Staking _

[This section gives advice about the second 'leg' of Profitable Professional Punting.
...............................THE STAKING PLAN.................
It is my contention, despite my "mathematical training" that optimised staking plans can increase returns on a long term basis by as much as 15%.]


First for those "following the Wizard thread".  If you do not know what this is go to the systems page and start from there.
From the Selections page we had the following:

"A" bets
Melb 2/1
Syd 7/2
"B" bets
none
"C" bets
Melb 4/1
Bris 2/1
Adel 8/2
"D" bets
Syd 1/1
Syd 4/1
Adel 6/6
Syd 8/2
Melb 8/2

About the average distribution I would expect.
I am going to follow two staking plans.  The
specials Plan I will follow myself. The Standard Plan is for those that want a little more action, but bet more conservatively.

The Specials plan will bet on A and B selections only.
We will simply start with a bank of 1000  units and bet a progressive 3% of progressive bank, on A selections and 2% of progressive bank on B selections.
Unless there are less than two selections.
Any day that has only one selection we will bet 5% of bank on that sole selection.
Any day that has no selections we will bet 5% per each selection the following day.
With these selections over the last six months this staking plan had 65 bets and increased a 1,000 unit bank to 1,522
therefore the potential is to at least double the bank each 12 months.
The month end targets become
Jan 1052 Feb 1108 Mar 1184 Apl 1249 May 1318 June 1410 July 1488 Aug 1570 Sep 1678 Oct 1771 Nov 1869 Dec 2000.
Do that for six consecutive years and you will have 64 times your original bank or 64,000 units with a 32,000 unit profit in the sixth year.
Just think of that in $2 units after  five to six years your bank grows from $2,000 to $100,000.
From then on you leave the bank at $100,000 and take the $100,000 a year profit. Sounds incredible but it is achievable.

So saturday 6th two "A" bets no "B" bets so start with 30 units.
Melb 2/1 bet 30 units no return .- progressive bank 970.
Syd 7/2 bet 29 units return 52.2 units- progressivbe bank 993.2   Loss for the day  6.8 units


The above Staking Plan proved dissappointing over the first two months January February 2001,
showing only 4.2% on Turnover so we adopted the  +50-50 method.
Start with 1 unit if loss add 50%  round up halves.
If winner reduce 50% round up halves.
E.G. bet 1 lose then bet 2 if bet of 1 wins then bet 1 again.
if bet of 2 loses then bet 3 - if 2 wins then bet 1.
if bet of 3 loses bet 5 - if 3 wins bet 2.
if bet of 4 loses bet 6 - if bet of 4 wins bet 2
if bet of 5 loses bet 8 - if 5 wins bet 3.
if bet of 6 loses bet 9 - if bet of 6 wins bet 3.......etc etc.


The Standard Plan will work on all bets.
A bets 5 units + 1 for each win that day to date.
B bets 4. units + 1 for each win that day to date.
C bets 3  units + 1/2 unit for each win that day to date.
D/E bets 2  units + 1/2 unit for each win that day to date.

Syd 1/1 bet 2 units  return 3.60   prof profit  1.6
Melb 2/1 bet 6 units return 0  prog loss 4.4
Bris 2/1 bet 2 1/2 units return 0  prog loss 6.9
Melb 4/1 bet 2 1/2  units return 5.5 prog loss 3.9
Syd 4/1 bet 3 units  return 0 prog loss 6.9
Adel 6/6 bet 3 units return 7.8 prog loss 2.1
Syd 7/2 bet 8 units return 14.4 units profit 4.3
Melb 8/4 bet 4 units return 0 prog profit 0.3
Syd 8/2 bet 4 unit return 7.2 prog profit 3.5  profit 3.5 units.   Profit for day 3.5 units  YTD profit 3.5 units


You will be able to follow these methods in "The Newsletter"

Here we will explore the world of Staking Plans, this logically comes after deciding on a "Selection Method" as the Optimum Staking Plan for a Selection Method that has a "Strike Rate" of 20% with an average Dividend of $6.00 has little resemblance to the Optimum Staking Plan for a "Selection Method" that has a win strike rate of 40% and an average Dividend of $3.00 despite the fact that both have the same "Profit On Turnover" of 20%.



It has long been my contention that Staking Plans are as important, if not more  important than the selection systems.
These days computer power has made form analysis an almost perfected science.  The more monied professional punters can afford the best in computer support.   The horses the better informed professionals support become favorites by means of the weight of their money. See where this is all leading ?.
Let us assume for a moment that this computer power was absolute and always selected the best horse in a race.
Let us assume that the professionals used this computer power and bet on every race.
Then only the "Chaos Factor" would prevent them making 100% of winners favorites or another way only the chaos factor would prevent 100% of their selections from winning races.
Of course the "Chaos Factor" that allows a horse with a "real" chance of  winning of  33% to defeat a horse with a 40% chance of winning is much less than that required for the 33% chance to turn the tables on a "real" 60% chance.
Therefore if there are 5 horses in a race with "real" chances of  40%  - 25% - 20% - 10% - 5% then the 40% has most chance of winning and the 25% the second most chance and so on.  So that with the computers accurately assessing the chances of horses and only being upset by the "Chaos Factor" you would expect that the percentage of winning favorites would increase, albeit at reducing prices, and the chances of horses other than the top three in the betting would be waning.
Indeed this is very much the case.
In the 12 months ending 1/8/98 all Sydney - Melbourne - Brisbane - Adelaide Saturday races
31% of favorites won,  18% of second favorites and 13% of third favorites, so that the first three represented 71% of winners.
But for the three months to 27/3/99 these figures became Favorites 35% second favorites 22%  third favorites 16%, the first three now representing 73% of winners.
Since the average field size was 14.4  20% of the average field represents 2.88 horses or say  3.
So we are fast approaching the old dependable 80/20 rule with 80% of the winners in the defined 'best' 20% of the chances. This indicates to me that the computer's identification of the main chances have gone about as far as they can go.
It also strongly suggests that the current markets accurately indicate the 'relative' chances of the contestants, that is to say the favorite has the best chance of winning, the second favorite the next best chance etc.
Another consideration is that the effect of random hazards has been scientifically evaluated at 50%, so that if the computer analysis reached the 80/20 perfection it would after the addition of the "hazard "  or "chaos" factor correctly predict the outcome of 40% of the races, and did so in 36% of actual races in the last three months on Australian Metropolitan courses.
All of this tells me that computers have it as right as they are likely to get it.
That the Professional Punters using the computer analysis will make the most likely winners the favorites.
All of this tells me that all successful punting methods or systems are going to come up with a mixture of the same three top horses in almost all races, and that 80% of races are going to be won by these system selected horses, that this percentage is unlikely of significant improvement, due to the "chaos" factor.  that these three top selections will win races in proportion to their position in the market, and they will continue to be in the proportion of approximately Favorites 5 to second favorites
3, to third favorites 2.
Over the last eighteen months (July 97 - Jan 99)also whilst favorites strike rates have increased from around 30% to 35%, their average dividends have decreased from around $3.10 to around $2.80 so that the average return has grown from around 93% to around 98% a most significant factor.  Second favs around 18% at $5.00 (90%)to around 22% at $4.20 (93%) and third favs around 13% at $7.00 (91%) to around 16% at $5.70 (91%).
This increase in returns particularly of favorites is of particular interest, as I have always claimed ( see below) that a good staking plan can turn a 30% strike rate or better with a 90% return or better into a return of 105%+.
I believe that if we can identify the 'true' favorites and get say a 35% strike rate at an average div of $2.60+, that I can turn that into a steady 10% profit with a sound staking plan.



