System update?

[Sportz]

Hi guys. Could someone possibly help me with a couple of system tests???

Bhagwan helped me create this system last year:

(1) Saturday metro races in Qld, NSW and Vic (delete jumps races)
(2) Fast, Good or Dead tracks only
(3) Field size 12-16
(4) At least 4 starts since a spell
(5) Finished in the first 4 at each of it's last 4 starts
(6) At least 1 win in last 4 starts
(7) Finished 1st or 2nd last start
(8) Won at this track
(9) Allotted weight: 48-55 kgs (ignore apprentice claims)
(10) Between 3 & 7 years old
(11) At least 10% wins
(12) Last start no more than 21 days ago
(13) Between $3 and $12 in pre-post betting

Bhagwan tested it over the previous 4 1/2 years or so (I think) and it produced a profit of around 75%. I must admit that for one reason or another, I didn't keep up to date with the results though and I'm just wondering if anyone can do another test to see how it's gone in say the last 12 months???



Also, I wonder if someone could do a test on this one. It's a variation on my old Number 1 system and it went VERY well over a 3 month period last year (August - October), but again, I haven't kept up to date with how it's going now.

(1) All TAB meetings.
(2) No 2yos, jumps, maidens or races less than 1300m.
(3) At least 8 starters after scratchings.
(4) Consider Number 1 providing it is clear pre-post favourite between $3.00 and $5.00
(5) Must have finished in the first 4 at each of it's last 4 runs.
(6) Must have had at least 4 runs since a spell.
(7) Must have won at this track at this dist (c) or be a last start beaten fav (b).
(8) Last start no more than 21 days go

[Chrome Prince]

Hi guys. Could someone possibly help me with a couple of system tests???

Bhagwan helped me create this system last year:

(1) Saturday metro races in Qld, NSW and Vic (delete jumps races)
(2) Fast, Good or Dead tracks only
(3) Field size 12-16
(4) At least 4 starts since a spell
(5) Finished in the first 4 at each of it's last 4 starts
(6) At least 1 win in last 4 starts
(7) Finished 1st or 2nd last start
(8) Won at this track
(9) Allotted weight: 48-55 kgs (ignore apprentice claims)
(10) Between 3 & 7 years old
(11) At least 10% wins
(12) Last start no more than 21 days ago
(13) Between $3 and $12 in pre-post betting

Bhagwan tested it over the previous 4 1/2 years or so (I think) and it produced a profit of around 75%. I must admit that for one reason or another, I didn't keep up to date with the results though and I'm just wondering if anyone can do another test to see how it's gone in say the last 12 months???



Also, I wonder if someone could do a test on this one. It's a variation on my old Number 1 system and it went VERY well over a 3 month period last year (August - October), but again, I haven't kept up to date with how it's going now.

(1) All TAB meetings.
(2) No 2yos, jumps, maidens or races less than 1300m.
(3) At least 8 starters after scratchings.
(4) Consider Number 1 providing it is clear pre-post favourite between $3.00 and $5.00
(5) Must have finished in the first 4 at each of it's last 4 runs.
(6) Must have had at least 4 runs since a spell.
(7) Must have won at this track at this dist (c) or be a last start beaten fav (b).
(8) Last start no more than 21 days go

Hiya Sportz,

The prepost rules preclude me from testing the systems for you, as I don't have prepost records in the database, this maybe something I implement in future, as I've found some fanastic edges all over the place ;)

But it's going to take months of work and research to get them in there.

Do you know of an online archive which has prepost prices?

[Sportz]

Sorry Chrome, no I don't.

[Chrome Prince]

Sportz,

If you have anything you want tested that doesn't involve prepost or unitab ratings, I'm more than happy to test it for you.

[Sportz]

Sure thing. Thanks.

If you want, you could try out that first system without the pre post price rule and see if it could actually produce a profit without that rule.

[Surround]

Hello Sportz
I'm new to this forum game, so please bear with me.
I ran your rules as best I could with your 2nd system from July 2004 to now.the rules as I put them in the data base follows:

RACE RULES:
exclude if Class = Hurdle, Maiden, Steeple
exclude if Age Rest = 2YO
exclude if Field Size between 1 and 7

HORSE RULES:
include if TAB number = 1
exclude if Spell Count between 1 and 3

FORM RULES:
include if Days Ago last run between 1 and 21
include if FinPos last run between 1 and 4
include if FinPos 2 runs ago between 1 and 4
include if FinPos 3 runs ago between 1 and 4
include if FinPos 4 runs ago between 1 and 4

RATING RULES:
include if Morning Line Rank = 1

STATS RULES:
exclude if Distance Wins = 0

The win POT (based on Unitab divs.) = 8.7%
Longest losing win sequence = 10
Win strike rate = 34.0%

Hope this helps.

