $3.30 and $1.50

Dombeen
3/4 - won, paying $1.70 and $1.20

Ascot
7/1 - won, paying $2.20 and $1.20

Goulburn
1/2 - ran second, but no Place divvies.

Gold Coast
7/2 - won, paying $4.60 and $1.70.

This is 4 winners totalling $11.80 from 6 selections and 4 placegetters from 5 selections for a return of $5.60.


So from today I'll list those that are in the top three in the Tele market, and see how they go.

Armidale
4/6
7/3

Cranbourne
8/6

Out of interest there is another selection but it is not in the top three in the pre-post market. It is Ararat R3 no.4.

In an earlier post I mentioned that almost always much higher Place prices for red hot faves can be obtained on Betfair then the TAB. This morning Zenyatta paid $1.04 on the TAB yet on Betfair it fluctuated between $1.40 and $1.55.

[partypooper]

Very interesting thread Michaelg, me being particularly interested in place betting and any new ideas. My problem has always been (like many others) as soon as those filters come in to increase the S/R the action is reduced again.

I'm working on one at the mo which produces a bet in EVERY RACE and stands at break even, I'm just going through the process of applying 1 filter at a time to see if I can tilt it into profit without cutting out too many bets, I will post the idea when I can get it into profit.

I am always amazed at some of the place divvies (differences) when backing best tote , its not uncommon to see 1st tote its not uncommon to see 104c on one tote and 1.70, 1.80 etc on another.

By the way, I have a data base covering 3000 favs, all opened at LESS than 2-1,

Best tote = break even, so applied a minimum reqd. price like so: win odds divided by 4......x 1.5 (i.e. 50% overlay) and then applied it to BF odds available,

the result? a bit disappointing, about 1.5% POT needs more work.

[michaelg]

Party, those are not bad results. Hopefully you will discover a filter that turns it into a profit.

There were no winners from yesterday's three selections, but two of them placed for a return of $3.30.

Today's selectionsare:

Queanbeyan
6/2

Colac
1/3
8/1

[bernie]

michaelg, why not:

Colac
7/2
8/3

[michaelg]

Hi, Bernie.

I only look at those that have placed at their last three starts. Both of them ran fifth at their third last start.

They are both in the top three pre-post faves so I would not be surprised if they run well. If they do, I might look at those horses that ran fifth or better at their third last start. This would also provide more action.

[thorns]

Very interesting thread Michaelg, me being particularly interested in place betting and any new ideas. My problem has always been (like many others) as soon as those filters come in to increase the S/R the action is reduced again.

I'm working on one at the mo which produces a bet in EVERY RACE and stands at break even, I'm just going through the process of applying 1 filter at a time to see if I can tilt it into profit without cutting out too many bets, I will post the idea when I can get it into profit.

I am always amazed at some of the place divvies (differences) when backing best tote , its not uncommon to see 1st tote its not uncommon to see 104c on one tote and 1.70, 1.80 etc on another.

By the way, I have a data base covering 3000 favs, all opened at LESS than 2-1,

Best tote = break even, so applied a minimum reqd. price like so: win odds divided by 4......x 1.5 (i.e. 50% overlay) and then applied it to BF odds available,

the result? a bit disappointing, about 1.5% POT needs more work.
Have you had a play with Barrier draw as a filter? Thats one of my party tricks that I am working with at the moment, and some of the results are very interesting to say the least. As we know, some barrier draws are far overbet comapred to others.

May not work for short priced favs, but something to consider if you haven't already.

[bernie]

Thank you for your prompt reply michaelg. Silly me, I must have been dreaming because I didn't pick up on the 5th placings.

Thanks for the great work you are doing.

Bernie

[michaelg]

Still, it might be worthwhile checking it out.

I made a mistake - Colac 1/3 is not a selection as it's last start is not better than the previous.

Sorry!

[michaelg]

Yesterday's selection placed paying $1.50.

Two selections today:

Cessnock
8/8

Townsville
3/2

[michaelg]

There was one winner from yesterday's two selections, paying $4.50 and $1.90.

I made a mistake with Monday's Place divvy - it paid $1.80 and not $1.50. I was going to correct it a couple of minutes later, but the Edit facility was not available.

Today's selections are:

Warwick Farm
6/1

Seymour
5/4
6/1

Ascot
7/6

[michaelg]

From yesterday's four selections there was one winner of $6.50 and two placegetters totalling $3.50.

Today's selections:

Bendigo
8/4

Rocky
5/1
6/4

[michaelg]

I haven't been checking the system but will see how it goes today:

Sandown
7/2
8/1

Eagle Farm
5/7

Morphetville
1/4
3/10
4/2
7/12

Inverell
4/3

Ascot
6/2

Darwin
4/2
5/1