 
 

I believe that if we can identify the 'true' favorites and get say a 35% strike rate at an average div of $2.60+, that I can turn that into a steady 10% profit with a sound staking plan.
I made that statement above sometime in 1998.
Throughout 1999 I have been refining my Ratings - see article in systems page.
For the calendar year 1999 I have achieved a strike rate of 35% and an average dividend of $2.67 whilst this shows a flat stake loss of 6.7% according to my statement above I should be able to come up with a staking plan to make 10% POT
well I have in the form of the WINSTEP  staking plan.  Which has shown a POT of between 10 and 20% averaging around 14-15%. 


THE WINSTEP STAKING PLAN

The winstep staking plan was born from the realisation that loss chasing plans were just that you chase your lossess and lose the lot.  Why not chase the wins??
The other problem with the loss chasing ideas is the rapid increase in bet size, brought about by tring to get it all back in one shot
Now there are staking plans that have divisors to address the large bet size problem.
They take a divisor - 6 seems to be popular for some or no reason - which means you achieve your target when win odds add to 6. That is a 6/1 or better winner - two at 3/1 - 3 at 2/1 or whatever.
The problem here for plans with a 35% strike rate is that average divs are around $2 - $3 and 6/1 winners are rare indeed.
So why not design a method that in fact depends or winning sequences - providing you can achieve the sequences often enough, and would it not be nice if at the same time you could achieve the target with a single longer priced winner.
Well the Winstep achieves all of the above.
To keep the bet size under control we do not divide by the popular 6 but by a conservative 10.
Now with a 35% strike rate we can expect 3 or more winners in a row around once in 25 bets
 Actually the ratings in the last 200 bets have produced the following sequences:
                                   2 - 8    times  - expected 12
                                3 -  3    times  - expected  5
                                4 -  3    times  - expected   2
                                5 -  1    time    - expected   1
So the ratings with a 35% strike rate with an expectation of producing 3 or more winners 8 times in 200 bets actually achieved 3 or more winners on 7 occassions.
So this should work - and it did.
So we use a divisor of 10 and seek to get out target with three consecutive winners.
How do we do this??
We chase our winnings rather than our losings.
When a winner is struck our next bet is doubled, if two winners are struck we double up again, and when three winners are struck we double up yet again.  Now providing we are getting better than even money this is not an all up approach but a bet increment that retains some of the winnings and chases the target at the same time.
The target will be attained if we strike any of the following:

                        A single winner at 10/1.
                        Two consecutive winners at average odds of 5/1.
                        Three consecutive winners at average odds of 5/2.
                         Four consecutive winners at average odds of  5/4.

In the last year the target was attained nine times, the last time on 18th December, at year end we were 7% behind target. Had made POT of 15.7% for the year achieved a bank turnover of 2.3 times and consequently increased our bank size by 36% for the year.  My target was 33% on bank and was over achieved.

So what are the rules for this staking plan??
1. The first rule is do not use this plan with any system with a strike rate below 30%.
2. Decide on a weekly required return.  Let us keep this in "hobby" rather than my terms say a target of $50 per week.
3. Divide this by the average number of bets per week - with a strike rate of 30% plus this would be unlikely to be more than say 7 and more likel;y in the 4/5/6 area - let us take 5.
4. Set a bet target goal of 2 divided by 3 that is $50 per week over 5 bets is $10 per bet.
5. So our "Target" will be the number of bets year to date minus current cumulative profit.
6. Now divide this "Target" by our safety divisor of 10 and that is our base bet size.
7. Multiply the base bet size by the win accumulator factor to arrive at the actual amount bet.
      That means that we will bet the base bet until a winner is struck
      After a winner the next bet will be base bets times 2.
      After two consecutive winners the bet will be base bet times 4.
      After three consecutive winners the bet will be base bet times 8.
      After four or more consecutive winners the bet will be base bet time 10.

        But what actually happens is after two or three winners you end up passing the target
      and then you have a "negative target".
      When the target becomes negative and whilst it remains negative the base bet  is increased to
       the minimum bet - the minimum bet is usually 1/4 of the bet target = $10/4 = $2.50.

So to illustrate with the first week of the 1999 equine year 7th August 1999.

An exceptionally busy week for a system that averages 5 bets per week - there being 11 bets.

First bet Melbourne race 2 No 3 the "Target" is $10 there being no accumulated losses at this stage.
So the base bet is $1 which we raise to the minimum and bet $2.50. the horse is unplaced.
Next bet is Adelaide race 2 no 1 the Tasrget is $22.50 ( 2 nd bet @ $10 per bet plus accum loss of $2.50)
Base bet is target divided by 10 = $2.25 raised to minimum of $2.50. The horse is unplaced.
Third bet is Melbourne race 3 no 1  target is now 30 + 5 = 35 base bet  35/10 = $3.50 A winner paying $2.20
We win $4.20 and are now "down"  80 cents.
Fourth bet is Brisbane Race 3 No 1  target is now 4 times 10 plus 0.80 = 40.80 base bet 40.80/10 =$4.08 BUT WE JUST HAD A WIN SO WIN FACTOR OF 2 MAKES BET  $8. Horse is unplaced.
Fifth bet is Melbourne Race 4 no 8 target is 5 times 10 plus 8.80 =58.8 base bet is 5.88 win factor now back to 1.
Bet is $6 Horse is unplaced
Sixth bet is Adelaide Race 4 no 1 Target is now 6x10+14.8=74.8 base bet 7.48
Bet is $7.50 horse wins and pays $2.10 we win $8.25
Seventh bet is Sydney Race 5 no 7 target is now 7x10+6.6=76.6 base bet = 7.66 win factor 2
Bet is $15 horse wins and pays $2.50 we win $22.50
Eighth bet is Adelaide Race 5 no 1  target is now 8x10 = $80 LESS our accumulated winnings of  $15.90 = 64.1
base bet is 6.4 win factor is now 4.
Bet is $25 horse wins and pays $2.40 we win $35 - cumulative winnings now $50.90
Ninth bet is Sydney Race 7 no 5  target is $90 less winnings of $50.90 = $39.10 base bet is 3.91 win factor is 8.
Bet is $31 Horse is unplaced.
Tenth bet is Brisbane race 8 no 2 target is $100 less cum wins of $19.90 = 80.1 base bet is $8 no win factor
Bet is $8  horse beaten in photo.
Eleventh bet is Adelaide race 8 no 1  target is $110 less cum winnings of $11.90 = 98.1 base bet is $10 no win factor
Bet is $10 horse loses we end up $1.90 in front for the day
Total out $119 total returns $120.90  POT  1.7%
Had we being using flat stake of $10 per bet outlay would be $110 for a return of $92 losing $18 or 16%.

Of course you can set up a spreadsheet to do all of the calculating, although if you go to the local pubtab or the course and do not have a portable then you still have to do it manually but it is not difficult.

From 1/8/99 to 31/12/99 I will do a week by week report on this combination of the Turf Accountant Ratings and the Winstep Staking Plan with the criteria of $10 per race and minimum bet of $2.50.
I would recommend a bank of $7,500 for this - with around 250 bets per annum our year target would be $2,500 profit per annum (33%).
A reasonable plan would be to plow back profits and increase target bet size by 25% each year.
Say $10 - $12 - $15 - $20 - $27 this way in 5 years you would have a bank of around  $20,000
and be making around $120 per week.
In ten years that would grow to a bank of $50,000 with a profit averaging  $300 per week.
From little acorns bloody big trees grow if you plow it back.

In an attempt to prove or disprove the theories two paragraphs back,  in early 1999 I analysed three commercially available selection methods and three of my own selection methods, each of which promised 30% or better strike rates, and 100% or better returns, with flat rate staking for win, place and eachway, and my own win, place, and eachway staking plans.
The results tabled below, astonished even me.
 