[Surround]

Sorry Sportz
I left out the rule of fav between $3.00 and $5.00.
That improved the bottom line greatly from 8.7% to 32.1%.
The losing sequence dropped to six but the strike rate dropped to 26.3%.
There were only 38 selections over the period (15 months), so it is very selective.
By the way Sunday's had 10 selections for no winners.
Take Sunday's out and you have 79.3% POT LLWS = 4 S/R = 35.7%

Good luck with whichever way you decide to go.
It does show promise but requires patience, something I'm lacking.

[Sportz]

Hi Surround,

Welcome to the forum. Thanks for that.

Go back to rule 7. You've only included distance winners, whereas I wanted horses which had won at this distance AT THIS TRACK. (not sure if you could test for that) And I also wanted to include last start beaten favourites (they don't have to have won at either distance or track). And you've also left out the distance rule of 1300m+.

Anyway, never mind. Thanks very much for your help.

By the way, in the 3 month test period last year, I got 14 qualifiers for 8 wins, 13 placings and a P.O.T of about 140%:
Aug 7 E.Fm R8-1 Strong Suit UNP
Aug14 G.Cst R6-1 Common Knowledge PLA $1.40
Aug21 Caul R1-1 Umbula PLA $1.50
Aug28 Rose R2-1 Canny Princess WON $4.40/$1.50
Sep 1 Sand R7-1 Batarshi WON $4.10/$1.50
Sep26 Hbt R9-1 Coalfields WON $3.30/$1.50
Sep27 Alb R5-1 Mary's Flag WON $5.70/$2.20
Sep27 P.Mc R7-1 Rodeo Chute Out WON $3.70/$1.50
Oct 4 Belm R4-1 Zero Engagement WON $2.20/$1.50
Oct 9 Chel R7-1 Lord Mick PLA $1.70
Oct14 Moe R5-1 Tortuga WON $5.70/$2.10
Oct16 E.Fm R2-1 Rue The Saint PLA $3.20
Oct22 Mack R2-1 Larrikin Prince PLA $1.40
Oct26 Tar R6-1 Desert Shaker WON $5.50/$2.30

[Chrome Prince]

Sportz,

Without reference to the prepost, it looked pretty good, but does not throw up many bets.

2000 to 2005

270 bets
38 wins
97 placings
281 win return
238 place return

11 units win profit
32 units place loss

4.07% win pot
-11.85% place pot

But take away the two longshots
$20.70
$21.60

And it's not that promising. The low strike rate with such good filters worry me a bit.

This was based on metro races only I should say.

[Sportz]

Chrome,

Bhagwan's test period was over 4 1/2 years leading up to the beginning of this year and with the pre-post rule in, he got:

148 bets
21% SR
71% POT

That was including fields of 17 which I deleted because they didn't go as well as 12-16 runners.

[Sportz]

This was based on metro races only I should say.

Yes, this system is for Bris, Syd, Melb on Saturdays only.

[Chrome Prince]

Yep, the prepost rule would have ruled out a lot of the longies included in mine.

You might see if he can tell you the longest two priced winners.

I reckon some of those $12 prepost might be the $20 plus winners.

[Surround]

Sportz
There you go. told you I didn't have too much patience.
Chrome has a much larger data base to work from so his analysis would be better than mine.
Anyway I plugged in the other rules I'd missed and this is the outcome.
Races 1300m and above = 14.% POT LLWS = 9 S/R = 34.1%
Must have won on track = 13.1% = 9 = 35.1
Btn favourite last start = 17.4% = 10 = 28.8 from 66 selections

I just plugged those rules in to the previous rules I'd applied.
I was wondering if you might be being too selective as the earlier one that I'd inadvertently mucked up showed a lot of promise.
I think you could do something with it.

