The home of intelligent horse racing discussion

# A Method of Staking

Home Forums Tipping and Research Systems A Method of Staking

Viewing 11 posts - 1 through 11 (of 11 total)
• Author
Posts
• #11820
dave jay
Member
• Total Posts 3386

Most punters have a problem with losing runs .. including me.

I thought it would be nice show some of the people on here a method of coping with losing runs and how to stake using a risk management model instead of a simple staking plan.

Beside that I haven’t made a proper race related post on here for about a year .. so it’s about time ..

The method basically works like this ..
1 – You create a price for each horse.

2 – You stake each horse to return 100 pts by backing the horse to a stake which is equal to your percentage chance, so a horse that you think has a 10% chance gets 10pts placed on it and so forth.

3 – You ignore the price that you are being offered about that horse and blindly stick to your prices.

By betting like this you will return your edge, be that negative or positive, so the ratings should be based in logical well thought out fact and not a hot-potch of back-fitted nonsense. Concentrate on the horse itself and leave the other clever stuff for everyone else.

I know staking in this method goes against a punters common sense with regards to value but quite a few mathematical problems appear to do that. When I stake like this I am not trying to win, I am managing my risk. That’s why it doesn’t make sense!! I’m going to opt for the long game. So long as my ratings are good then I don’t have a problem.

When you think about risk management (mitigation) you have to think big, from a numbers point of view. I need to invest 100 points on each race, say I decide to use £2 points, that’s £200 per race.

You then have to decide how many banks of £200 you are going to need to never blow your bank, this is a bit more complicated. For calculation sake we will say that we will back 1 point on each race, so 1pt = £200.

For anyone that’s uploaded that into excel you will see that a couple of races returned -70pts. That’s a losing race obviously (-£140) -0.7 of a point. But with the correct amount of points (£200’s) this becomes just a part of the cash flow .. the bank needs to be deep enough to absorb the random runs of results.

There is no right way, or a calculation you can pull out of the air to determine how deep the troughs will be. Peter May suggested that chi-test could be used, others have gone down the binomial route and others have tried Bayesian inferencing. I prefer rating no less than 700 races and then using Bayesian calculations for the top half of the rated horse versus the bottom half, the Yanks call these Impact Values (IV’s), I think. Although it’s not perfect, it really does sort the wheat out from the chaff.

This is a file I pulled off my own ratings for today, you should be able to drag and drop this into an excel spreadsheet, if your interested.

A few of the selections have a ‘0’ value, this is rounding error, but I don’t have time to fix it. This value is ‘1’.