 SYSTEM   STAKING   INVESTED   RETURN  PROFIT/LOSS   P . O . T .    MY 
 RANKING		  RANK POSITION
MAJIK OZZIE   $3,480.00     $6,263.00      $2,783.00       80% ...*****++       1 st.
MAJIK  FLAT-WIN    109 @ $10. 
   $1,090.00 		    48 
  $1,752.00 		     $ 662.00      61%     *****       2 nd. 
DISPEC  OZZIE   $4,870.00   $8,168.00     $3,298.00       68%      *****        3 rd.
MAJIK  $1000/MONTH WIN     $5,175.00     $7,827.00     $2,652.00       51%      *****        4 th 
MAJIK    100/20+10   $6,950.00    $10,475.00   $3,525.00      51%      *****        5 th
MAJIK  FLAT- 
EACHWAY		 109 @ $5 E/WAY 
  $1,090.00		  87 
  $1,653.00		      $563.00       51%        ****         6 th
DISPEC    100/20+10   $4,430.00     $7,092.00   $2,662.00         60%       ****          7 th
MAJIK  E/WAY ESC.   $7,000.00    $10,921.40    $3,921.40        56%       ****          8  th
MAJIK  SATAN   $10,876.00    $15,722.50    $4,846.50         45%       ****          9 th
MAJIK FLAT - 
PLACE.		  109 @ 
  $10 Place. 
  $1,090.00		   87 
  $1,548.60		    $ 458.60          42%       ****          10 th
These Top Ten combinations rated 4 stars or better, a 4 Star rating means that by investing 1% of bank per base bet
you would make a profit of 150% of bank in one year, or better, and have less than 1% chance of "going broke"..
 
 
 SYSTEM STAKING   INVESTED RETURN PROFIT/LOSS  P . O . T .   MY 
 RANKING		 RANK POSITION
BUST  100/20+10 $2,050.00 $2,316 $266.00       13%       ***      11th
DISPEC  $1000/MONTH WIN  $7,380.00  $10,450.00 $3,070.00       42%      ***      12th
DISPEC  FLAT-STAKE WIN   83 @ $10 Win. 
 $830.00		   25 
 $1,165.00		 $ 335.00      39%        **      13th
BUST  OZZIE $1,140.00 $2,071.00 $931.00      82%        **      14th
MADIS  OZZIE $2,380.00 $4,071.00 $1,691.00      71%         **      15th
BUST  $1000/MONTH WIN $3,640.00 $6,445.00 $2,805.00       77%        **      16th
MADIS  FLAT-STAKE WIN  43 @ $10 Win. 
  $ 430.00		  19 = 
  $ 779.00		  $ 349.00      65%        **      17th
These Second Seven combinations rated 2 - 3 stars.
3 Star rating means that by investing 1% of bank per base bet you would make a profit  in excess of 100% of bank in one year, and have less than 5% chance of "going broke".
2 Star rating means that by investing 1% of bank per base bet you would make a profit of 100% of bank in one year, and have less than  10% chance of "going broke."
 
 
 
SYSTEM STAKING  INVESTED RETURN PROFIT/LOSS P . O . T .   MY 
 RANKING		 RANK POSITION
MAJIK $200 a day Place Plan. $10,700.00  $13,393.00 $2,693.00       25%        *      18th
BUST  FLAT-STAKE EACHWAY  41 @ $5 e/way 
  $ 410.00		    22 = 
  $ 488.40		  $ 78.40       19%        *      19th
MADIS   100/20+10 $2,890.00   $4,594.00 $1,704.00       59%        *      20th
DISPEC  E/WAY ESC. $25,120.00  $31,551.60 $6,431.60        26%        *      21st
MADIS  $1000/MONTH WIN $4,670.00  $7,687.00 $3,017.00        65%        *      22nd
MADIS  FLAT-STAKE EACHWAY   43 @ $5 e/way 
   $ 430.00		  30  = 
  $ 663.00		   $ 233.00       55%         *       23rd
DISPEC  FLAT-STAKE EACHWAY   83 @ $5 e/way 
  $ 830.00		   47 = 
  $ 1,057.00		   $ 227.00      57%          *       24th
MADIS   E/WAY ESC. $3,760.00  $5,439.20 $1,679.20       45%          *      25th
MUST  $1000/MONTH WIN  $4,180.00 $6,661.00 $2,481.00      59%          *      26th
BUST    41  FLAT-STAKE WIN   41 @ $10 Win. 
  $ 410.00		  13 = 
$ 556.40		   $ 146.40       32%           *      27th
MADIS   SATAN $3,880.00   $4,535.00 $655.00       17%           *       28th
DISPEC  FLAT-STAKE PLACE   83 @ $10 Place 
  $ 830.00		  47 = 
 $ 949.40		   $ 119.40       15%           *       29th
WHOS  $1000/MONTH WIN  $9,915.00  $12,493.50 $2,578.50        26%            *        30th
MUST  FLAT-STAKE 
  WIN.		  27 @ $10   Win.. 
  $ 270.00		   14 = 
  $ 390.60		   $ 120.60       45%            *        31st
WHOS  OZZIE  $3,280.00 $4,076.00 $796.00       24%            *        32nd
MUST  OZZIE $460.00    $ 620.00 $160.00        35%            *        33rd
MADIS  FLAT-STAKE PLACE   43 @ $10 Place. 
  $ 430.00		   30 = 
  $ 546.00		   $ 116.00         27%            *         34th
All of these combos deserve their Star Rating - some more so than others - each showing in excess of 10% POT
with less than 25% chance of  "going broke".
 
 
 
SYSTEM STAKING  INVESTED RETURN PROFIT/LOSS    P . O . T . MY 
 RANKING		 RANK POSITION
MADIS  $200 a day Place Plan $10,235.00 $11,533.00 $1,298.00         13% 
BUST   E/WAY ESC. $6,080.00  $ 6,810.00 $730.00          12%
DISPEC   SATAN $5,340.00  $ 5,862.60 $522.60          10%
BUST   SATAN $2,780.00 $3,011.50 $231.50           8%
MUST  FLAT EACH-WAY  27 @ $5 E/WAY 
  $ 270.00		   19 = 
  $ 323.00		      $ 53.00          20%
DISPEC    $200 a day   Place Plan $14,785.00  $15,568.00 $783.00          5%
WHOS  FLAT-STAKE WIN   58 @ $10 
Win. 
  $ 580.00		  21 = 
  $ 657.30		   $ 77.30          13%
MUST   100/20+10 $2,060.00  $2,333.00 $273.00          13%
MUST   E/WAY ESC. $1,560.00   $ 1801.00 $241.00            15%
WHOS   100/20+10 $4,350.00   $4,671.00 $321.00           7%
MUST   $200 a day   Place Plan $6,470.00  $6,745.00 $275.00             4% 
MUST FLAT-STAKE 
 PLACE.		  27 @ $10 
  Place 
  $ 270.00		  19 = 
  $ 290.70		   $ 20.70           7%
WHOS  FLAT-STAKE EACHWAY  58 @ $5  E/WAY 
   $ 580.00		 41 = 
$ 594.50		   $ 14.50           3%
BUST  FLAT-STAKE PLACE  41@ $10 
Place. 
  $ 410.00		  22 = 
 $ 420.20		    $ 10.20           3% 
AND THE LOSERS:
BUST   $200 a day 
Place Plan		 $11,460.00  $ 11,455.00   - $ 5.00
WHOS   E/WAY ESC. $7,560.00   $7,546.40   - $ 13.60
MUST    SATAN $1,975.00  $ 1,842.00   -$133.00          - 7%
WHOS  $200 a day Place Plan $20,585.00   $19,422.00 -$1,163.00        - 6%
WHOS  FLAT-STAKE PLACE  58@ $10 Place
  $ 580.00		   41  = 
  $ 528.90		     - $ 51.10         - 8% 
WHOS   SATAN $4,375.00   $3,834.50   -$540.50         - 12%
This is the "balance" of the survey for completeness sake - I could not, nor would not recommend any of these.
 

Today 26th April 1999 I am posting on the Systems Page full details of the MAJIK selection method.
On this the staking page you will find below full details of the OZZIE staking plan, and on the Money Management Page
there is now a table relating these two together.
I will do this with the three systems MAJIK MADIS and DISPEC,  and their combination with the various staking plans - the
three win plans    -  100/20+10, The Ozzie One Step, The $1,000 a month Plan,
the two place plans - the $200 a day plan, and SATAN,
and the Eachway Escalator Plan.