Good luck

[Sportz]

Hey, no worries. Just throwing a few ideas around. It was purely a more restrictive version of another system I had on the forum a while ago, and as you can see, it worked very well in the test period I had last year. I simply wanted to see how it's been going lately.

[Dale]

Bhagwan helped me create this system last year:



(12) Last start no more than 21 days ago


my old Number 1 system
(7) Must have won at this track at this dist (c) or be a last start beaten fav

Hi Sportz,

Both systems look promising but....

Why does everyone include rule 12) in their systems,imo its a myth,unless someone can show me a comparison of win % for those that started no more than 21 days ago & those over 21 days i will continue to totaly disregard it.



In the 2nd system i dont get rule 7),could you explain the reasoning behind asking for something as specific as won at track and distance and then turning around and ignoring that rule just because the horse was a last start loser at short odds.

[NANOOK]

Days since last run these are the results I get for this year

[marcus25]

In the 2nd system i dont get rule 7),could you explain the reasoning behind asking for something as specific as won at track and distance and then turning around and ignoring that rule just because the horse was a last start loser at short odds.

Hi Dale!

"(7) Must have won at this track at this dist (c) or be a last start beaten fav"

If you read this carefully you will find that this statement is not a mutual exclusion, but an extra inclusion instead.
Ie. if it had not won on this track AND distance, it must have been a last start beaten fav. at least, to be included!
Cheers.

[crash]

Sportz,

The low alloted weight is possibly a poor rule and having to have had 4 starts might be more suited to 1600m and up races. The 4 start rule will also collect a lot of selections that are past their peak.

4 to 8 starts only
54kg to 57kg
1600m to 2200m

Might offer a good improvement.

[Sportz]

In the 2nd system i dont get rule 7),could you explain the reasoning behind asking for something as specific as won at track and distance and then turning around and ignoring that rule just because the horse was a last start loser at short odds.

This system is simply an improved version of another system I previously used. In the original system, there were four different plans. One of them required a win at this track & dist. Another required a last start beaten favourite. Then there were two other plans which did not do so well, so I simply deleted them and kept the first two.

[Sportz]

Why does everyone include rule 12) in their systems,imo its a myth,unless someone can show me a comparison of win % for those that started no more than 21 days ago & those over 21 days i will continue to totaly disregard it.


Well, Bhagwan worked out the first system and I assume that he found out that 21 days or less worked out the best. With the 2nd system, there were not a huge number of horses outside the 21 day range, but removing them improved the profitability in the period I tested it.

[Dale]

Nanook,

thanks for the figures,

THE FIRST 3 LINES OF SP FOR HORSES THAT HADNT RACED WITHIN 21 DAYS REURNED A PROFIT!

THE FIRST 3 LINES OF SP FOR HORSES THAT HAD RACED WITHIN 21 DAYS DID NOT RETURN A PROFIT.

I REST MY CASE!




Marcus25,

yeah i understood the rule but i didnt get why the beaten favorite bit was included.



Sportz,

Thanks for the explination but it smacks of retro fitting to me,ie it wasnt based on what works best over a lengthy period of time but was introduced because a few good winners fitted the rule at the time of inclusion.

[Sportz]

Yes, but I included the 21 days or less ONLY because it was a rule that I often used anyway, and deleting horses outside that range did seem to improve the situation in the test period, so why not.

It's only "retro fitting" if you introduce rules which you wouldn't normally use or which have no real logic behind them, purely because it makes the past figures look better.

[partypooper]

This is meant only as constructive comment, I would have thought that the 21 day rule is self explanantory i.e.,. surely this relates to the horses fitness??
Why 21 days ?? .... why not 22 or 19 or 18, well I guess you just have to draw the line somewhere. Today I ruled out a place bet in PR2-2 Brockeys Ace purely on the 21 day rule (thankfully) but then missed PR4-5 Attadale on the same rule so there you go!

[Chrome Prince]

When relating to Metro class horses the 21 days refers exactly to three weeks.

Good horses don't race at flemington on saturday and then murtoa on friday in 20 days (usually). That's why the rule in general Weds to Weds, Sat to Sat etc.

When applying this rule it has to be flexible to account for 19,20,22,23 days based on the individual circumstances.

I still maintain horses stepping up in grade are a lot better value and do better than those stepping down. How many city placegetters go to the country and go around half the price they should be and run nowhere.

A real mugs bet in general.

