1200 Metre Race Plan-150% POT

[Bhagwan]

Here`s a plan that produced 150% POT over 967 races , on all races , 7 days a week

1)The race distance has to be EXACTLY 1200 Metres only

2)Must be a last start winner by 1.80 Lngs or more

3)Barrier has to be 2-12 only

4)Field size can be anything except 10-12 starters(Do not bet on fields with 10-12 srarters, poor strike rate.)More is fine.

5) Days last start must be 12-20

6)Must be showing $4.50+ in newspaper pre-post market.

You usually end up with 90 selections per month. 7 Days a week.
If 2 or more selections back them all.

Strike rate is 17% so you will need a bank of $1.00 for every $100 bank .
It producers some very good payers.

[becareful]

For those of you following the "Why Do Most Systems Fail" thread this plan has some of the warning signs of backfitting rules to make a profit without having a rational reason for the rule. In particular rules 3 and 4. I can't see any logical reason for excluding barrier 1 or for excluding races with 10-12 runners apart from boosting the historical POT. I NEVER include a rule in any system unless I can justify it with a logical reason.

[Dr Pangloss]

Bhagwan

In complete agreement with be careful. The system as described reeks of curve fitting.

For example, distance of exactly 1200m suggests distances < 50m > will destroy the results. Last start winner margin > 1.8 lengths makes you wonder what results >1.5 lengths produced.

You can make data jump through hoops at the click of a mouse these days. I think Equine Investor used the term 'Fundamental Principle' in another thread. In my view a robust system contains a MINIMUM OF RULES based on Fundamental Principles with selection filters ( discretionary and non-discretionary) over-layed. Just my thoughts.

[Fryingpan]

Bhagwan

What do the results look like when you do add barrier one and you do include fields of 10-12 starters and perhaps include races that are 50m over. Maybe there is something worth learning for all of us.

[Equine Investor]

Bhagwan,

Try this little gem -

Last start winner within 7 days!

That's the only rule.


Think you may get a surprise.

:grin:

[Bhagwan]

You bet I used passed results for these figures & I went through the process of shoehorning the data to produce a profit .

I am not interested if you can`t see any logic in it , so my advice is DON`T use it.

What I would like to see from the detractors is their proposed plan on results based on the future before they have run , seeing that they beleive the past is irrellavent.

[Bhagwan]

Results for the 7 Day idea,
made a loss on turn over -60% , 1200 m lost the least of all the distances.

Barrier No.1 produced 60% loss on turnover.

Races with 10-12 starters in 1200M races along with the other rules produced a profit of 79% POT,

1250 Metre races produced a loss with the criteria.

[Bhagwan]

I guess one could ignore the 10-12 runners rule & back them anyway.
Which means , back races with 7-14 runners in 1200M races, following the previous rules.

[becareful]

"If you can't see any logic in it ... don't use it" - now that's my point exactly Bhagwan - glad to see you agree!

For the record I never said the past was irrelevant or couldn't be used to develop a system. What I did say is that if you are developing a system based on past results you must be sure your rules are logical and are not just made up to get the figures to show a profit. Unfortunately, as you say, these rules were specifically made up to get the past data to show a profit so it is very unlikely that profit will continue in the future (not anywhere near the 150% anyway). Do you really believe that Barrier 1 is a such a huge disadvantage that will turn a 150% POT into a 60% loss? Are last start winners able to count and work out there are 10 other runners so they should not win? Is there really such a big difference between 1200M and 1150M or 1250M that would cause a huge variance in results?

I have previously posted a method for using past data to develop and then verify a system - you might like to try it as it as it does help you get a more realistic idea of whether a system is valid or is simply a statistical anomoly in the data.

[Equine Investor]

60% LOT ?

Did you apply filters to this 7 day rule Bhagwan?

I am shocked at the results, as this does not reflect my results.

My results -

Total stakes: 500 units

Total Returns: 759.28 units

Profit 259.28 units

You get many horses either favourite or second favourite. However, you do get 14/1 shots etc.
Last winner I can remember is Energy Abounds at a country meeting and it ran second or third at Caulfield I think yesterday. Excellerator was also 7 day system selection when it won in Sydney and paid $6.00 win.



<font size=-1>[ This Message was edited by: Equine Investor on 2002-10-17 08:45 ]</font>

[The Phoenix]

I have been looking at a couple of issues related to the underlying two principles of the current discussion, ie races at 1200 metres and last start 7 day winners.

With respect to the 1200 m issue compared to 1250 m, my reviews have also shown that the 1200 m races are far more consistent than 1250 m, by the time you get up to 1350 m races, the results are getting worse. 1400m like 1200m seems to be a bettable proposition. 1600m - 1650 m were I thought until recently the most consistent distance however it has turned out that I have lost more money on these races than any other distance. Now the big question for 1200m and also barrier 1 issues.

I propose that the reason why these results occur and can be relied on are that many horses are used to running 1200 m and are also trained over a similar distance. The extra 50 m small as it is is therefore significant. 1400 m races are similar in that horses that have run good times over 1400 m - down towards 1.22 - often repeat the effort even in higher grades and often at good odds.