 

                                                    The Ozzie One step staking Plan
 

Most - not all, good staking Plans are simple, and this is one of the simplest.
It is a win only Plan, and it is in the category of loss chasers.
It requires a marginal long term Win POT,  just 1%  will do, in fact it can turn marginal long term losses into profits.
Just bet one unit for a win.
If the bet loses increase stake by one until win struck, that is in a losing streak bet 1 2 3 4 5 6 7 8, when a win is struck revert the odds.
This means that if you are at an 8 unit bet and you have a 3/1 winner the next bet is 8 units (the last bet) minus 3 (the odds of the winner ) = 5.
If the bets require an outlay of 10 or more units then revert twice the odds on a winner until below 10 again.
for example our next bet is 5 units as above we strike 7 more losers and then a 5/1 winner the bet sequence is:
5 6 7 8 9 10 11 12 (5/1 winner) next bet 12 - 5x2 = 2 units.

I have just completed a twelve month study of staking plans, and I am extremely pleased with the results, which I will now share with you.
Last year I was involved with a Group as a kind of "coach" in successful punting.
The Group got together each week for a three hour "session" and then about half of us - not always the same members - would get together at the local club on Saturday afternoon to put the theory into practice.
It was Melbourne Cup day 1997 and we were cheering like mad as the commentator called one of our
Trifectas Might and Power - Doriemus - Linesman  -  our celebration soon stopped when the third placing went elsewhere.
Then the postmortems went on for some hours, during this time one of the Group once more brought up my claim that if a system had a 30% or better strike rate and better than $3.00 average dividend I could construct a staking plan to turn the -10% into around +10%.
Then someone suddenly realised that there was a very easy system that fulfilled the 30% strike rate - $3.00 criteria - just back all the Saturday Metropolitan Favorites!!
Well how could I refuse the challenge??

So the challenge was set starting Melbourne Cup Day 1997 to Melbourne Cup Day 1998 Design a Staking Plan that will show a profit backing every Metropolitan Favorite.
So I got out the databases and the analysis tools and came up with three suggestions, each of which was to be monitored for the full twelve months.
After much discussion some extra rules were set.
1.  Maximum bet $100.
2.  Bank available $2,500 per  system.
3.  Not only Metropolitan Favorites Saturdays but also Public Holidays.
    ( this resulted in 1,824 races over 61 meetings).
Within a matter of a couple of months one bank was lost and that Plan laid to rest.
However, the other two Plans - the almost exact opposites of eachother both succeeded.

These thoughts and ideas are the basis of both plans.

The plans came from years of statistical analysis.
After many years of trial and error and study of results, I have developed a staking strategy that can usually turn, up to 10% loss systems, into profit plans.
This I believe comes from realising that the long term POT, is the long term strike average
 - times -  the long term average dividend -  less -  the long term average bet.
To turn losses into profits or increase profits we need to do one of four things.
Either raise the long term strike rate.
Increase the long term average dividend.
Increase the average win bet size
Decrease the average loss bet size

The first matter is a factor of the selection method and beyond the control of a staking plan
The second is a factor of the market and also beyond the control of a staking plan.
So we have to concentrate on the latter two items.

Now the old double-up strategy does this.
By doubling up on each loss and ending the series at the "peak" when a win comes, the win
bet is obviously the highest bet in the series and this both raises the average win bet size and lowers the
average losing bet size.
BUT it takes an enormous bank.

However - for the long term strike average to be what it is - or to become again what it was -
then inevitably at some undefinable point a lower than average strike rate sequence or period will - eventually - be followed by a higher than average strike rate period.

The solution is in:
(i) Having the sequences of reasonably short duration,
(ii) In increasing the base stake slowly over the longer losing sequence,
(iii) Limiting losses in long losing sequences,
(iv) Using the stake increase to accelerate the recovery when it comes.

The "key" is the long term strike rate average, and the pattern of short term strike rates,
that in turn, make the long term strike rate.
A simple illustration of the point:
Sequence   L L L W  L W L L L L L L W L L W  L W L L W W  L W L W
 L L L L L L L W L L  W L L W W .
Strike rate 13/40 or 32.5%
What if we designed a staking plan that increased bet size when the short term strike rate
was below par and reduced bets when the short term strike rate was above average.?
Let us say for ease of illustration that each winner paid $3.00.
13 winners @ $3 =$39  for loss of $1 or -2.5%.
Now after two bets strike rate is 0% so increase to stake of 2
after three bets strike rate is 0% stay at 2.
after four bets strike rate is 25% - below average stay at 2.
after five bets below average stay at 2.
after six bets strike rate is 33% and above average go back to 1,for bets 7 and 8
and then because of lower strike rate increase to 2 for bets 9 10 11
then 4 units for bets 12 and 13 after bet 13 strike rate is still low stay at 4 units for bets 14 15 16.
after bet 16 strike rate still down around 25 stay on 4 units for bets 17 18 19
strike rate now over 25 but below 30 bet 2 units bets 20 21 22.
strike rate now 33 % revert to 1 unit bet bets 23 24 25 26 27 28.
now strike rate has fallen bet 2 units bets 29 30 31.
then 4 units bets 32 33 34. After 34 bets strike rate above 25 below 30
bet 2 units bets 35 36 37  after 37 bets strike rate still below 30 bet 2 units bets 38 39 40.
So now bet return was -1 -1 -2 +4 -2 +4 -1 -1 -2 -2 -2 -4 +8 -4 -4 +8 -4 +8 -4 -2 +4 +4 -1 +2 -1 +2
-1 -1 -2 -2 -2 -4 -4 +8 -2 -2 +4 -2 -2 +4 +4  = +2 so a 2.5% loss becomes a 5% gain.
Now this is a simplistic illustration but is the basis of the first Loss Chasing Favorites Plan.

The "Rules" then for the first Favorites Plan are as follows:

Consider the Plan as a "Stepping Loss Chasing Plan".

The "Steps"  are $10 - $20 - $40 - $80.
This represents the "Bet target" for each bet at that step.
One stays on a step until the "Step Target" or "Stop Loss" level is reached.
The Step Target is 5 times the Step Size.
The Stop Loss is 4 times the step number times the Step Size.
The "Bet Target" is always the next step.
That is all there is.

To explain:

The first "Target" is $10.
Why?
1. The first step size is 10.
2. We stay on step 1 until we reach either $50 (The Step Target = 5 times the step value)
                                                       or -$40 (The Stop Loss for step one = 4 times 1 (step 1) times $10 ( the step value).
3. Since the step size is 10 then the steps are -40 -30 -20 -10 0 10 20 30 40 50 there are no others
in this step range. ( total 10 steps)
4. We start with Progressive total 0.
5. The next step is 10.
6. So our bet size is to bet to win 10.

Say our first bet is at 2/1 we bet $5 and lose.
Our progressive total is now -$5.
Our next step is 0.
We bet to win $5.
The horse is at 3/1 - we bet min bet of $2.
The horse loses we are now progressive total -$7.
Our next step is still 0.
We bet to win $7.
The horse is at 6/4.
We bet $5 - the horse loses now -$12.
Our next step is now not 0 but -10.
We bet $2 on 1/1 winner.
Our progressive is now -10 our next step is now 0 again.
Say we win again and our progressive is +$2
Our next step is now +10 .. and so on.

If we come to a progressive total of -$40.
We now go to step 2.
We stay at Step 2 whilst the Progressive is between -$200 and +$100.   ??

Well we start step 2 at -$40 and our stop loss on this step is 4x2x20(4x step number (2) x step size (20)) =$160   so start -$40 and lose $160 = -$200.
and the Step Target is 5x20(step size) = $200.
So the range of step 2 is -$200 to +$100
So Step 2 sub steps are -$200 -180 -160 -140 -120  -100  -80 -60 -40 -20 0  +20 +40 +60 +80 +100.
(16 steps)

Likewise Step 3 ranges from -$840 to +$200 in steps of $40 (27 steps).