[Dale]

Sportz,

We seem to have different veiws on retro fitting,i maintain that if you base your rules on what works in a certain test period and not what will be best for long term profit then the system is doomed to failure.


Partypooper & Chrome,

You talk about fitness and good horses like they are the most important factors in this game,what about having a profitable edge or angle that most people are ignoring,surely that is more important,

Nanooks figures prove that the 21 days rule is a myth!

[the rostram]

well i turned 21 21 yrs ago does that count

[darkydog2002]

21 +21 = 42 = 4 + 2 = 6

So today in every meeting its race 6 horse 6 .

WHAT???????

Cheers.
darky.

[crash]

Surely the 21 day rule would be dependant on whether the horses are sprinters, middle distance horses or stayers and what class of races are being targeted. quick turnarounds of 7 to 14 days seem to favor sprinters and it tells the punter one important fact, there have been no 'issues' with the horse since it's last start and it's fit and ready to go [again].

Blanket stats covering everything don't suit all systems and indeed all race types either. They can be quiet misleading. Barrier positions are a classic example of misleading stats. For all races they suggest there is no real inside barrier advantage [minute anyway]. Break them down to field sizes and I'd rather a draw from 1 to 7 in a 16 or 18 field sprint.

[Sportz]

Oh no Crash, not that damn 1-7 thing again. :D

I feel exactly the same way about that rule as Dale does about the 21 day rule. I have no argument with anyone who wants to delete horses from wide barriers. I just can't work out what's so special about barrier 7 that he is left in while barrier 8 is out. Weird.

Similarly, I guess you could say why is 21 days allowed but 22 days not. Well, as Chrome suggested, when concentrating on Metro races, the 21 days refers to exactly three weeks, Sat to Sat, or even Wed to Wed. Now, it may not always work out, but over the years, I've pretty much stuck to that 21 day rule.

If you have a mechanical system with set rules, you have to have a cutoff point somewhere. I guess that is Crash's answer to the barrier 7 thing too. :D

[Sportz]

When applying this rule it has to be flexible to account for 19,20,22,23 days based on the individual circumstances.


I guess that's the difference between a mechanical system which you stick rigidly to and a general set of rules which you can vary slightly depending on the situation.

[crash]

Your right Sportz,
you have to have a cut off point somewhere and that's where mechanical systems fall down somewhat. When it comes to days since last start or barrier draw and I'm doing the form on a race [my preferred mode of betting as most here know], I take into account what suits an individual runner. Is it a swooper [outside half of the barriers please]?, an on-pacer or leader [inside half please]? and that too will depend on the track, race distance, field size and track cond't. Mechanical systems can't do that.

The days since last start rule depends on the horse. I do a history check and see what the trainer has decided suits the horse. If I see regular quick turn-a-rounds and then see it is suddenly 21 days since it's last run, my 'didn't pull up well or there has been an injury or some other issue', alarm bell rings. Usually I have saved my money by not betting the horse. Some horses run only once a month because that suits them.

I think any system can be improved immensely by doing a little investigating on the selections it throws up [that's why I advocate handicapping as the best selection system of all]. The last rule of any system should be: 'Handicap the selections'. It doesn't defeat the purpose, because the purpose is money in our pockets.
Yesterday was a classic example of how that works. 5 selections for one of my systems. A quick check showed one selection with an excellent win rate at 1000m and 1100m was trying for the 1200m for only the second time [hopeless first time]. The runner was 5yrs old. Only the 2nd. time at 1200m?
I didn't back it and it was unplaced. Another runner was coming back to 1200m from 1350m. No history of winning coming back in distance so no bet and unplaced. 3 only bets for 1 win at $5.40 and 2 seconds from 3 bets rather than 1 winner from 5.

There are no rules that can be added to any good system to let you know when a mechanical selection is a stupid bet. A quick handicapping of the selections sure can. Sometimes I get it wrong but not often, as the the reason for ruling out any system selection is generally easy to spot, not deep and meaningful. "!0 starts at this track and never placed" ? Cross that selection off as a bet. Obvious things. So why do systemites insist on Lemming like bets just because the system tells them too? Dumb and dumber or haven't got a clue about handicapping?
When I bring up the idea of [quick] handicapping system selections as a great way to improve SR, I am usually deafened by the silence, so I suspect the latter of the above reasons mostly applies. It's an observation, not a criticism.

Cheers.