The barrier 1 issue for 1200 m races occurs because unless the horse has reasonable early speed, they can be crossed and then get progressivly shuffled back along the fence. At the end of the race, other horses are going well enough that even though they can not win, they block the run for the barrier one horse. Conversly, a front running horse is generally suited by barriers up to about 8 or 9 although ideally I like to see them drawn 3 or 4 to about 6 or 7. These barriers give them the best chance to bowl out of the barriers and take up a front running position without having to be hussled along as they sometimes have to be comming from Barrier 1.

On the issue of last start 7 day winners, as a group I looked at them for many years and blindly following them will make a level stakes loss. It is possible however to make a profit on them but I dont bother. What I do though is take great interest in horses that have had last two runs at <=10 days, <=7 days coupled with some very specific speed and class ratings figures. These are a very good group of horses to bet on as they are fit, have form and the owners and trainers have backed them up usually expecting them to peak.

regards

The Phoenix

[Equine Investor]

Phoenix,

How can my figures be so different to yours and Bhagwan? I think something is wrong here.
Maybe adding filters or something.
My figures are based on my last 500 bets using this system as a test -
Only using last start winner within 7 days - anywhere - no filters.
7 days qualified, 8 days not.

The only thing excluded were jumps, but applied to even country and metro meetings.
How this could go from a 50% profit to a 60% loss with a sample of 500 bets has me really puzzled.

:eek: :???:


<font size=-1>[ This Message was edited by: Equine Investor on 2002-10-17 11:22 ]</font>

[The Phoenix]

Equine Investor

the main difference between my figures and yours is most likely to be that I only consider Saturday Metro racing - 5 States.

At the moment due to having revised my database at the beginning of May this year, I have had to look at two different time periods with what you have proposed being back to 4.5.2002 for this year and also for an earlier period of 1991 to 1996.

For the current period - 132 unique ie only horse in race qualifiers. Level stakes win of 7 % POT however statistically suspect as deletion of the highest price winner leads to a level stakes loss.

For the earlier period 924 selections for a level stakes return of $ 999.87 ie a level stakes POT of 8.2 % which is pretty comparable. The larger sample also shows that the occassional higher dividend - $ 21 to $ 54 is where the profit comes from and there were 9 winners at >= $ 21.

I would not in any way suggest that you are wrong as clearly we have different populations. I am wondering though where you refer to "bets" whether you mean that you actually have had 500 bets or that it is a theoretical method.

In general terms, a level stakes POT of 7 to 8 % indicates that with appropriate filters there is probably a viable method within the general principle of 7 day last start winners. I however prefer higher profitabilities for the starting point.

It is quite possible that enough winners pop up in the country to make the difference but would not be any use to me as I could not be able to get enough money laid to make it worthwhile.

regards

The Phoenix

[Equine Investor]

1. Apologies to Bhagwan as I should have started a new thread for the 7 day discussion as it's a little off topic - sorry.

2. Thanks for that info Phoenix. Yes we are obviously looking at two separate sample periods. They were not actual "bets" of mine, rather I followed it on paper daily - not retrospective.
And yes the average country dividend is a little higher, as last start 7 day city winners seem to pay less - on average.

[shy]

EI, I just ran winner in 7 days through Price Predictor and it came up with the following based on NSW Divs from July 1 2002 to Sept 28...

Meetings considered : 457
Races selected : 245
Horses selected : 257
- Number that won : 51
- Strike Rate/Race : 20.8%
- Strike Rate/Seln. : 19.8%
- Place Strike/Seln.: 46.3%
Return at $1/Seln. : $256.40
- Average Win Div. : $5.03
- Absolute Profit : $-0.60
- Profit on Turnover: -0.2%

I noticed as it was running through that it was showing a 50% profit until July (the program tests backward... so Aug & Sept were 50%+ POT).

There were lots of good divs though so I added 4 filters to knock out the better supported runners and got the following...

Meetings considered : 457
Races selected : 34
Horses selected : 35
- Number that won : 10
- Strike Rate/Race : 29.4%
- Strike Rate/Seln. : 28.6%
- Place Strike/Seln.: 45.7%
Return at $1/Seln. : $92.00
- Average Win Div. : $9.20
- Absolute Profit : $57.00
- Profit on Turnover: 162.9%

A lot less selections... but crikey!!

[Equine Investor]

Thanks shy, confirms my earlier calculations were spot on for the selected period.

And your filters would be?

Just kidding!

:wink:

[trench digger]

Last start winner 7 days
1200m only
Syd/melb only
no change in distance from last start
dec 2000-oct2002

RACES BET 20

RACES WON 3 RACES PLACED 12

STRIKE RATE 15% STRIKE RATE 60%

POT 24% POT 35%

same rules as above but expand distance range to 1200m-2400m

RACES BET 97

RACES WON 23

STRIKE RATE 23

POT 35%

[dinodog]

hi shy

what race length and runners did you run on your first system check

i did the straight win 7 days and had very dissimilar results


regards

dinodog

[shy]

Hi dinodog

all race lengths and all runners... just straight out last start winner, max 7 days