And Step 4 from -$3,300 to +$400 in steps of $80 (54 steps)
** Note in the "Experiment"  this changes to -$2,500 to +$400 (38 steps) due to limit of $2,500 imposed by rules
of the challenge.

The Plan as I said was based on statistical analysis and most particularly the coefficient of correlation, standard deviation,
and the "natural" observation that over time the actual performance "oscillates" around the long term average.
In this case the long term average strike rate.
The "natural" effect is the observation that the further from the "mean" line the actual deviates the stronger the
probability that the actuals are about to change trend and head back to the trend line.

The actual results over the 1824 bets trended as follows:

Bet 1 start at Step 1 until bet no. 46 when progressive was -$43 and step 1 was 'lost' go to step 2.(this step -43)
                               until bet 141                               +$108 and step 2 was "won" revert to step 1.   (+151)
                              until bet 166                                +$160 and step 1 was 'won' revert to start of step 1.(+52)
                              until bet 204                                +$117 and step 1 was 'lost' go to step 2  (-43)
                              until bet 220                                -$43   and step 2 was 'lost' go to step 3.   (-160)
                             until bet 256                                 +$421 and step 3 was 'won' revert to step 1.(+454)
                             until bet 277                                 +$377 and step 1 was 'lost' go to step 2. (-44)
                             until bet 307                                 +$213 and step 2 was 'lost' go to step 3.(-164)
                             until bet 347                                 +$626 and step3 was 'won' go to step 1.(+413)
                            until bet 365                                  +$584 and step 1 was 'lost' go to step 2  (-42)
                            until bet 442                                  +$426 and step 2 is 'lost' go to step 3    (-158)
                            until bet 594                                  +$828 and step 3 is 'won' revert to step 1.(+402)
                            until bet 638                                  +$789 and step 1 is 'lost' go to step 2  (-39)
                            until bet 651                                  +$942 and step 2 is 'won' revert to step 1.(+153)
                            until bet 666                                  +$897 and step 1 is lost go to step 2  (-45)
                            until bet 780                                  +$1,054 and step 2 is won revert to step 1(+157)
                            until bet 788                                   +$1,014 and step 1 is lost go to step 2  (-40)
                            until bet 809                                  +$1,235 and step 2 is won revert to step 1  (+221)
                            until bet 825                                  +$1,190 and step 1 is lost go to step 2  (-45)
                            until bet 946                                   +$1,033 and step 2 is lost  go to step 3  (-157)
                            until bet 1022                                 +$1,418 and step 3 is won revert to step 1 (+385)
                            until bet 1050                                 +$1,374 and step 1 is lost go to step 2 (-44)
                            until bet 1112                                 +1,213 and step 2 is lost go to step 3 (-161)
                            until bet 1625                                 +$1,618 and step 3 is eventually won revert to step 1 (+405)
             ###         this long saga was really in two parts
                            to bet  1464                                   +$822 (the bottom of the graph) (-391**)  unfinished..
                            to bet 1625                                    +$1,618 (+$796).revert to step 1.
                            to bet 1779                                    +$1,677 and step 1 is won go back to start of step 1 (+59)
                            to bet 1801                                    +$1,633 and step 1 is lost go to step 2 (-44)
                            at bet 1824 still in step 2                 +$1,644 + 11 and continuing ..........
 
 

So lots of observations:

The $1,644 profit on a turnover of $18,932 for a POT of 8.7% is unlikely to make anyone rich.
I would anticipate this system going broke once in 100 series which would be about 7,500 bets,
the Gross Profit in 7,500 bets should be around $6,600 and you would lose $3,300 once for
a net profit of around $3,300 on say $75,000 turnover - POT  4.4% -  hardly exciting - BUT - this is based on a "system"
of backing favorites which had a strike rate for that period of  32% at an average dividend of $2.97 (TAB)
for a flat stakes loss of  4.7%.
I am confident that this approach applied to a break even or better selection method with a strike rate around 30% will produce good safe and steady profits.

############################################################################################

My expectations and the results:

I expected 80% of the series to go to step 2 in fact 92% did so.
I expected Step 1 to be successful 20% of the time in fact it was only 16%  2 from 12.
I expected Step 2 to succeed 50% of the time in fact it was 44% - 4 from 9.
I expected 30% of the series to go to step 3 in fact only 19% did so 5 from 27.
I expected step 3 to succeed 80% of the time in fact it was 100% 5 from 5.
I expected  11% of the series to go to step 4 - in fact none did.
I would expect step 4 to succeed 90% of the time.
In other words I would expect the plan to fail 10% of 11% of the time that is 1.1 times in 100.

###############################################################################################

   My expectations come from statistical analysis mixed with 'gut' feelings and years of experience
but as you can see after 50 years I am getting most of it pretty right.
Based on my analysis of the tendency of the actual short term trend to come back to the long term trend,
I anticipated a 10% chance of a winning sequence being followed by another winning sequence, but a 33% chance of a winning sequence following a losing sequence.  but a 70% chance of a winning sequence following two consecutive losing sequences, and almost 90% chance of a winning sequence following after a series of three losing sequences.
Our actual performance - by sequence or step-set as above -

L  W  W  L  L  W  L  L  W  L  L  W  L  W L  W  L  W  L L W L L W W L

    Let us take the "expectations" win followed by win expectation 10% actual  2 in 26 = 8%
                                            lost followed by win expected 33% actual 9 in 26  = 35%
                                           double loss followed by win expected 70% actual  5 in 5 = 100%
                                          treble loss followed by win 90% there were no treble losses.

I also anticipated 66% loss sequences, actual was 15 in 26 = 58%..
50% double loss sequences, actual was 5 in 13 (26 series = 13 doubles) = 38%.
30% triple loss series there were none.
10% quadruple loss series - there were none. 

.
So there is a summary of how a plan was developed.
What the expectations were and how the results stacked up to the expectations.
I think I got things fairly right and would appreciate feed back.

This bit is a bit deep but for those interested in my more "way - out" ideas on statistics:

I said above that I have based all of this on my interpretation of statistical analysis mixed with gut feeling and years of
observations.
What I was expecting was that as the" short term" - immediate losing steps since last win step - strike rate got further from the long term average the tendency to correct increases.

Look at the actual happenings in that light, bearing in mind long term strike rate 32%:
STSR = short term strike rate.
Change in Step.  at bet sequence number. step strike              prog strike rate   seq result
    1.                        46                             11/46 =   24%               24%            LOSS
    2.                      141                              36/95 = 38%                33%            WIN
                LOW STSR followed by HIGH STSR and WIN
                Progressive strike rate switched from below LTA (Long Term Average) to above LTA
   3.                       166                             10/25 =40%                  34%             WIN
        Two consecutive highs and wins not normal pattern.
   4.                       204                              9/38=24%                    32%             LOSS
        "Switch" as expected  - more especially after double high - LTA  re-instates.
  5.                        220                              0/16=0%                      30%             LOSS
                       70% Probability of Win to follow  since two losses and STA below LTA
  6.                        256                              18/36=50%                  33%             WIN
                          and it happens!!!!!!
                      Loss now odds-on.
 7.                        277                                 4/21=19%                  32%              LOSS
 8.                        307                                 4/30=13%                  30%              LOSS
                    70% Probability of Win to follow  since two losses and STA below LTA
 9.                        347                                 19/40=47%                32.0%           WIN
                              and it happens!!!!!!
                      Loss now odds-on.
10.                       365                                  3/18=17%                  31.2%           LOSS
 11.                      442                                  23/77=30%                31.0%           LOSS  Long slow loss poor divs.
                     70% Probability of Win to follow  since two losses and STA below LTA
 12.                      594                                  51/152=34%              31.6%            WIN
                        and it happens - but STA still below LTA so another Win 30%
   13.                    638                                  12/44=27%                31.3%           LOSS
         Loss and below LTA win now 50% chance..
 14.                     651                                  8/13=62%                   32.0%           WIN
 15.                     666                                   3/15=20%                   31.7%           LOSS
          Loss and below LTA win now 50% chance..
  16.                    780                                   46/114=40%                32.9%          WON
           This series more than 100 bets - series 100+ more reliable to get expected result.
 17.                     788                                   0/8=0%                        32.6%           LOSS
          Very short series STA still above LTA  - win to follow  only 30% chance.
  18.                    809                                    10/21=48%                 33.0%           WIN
          win and STA above LTA Loss to follow odds - on.
  19.                    825                                    4/16=25%                    32.8%           LOSS
              Loss but STA still above LTA win/loss  to follow 50/50
  20.                    946                                    32/121=26%                32.0%           LOSS
                          Two consecutive loss series but STA = LTA win to follow 50/50
   21.                   1022                                  33/76=43%                   32.9%          WIN
                          Win and STA almost 1% above LTA Loss to follow 80%
    22.                  1050                                  8/28=29%                      32.8%         LOSS
           Loss but STA still almost 1% above LTA Win to follow 40%
   23.                   1112                                  17/62=27%                    32.5%         LOSS
           Two consecutive loss series but STA STILL ABOVE LTA win to follow 45%
  24.                    1625                                   158/513=31%               32.3%         WIN
                         This was a "danger" series - in step 3 BUT STA above LTA -
                         Actually if you split this into two halves  bets 1112 - 1464
                         and 1465-1625  then 1112-1464 = 95/352 = 27% and STA falls below LTA
                         making it 70% certain that WIN will then follow
                        as it does in 1465-1625  =63/161=39% = WIN.
 25.                   1779                                     55/154=36%                 32.5%          WIN
                            Loss to follow now odds on
26.                    1801                                     4/22=18%                     32.1%           LOSS
            Loss but STA still marginally above LTA WIN to follow 50%

The 100/20+10 Plan.
 

There are no"Rules" as such for the second Staking Plan that proved profitable for all Saturday/Public Holiday Metropolitan Favorites over the twelve month period, but rather a Formula..
This Plan takes the complete opposite "Tack" instead of building up bets on Losses
It makes it's major build up on Wins.  The Plan relies on the occasional 3 or more consecutive win sequences to succeed, by waiting for these sequences and betting virtually "All-up" we bet small
until a winner is struck and thereby limit losses in the wait between "drinks".
The "Factors" in the formula have to be adjusted for the expected strike rates and expected "run of outs".
With the 30% strike rate I calculated the probability of 3 consecutive winners at 2.7% and allowed for my estimate of the 1/100,000 probability of 165 bets "misses" between trebles.
The Plan was then tested over the 1824 bets as above, and I would have expected a longest run of outs
as in the table below - compared to the actual happening.
 

                        Run of outs between trebles
R/O/O/           Frequency     Expected.             Occurred.

110+                            0                                    0
100+                            1                                    1      (101)
90+                              2                                    2
80+                              3                                    4
70+                              5                                    6
60+                              8                                    6
50+                            11                                    8
40+                            15                                   16
30+                            24                                   23
20+                            41                                   39
10+                            51                                   49
0+                              57                                   56
 

The degree of correlation surprises even me, but validates the 165 provision in the Plan.
 

The Formula is:
Next Bet = T/I*D*W.
Where:
           T = Targeted Return $.
           I = Insurance Factor.
           D = Dividend Factor.
           W = Win Sequence.

           T = Targeted Dollars = B + L*BT + CCL
            Where:
                        B = Target Base.
                        L =  Losing sequence.
                       BT = Profit Target per Bet.
                       CCL = Current Cumulative Losses this Current Sequence.

Sounds very complicated - IT IS - but it works.

"T" combines Profit and Loss chasing.
  If at a stage in the current sequence you have had 3 losing bets of say $5, $10 and $12, and you
have set "B" at $100, and BT at $10 (That is the long term Plan is to profit by $10 per bet). Then C will be $157, that is
the addition of the base target profit of $100, plus $10 per bet for the three bets, = $30 plus recouping the $27 loss this sequence to date. (Cumulative Losses)
T = 157.
Now I is the insurance factor - it is the realisation that we cannot win all our target each bet, as the bets would become too high to manage - so we go for a portion of the overall target each bet.
I use an Insurance Factor of 20. So I = 20.
D = Dividend Factor.
It is just a natural fact of the game that more TAB <$2 win than $2-$3 and even less >$4 win etc.
In the first Plan we increased our bets on the short priced/better chances by "Target Betting" here we use a Dividend Factor.
I only use 1 - 2  - 4 -KISS - but other methods could prove more accurate.
Factor 4 is used for any Win Divs under $3.
Factor 2 is used for Win Divs $3.00 - $5.00.
Factor 1 is used for $5+ Win Divs.
W = Win Sequence.
As I said we chase the win sequences.  With an Insurance Factor of 20 we would never achieve
our target otherwise.
The Win Sequence is 1- 3- 6- 10- 16- 26- 42. ( This is related to average expected Dividend and is adjusted accordingly.)
It rarely goes beyond a series of 4 and the factor of 10, since a horse paying say $3 with a D Factor
of 2 and a W factor of 10 would have a  factor of 20 and since the Insurance factor is 20 this 20/20 = 1,
in other words the bet would be the same size as the target and the target must be exceeded if the bet wins, and so the target would become 'negative'.  In the case of the target dropping below 100 re reset to 100.

For those contemplating setting up a Spreadsheet like this a few "tips".
All of these "Factors" such as Insurance size,  Win sequence factors, Dividend Factors etc, should be set up in "named" cells and the formulas related back to the named cells.  Then you can "experiment"
by changing the insurance factor to say 10 or changing the  third win sequence factor to 12 instead of 10,
you can then change certain factors in isolation or in combinations and see the effect on the entire series of bets at the push of a button. I have used that type of methodology in analysing systems and staking plans and the two in combination for the last 20 plus years.

So using this Formula Next Bet = T/I*D*W.
with I set at 20.
B set at $100.
BT set at $10.
D at 1 - 2 - 4 for   <$2  #  $2.10 - $4.00  # $4+
and W at 1 3 6 10 16 26 42....

The 1824 bets on Saturday Metropolitan favorites resulted in a net Profit of $4,352 on a total turnover
of  $53,383 for a POT of 8.15%  the minimum bet size was $1 the maximum $500 and the average bet
size was $29. After we had had our third collect on the second day we were always in front and using the
TAB's money, except for an 8 bet sequence on the tenth day where we reached our worst situation of minus $499.
 
 
 


 

    This graph is a Punters Delight or Bookies Nightmare - and remember this was not any fancy system - just
     backing every favorite - with a sound selection system this staking plan is damn near miraculous.
     Notice that after week 2 we were ever "in the red" only once - for a period of 8 bets and a maximum of $499.
     Notice although there are 5 major down turns each is of short duration and the bottom of each trough is
     higher than the bottom of the previous trough.
     Both tops and bottoms on a consistent rising trend.  Particularly the Progressive Profits whose trend line is
     almost unbelievably constant.

    I guess the only thing that remains is an illustration of the main points - taken from weeks 48- 49:

        Win Div    Plc Div    "D"  "W"   Bet   Return     Prof/Loss Bet     Bet Bank   Target     Prog Bank   Loss Seq.

        $2.70       $1.30         2      1     $11      $30          $19                  ($58)        126         $4,001           33
                      We bet $11 on a $2.70 winner reducing our loss this sequence to $58 this was the 33rd
                      bet in the series.

           0          $0.00         1      3      $19        $0         ($19)                  ($77)       147         $3,982            34
                         Our "Target" is now 34 bets times $2 plus the $58 we are losing = $126
                          Divided by "Insurance" 20 = 6.3 times D of 1 as div is $5+ and W = 3 since
                            previous bet was winner = $19.
 

       $3.20     $1.60           2      1     $15         $48          $33                  ($44)        116       $4,015            35
                        Our target grows by $2 for the bet plus loss of $19 = +21 =    147
                         divide by 20 - D is 2 as Div range $3 - $5 W is 1 since prev bet lost.
                                                =147/20*2*1     =$15
 

      $4.70   $2.40             2       3     $35        $164        $129                 $85 =$0        2     $4,144               1

                                     Target reduces by win and increase by $2 =  116
                                     D = 2 and W = 3 as following win bet = 116/20*2*3 =$35

          0     $1.10            2       6      $1             0            -$1                    ($1)               5   $4,143                2
                                     Target reduces by win and is negative.
                                     Start new series with target $2   Losing sequence now back to 1
                                     Bet Bank to '0'
                                     bet = 2/20*2*6 (W=6 as follows 2 win sequence) =$1.

       $3.50  $2.30          2        1      $1             $3            $2                    $0                  5   $4,145              1
                                    Target m= 2
                                    bet = 2/20*2*1  =1

            0    $0.00         4         3      $2            $0             ($2)                 ($2)               4  $4,143               2
                                    target is 2 + 2 = 4
                                    bet is 4/20*4*3=2

          0     $0.00        2          1        $1          $0               ($1)                 ( $3)              8  $4,142             3
                                    target is 4+2+2 = 8  ( which is the same as saying  $2 per bet times L (3) plus bet bank loss $2)
                                    bet is 8/20*2*1   = 1   D = 2 as div $3.30  W =1 until winner struck

          0      $0.00      4           1         $2         $0                ($2)                 ($5)             11  $4,140             4
                                   target is 8 +2 + 1 =11
                                   bet is 11/20*4*1  = $2
                                    of course this is just explanation the spreadsheet works this out for you once the formula is set.

        0       $0.00      4            1          $3         $0                ($3)                  ($8)            15  $4,137           5

       $3.10 $1.50      2            1          $2        $6                 $4                     ($4)            20  $4,141            6

      $1.80  $1.30      4            3          $11      $20               $9                     $5              18  $4,150          7
                                            Target is L=7 x 2 = 14 + Prog Loss $ 4 = 18
                                            bet = 18/20 * 4 (D=4 div less than $2) * 3 following winner.
                                            =11.
                                            Win is less than target despite betting bank in + so sequence continues.

     $3.30  $2.20      2            6           $7      $23            $16                      $0              11       $4,166          1
                                            Target = L (8) x 2 less bettg bank $5 = 11
                                            bet = 11/20*2*6 (W=6 as follows two wins) =7
                                            $16 win exceeds target start sequences again.

    $2.80  $1.70       2          10             $4       $11              $7                      $0              4         $4,173        1
 

        0    $1.60       4            2              $2        $0            ($2)                     ($2)              4          $4,171       2

       0     $1.10       4            1             $1          $0            ($1)                    ($3)              8           $4,170      3

       0   $0.00        2             1           $1           $0             ($1)                   ($4)               11         $4,169      4

       0     $1.70      2             1             $1         $0             ($1)                   ($5)                14       $4,168      5

       0    $0.00       2               1           $1         $0              ($1)                  ($6)                17       $4,167       6

      0    $1.40         4            1             $3         $0            ($3)                   (9)                 20       $4,164       7

                                        Small bets as we wait for the tide to turn

    0     $1.40         2            1               $2         $0               ($2)                 ($11)             25        $4,162     8

   0      $2.00         2            1              $3           $0              ($3)                 ($14)             29        $4,159     9

 $2.80 $1.60         4             1             $7           $20               $13                 ($1)              34        $4,172     10

 $2.50 $1.60         4            3              $8           $20              $12                  $11               23         $4,184      11

  0       $0.00        2              6              $8            $0               ($8)                  $3             13        $4,176      12
                                            although we failed to get our three in a row we recovered all losses on the previous 8 losers.

  0     $2.30          4             1               $4           $0               ($4)                ( $1)              23       $4,172     13

0      $1.30           2             1               $3           $0                ($3)               ( $4)               29        $4,168    14

0       $1.60           4            1              $7        $0            ($7)                     ($11)              34          $4,157       15

0     $1.10            4            1             $9          $0            ($9)                    ($20)              43           $4,137      16

0   $0.00              2             1           $5           $0             ($5)                   ($25)               54         $4,132        17

0     $1.70            2             1             $6         $0             ($6)                   ($31)                61            $4,126      18

0    $0.00           2               1           $7         $0              ($7)                  ($38)                  69           $4,119        19

0    $1.40         4            1             $15         $0             ($15)                   ($53)                 78            $4,104      20

                                        bets are gradually building up to take advantage of the wins when they come.

0    $1.40         2            1               $10         $0               ($10)                 ($63)             95            $4,094     21

   0      $2.00         2            1         $11           $0              ($11)                 ($74)             107            $4,083     22

0       $1.60         4             1             $24           $0            ($24)                ($98)             120            $4,059     23

0       $1.70         4            1              $29           $0             ($29)                  ($127)         146            $4,030      24

  0       $0.00        2        1             $18            $0               ($18)                 ( $145)          177             $4,012     25

  0     $2.30          4             1        $38           $0               ($38)                ( $183)             197           $3,974     26

0      $1.30           2             1       $24           $0                ($24)               ( $207)              237        $3,950    27

$3.20 $1.70         2            1        $26            $83                 $57               ($150)              263        $4,007    28

$2.40 $1.70         4            3        $125           $300            $175                $25             208            $4,182       29
                                                        although bettg bank positive target not reached continue series
$2.20  $1.40       4            6         $41             $90                $49                 0                   35         $4,231         1

                                                    Target attained return bettg bank and losing sequence to 0 and 1

                13 winners from 43 = around average 30%  TAB return on 43 bets $38.20 LOT  10%

                but our outlay $553  average bet size $13   return   $802   POT  $249   45%


A further study of all Saturday Favorites Melbourne - Sydney - Brisbane - Adelaide for the three months 1/1/99 to 27/3/99 confirmed the above with an 8.4% POT.

Also note my new EMail address I have changed ISPS


 
 
 
 
 
 
 
 

                        SATAN'S  STEPLADDER

But in the "tradition" of trying to teach you to develop your own approaches and ideas, let me explain how "Satan's Stepladder" developed as a Staking Plan.

For some time I have advocated "Parlay" betting for a place on selection methods that can consistently
produce 60% Place Strike Rates. This was based on the ability of such methods to achieve a 20% strike rate for a four leg parlay, that is once in five. A four leg Parlay that can achieve an average place dividend of $1.75 will return around $55 for the outlay of  $11 and will break even with the 20% strike rate, so if either the 60% strike rate or the $1.75 average dividend can be improved upon a profit is made.  In fact if the average dividend is $2 the return jumps to $72  with a  P.O.T. of 25%+ or if the strike rate is one in 4 with average dividend of $1.75 again a 25% P.O.T.  is achieved.
This is all very well - BUT - for the "stay at home" punter betting at the local T.A.B. one can only place Parlays on horses at the same track.  If you have 2 selections in Melbourne, one in Sydney and one in Brisbane, you have to recalculate and place 11 separate bets - 6 doubles - 4 trebles - one quadrella.  This is very difficult, mathematics wise for many punters, and often almost impossible time wise.
Therefore I developed the "Flexible Parlay" I found that if you started with a bank of 11 'units' and simply bet 60% of the progressive bank on each following bet over the 4 legs you will end up with almost the same result as the T.A.B. Parlay - and there is no recalculation of the 11 bets just 4 calculations of 60% of the then progressive bank, and the horses can be at different tracks and even different days or racing types - i.e. you can mix a greyhound bet and a thoroughbred bet today with a trots bet the next day and another thoroughbred bet a couple of days later.
To make the mathematics of an illustration easy take a four leg all placed mixed bet where all are placed and the dividends are all $2.  The T.A.B. Parlay return would be $72 - a profit of $61 -  if you could place such a bet.
doing it "manually" you would have to realise that bet
"A" is in 3 doubles 3 trebles and one quad and place $7 units on it for a return of $14.
Then bet "B"  "inherits" $2 from bet "A" for one double, two trebles and the quad for $8 plus
"B" commences two doubles and one treble for another $3 a total of $11 to bet on "B" for a return of $22.
Then "C" inherits $2 in doubles from both A and B, plus $4 in one treble plus $2 in two other trebles, plus $4 in the quad  as well  as commencing one double so a total of $17 is invested on "C" for a return of $34.
The "D" inherits $2 in doubles from each of A B and C, $4 in each of three trebles, and $8 from the quad a total of $26 for a return of $52.
So we win $7 from "A" plus $11 from "B" plus $17 from "C" plus $26 from "D" total winnings of $61. as with the T.A.B.  Parlay but requires 11 complex calculations.
Now take the progressive 60% method - we have a starting "bank" of $11 our first bet is 60% =$6.60
so we round it to $6.50 we win $6.50 and now have a "bank" of $17.50 so our next bet is 60% of $17.50
= $10.50, we win $10.50 so now our bank is $28 and our next bet is 60% of $28 = $16.80 which we round to $17 and we win $17 so our bank is now $45 and our final bet 60% of $45 = $27 and we win $27 for a final bank of $72 - exactly the same $61  profit, but a lot easier in the "execution".
I have used this method for years, and another "flexibility" with it is that you do not have to have a fixed number of legs you can do this for three legs, four legs, or forty four legs it does not matter.
However, if you continue this on at 60% you will go bust when you strike five or six consecutive losers, as your bank after a loss is down to 40% after two losses 16% three losses 6.4%  four losses 2.5%, and five losses 1%, even a "progressive" bank of $100 is then reduced to $1, which all goes onto the next bet. However when you get a couple of collects and then miss one and then collect a few and then miss a couple and so on you can fluctuate up and down for a very long time, but eventually the 7 losers hit and you go broke. So how do you use this idea "profitably"? - you set yourself a "Profit Goal" say 10 times the original stake.  For example start with 100 units, first bet is 60 units etc, you go up and down, if the bank recedes to less than 10 units add a fresh 100 and continue, when the bank reaches the profit target in this case 10 times the 100 or 1000 units - take out the profit of 900 units and start again with 100 units.
 



 
 
 
 

The A B C Staking Plan for the Simple Systems:

If a horse is selected in more than one "Simple System" the "weighting" is 1 for each time selected, except for the addition of the following 'extra' weightings.

When a horse is selected by the 'a' simple system - that is the simple ratings method
then if the 'margin' over the next horse is 100-200 the weighting is 1,
but if the margin exceeds 200 then the weighting becomes 2
and if the margin exceeds 300 the weighting becomes 3 and so on.

When a horse is selected by the 'b' simple system that is the 1st/2nd favorite method
the weighting becomes 2 if the selection is the Favorite and the second Favorite IS NOT a nines horse.
you also add 1 weighting for each extra consecutive win other than the last start.
i.e if won last two starts add 1 if  won last three starts add 2 if won last four starts add 3.

A selection becomes an "A" Category selection where it is selected at least 3 times and has a
'weighting' of at least 6.

A selection becomes a "B" Category selection where it is selected at least twice and has a
'weighting ' of at least 4.

Any other selection is a Category "C" selection.

You can just bet on all of these evenly if you like, but a more methodical and usually more profitable approach is to "bet to prices" that is you only bet if the price available represents 'value'.
I find that over time I can achieve around 50% win 80% Place strike rate with "A" selections, so to bet to prices or value means only backing "A" selections at 1/1 or better, that is do not bet "Odds on".
With "B" category selections my long term average win strike is around 37%,(68% Place) so if we could get 2/1 about all of these we would show a profit, so to bet to prices with "B" selections only bet if 2/1 or better is available.
With "C" category selections my long term average win strike is around 27%,(60% Place) so if we could get 3/1 about all of these we would show a profit, so to bet to prices with "C" selections only bet if 3/1 or better is available.

The recommended "Simple" staking Plan, therefore, is:

Category "A" 6 units to win providing at least 1/1 available.

Category "B" 4 units to win Providing at least 2/1 is available.
bet 2 units eachway where 4/1 or better available.

Category "C" 2 units to win Providing at least 3/1 is available.
bet 1 unit eachway where 4/1 or better available. ###

Where the Place Dividend of any selection is 50% or more of the Win Dividend (3 mins before Post)
double the wager and bet for the place only. - (50% for 1/1 - 10/1 Refer  to article below
on value place betting for complete strategy).     ###

###Where there is a clash bet eachway.

If more than 1 Category A selection link all Category A selections in Win Doubles.
If only 1 Category A link in win doubles with all Category B selections.
Where 6 or more selections in Category A + B take Place  Parlay 60% Simulator.
If less than 6 then add Category C if more than 8 drop Category A.


The 50/50 Staking Plan.
 

Many punters seek a Staking Plan for high strike rate Place betting, where the strike rate is
high the Dividend low and the flat stake profit minimal, for this I suggest the 50/50 Plan.

The first bet on the 50/50 Plan is 1% of bank, for every loser increase the % of Bank
by 50% and round fractions up so if you keep losing the sequence would go:
1% 2% 3% 5% 8% 12% 18% 27% 40% 60% 90%
on each 'collect' multiply the last bet by 100 minus half the 'profit'
most find this difficult to understand until they study the following table:

If the Dividend is $1.50 and the last bet was 12% of bank then the next bet will be
9% of bank - as follows:
$1.50 represents 50% profit - so 100 - 1/2 profit = 100 - 25 = 75%
so the next bet is 75% of the previous bet that is 75% of 12% = 9%.

Another example if the collect is $2.00 then the profit is 100% and the
next bet is 100 - 100/2 = 50% of previous bet which would be
50% of 12% so the next bet would be 6% of bank.

Complete Table:
 


The Pen - Step.
 
 

The Win Bet Staking Plan I am using for the Simple Systems is the PenStep invented for the
study I did on Don Beggs excellent Book Walkaway a Winner.

It is a little complicated in that it works on 'Divisors' - The purpose of Divisors is to
take advantage of the 'natural' pattern of winners - winners come in an 'unpredictable'
sequence or pattern so the natural pattern is unpredictable - but yet predictable in that
once a strike % is well established - say over 2,000+ bet history, it almost inevitably
re-asserts itself - IN THE LONG RUN.
Now the success of any Staking Plan lies in achieving a pattern that results in the
AVERAGE WINNING BET being at least 10% greater than the
AVERAGE BET. This is usually achieved by PROGRESSION after loss and
REGRESSION after a win, so that if you made a graph of the bets all of the
wins would occur at 'peaks' in the graph.

In this Plan I use a Divisor of 5 and a maximum bet of 50 units. All 'Divisor' plans are
a form of target betting and therefore you base your bets on a 'Target' - this Plan works
on a minimum 'Target' of 50 units. In order that the maximum bet will be 50 units with a
Divisor of 5 you must change the Divisor when the "Target" reaches 250 - as 250/5 is
50 or your maximum bet, so for each 250 units in the 'Target" the Divisor is increased by
5. So the Divisor "Chart" would be:
 

You commence with a target of 50 and a divisor of 5 so the first bet is 10 units.
you add losses to the target and deduct wins from the target - after a win if the
target is less than 50 change target to 50.
if you strike a run of 'outs' the bets go like this: Send your questions and comments to:......turfacts@gl.hardnet.com.au  - only.
    If you would like access to a lot more free advice copy your comments to "ausrace".by clicking this line..

Copyright © 1996
Last modified:26/4/99