Name Race BSP Return
Joinedupwriting 15:15 Redcar 38% 2.91 -38
Just Observing 15:15 Redcar 21% 9.2 -21
Spirit of France (IRE) 15:15 Redcar 19% 3.85 -19
Trouble Mountain (USA) 15:15 Redcar 13% 35 447
Eijaaz (IRE) 15:15 Redcar 4% 6.6 -4
Little Sark (IRE) 15:15 Redcar 4% 28 -4
McCormack (IRE) 15:15 Redcar 0% 14 0
Johannes (IRE) 15:55 Redcar 21% 3.6 77
Vhujon (IRE) 15:55 Redcar 20% 6.5 -20
Wyatt Earp (IRE) 15:55 Redcar 18% 9.2 -18
Jaconet (USA) 15:55 Redcar 13% 9.3 -13
Tabaret 15:55 Redcar 10% 28 -10
Fol Hollow (IRE) 15:55 Redcar 9% 8.8 -9
Benllech 15:55 Redcar 8% 7.8 -8
Flash McGahon (IRE) 15:55 Redcar 0% 19 0
Aussie Blue (IRE) 17:05 Redcar 20% 11 -20
King Pin 17:05 Redcar 15% 5.22 79
Efidium 17:05 Redcar 14% 14 -14
Kalasam 17:05 Redcar 13% 9.6 -13
Rosko 17:05 Redcar 10% 11 -10
Daaweitza 17:05 Redcar 9% 9.4 -9
Hilbre Court (USA) 17:05 Redcar 9% 17 -9
Minority Report 17:05 Redcar 5% 7.6 -5
Regal Lyric (IRE) 17:05 Redcar 4% 11 -4
Moonage Daydream (IRE) 17:05 Redcar 1% 19.5 -1
Dolly No Hair 17:05 Redcar 0% 150 0
Northern Flyer (GER) 15:30 Musselburgh 25% 1.87 47
Minturno (USA) 15:30 Musselburgh 20% 8 -20
Cash In The Attic 15:30 Musselburgh 15% 30 -15
Quick Gourmet 15:30 Musselburgh 15% 8.8 -15
Lost In Paris (IRE) 15:30 Musselburgh 12% 13 -12
Meydan Groove 15:30 Musselburgh 7% 23 -7
Our Apolonia (IRE) 15:30 Musselburgh 5% 200 -5
Reel Bluff 15:30 Musselburgh 2% 130 -2
Ask Dan (IRE) 15:30 Musselburgh 0% 40 0
Flynns Island (IRE) 15:30 Musselburgh 0% 46 0
Mountain Cat (IRE) 16:45 Musselburgh 29% 2.2 63
Shunkawakhan (IRE) 16:45 Musselburgh 27% 16 -27
Kabis Amigos 16:45 Musselburgh 20% 12.5 -20
Ninth House (USA) 16:45 Musselburgh 12% 17.5 -12
Grand Diamond (IRE) 16:45 Musselburgh 8% 5.7 -8
Bold Indian (IRE) 16:45 Musselburgh 4% 6.7 -4
Fern House (IRE) 16:45 Musselburgh 0% 150 0
Red Skipper (IRE) 16:45 Musselburgh 0%
Argentine (IRE) 17:20 Musselburgh 36% 1.97 -36
Spirit of Coniston 17:20 Musselburgh 19% 8 -19
Ridley Didley (IRE) 17:20 Musselburgh 15% 8.6 -15
Rocketball (IRE) 17:20 Musselburgh 13% 9.29 119
Distant Vision (IRE) 17:20 Musselburgh 11% 25 -11
Mr Rooney (IRE) 17:20 Musselburgh 6% 55 -6
Blazing Heights 17:20 Musselburgh 0% 11.5 0
Saloon (USA) 19:05 Newmarket 34% 3.75 -34
Cwm Rhondda (USA) 19:05 Newmarket 31% 3.2 -31
Graylyn Ruby (FR) 19:05 Newmarket 22% 11 -22
Apache Fort 19:05 Newmarket 6% 17.58 -6
La Rosa Nostra 19:05 Newmarket 3% 8.6 -3
Silent Applause 19:05 Newmarket 3% 7.8 24
City Stable (IRE) 19:05 Newmarket 0% 39 0
Baby Queen (IRE) 21:10 Newmarket 29% 5 -29
Amosite 21:10 Newmarket 21% 9 -21
Deckchair 21:10 Newmarket 13% 11 -13
Halaak (USA) 21:10 Newmarket 10% 6.5 -10
Hi Shinko 21:10 Newmarket 10% 5.8 -10
Red Rosanna 21:10 Newmarket 10% 8.6 90
Keep Dancing (IRE) 21:10 Newmarket 6% 8.5 -6
Brynfa Boy 21:10 Newmarket 0% 6.5 0
Oceana Blue 19:55 Goodwood 22% 9.4 -22
Perfect Friend 19:55 Goodwood 22% 15 -22
Balaagha (USA) 19:55 Goodwood 20% 3.1 -20
Labisa (IRE) 19:55 Goodwood 20% 6.6 -20
Al Sabaheya 19:55 Goodwood 8% 16 -8
Fleeting Star (USA) 19:55 Goodwood 4% 10.6 -4
Victoria Sponge (IRE) 19:55 Goodwood 4% 7.5 29
The Jostler 19:55 Goodwood 2% 29 -2
Nice To Know (FR) 19:55 Goodwood 0% 36 0
Greenwich Village 20:30 Goodwood 28% 12.5 -28
Yes Mr President (IRE) 20:30 Goodwood 19% 12.5 231
Presbyterian Nun (IRE) 20:30 Goodwood 17% 7.8 -17
Sevenna (FR) 20:30 Goodwood 11% 6.7 -11
Relative Strength (IRE) 20:30 Goodwood 7% 7 -7
Storyland (USA) 20:30 Goodwood 7% 2.99 -7
Aaim To Prosper (IRE) 20:30 Goodwood 6% 9.8 -6
Goodwood Starlight (IRE) 20:30 Goodwood 6% 19 -6
Bureaucrat 20:30 Goodwood 0% nr 0
Chookie Hamilton 19:45 Ayr 41% 4.7 194
Summer Soul (IRE) 19:45 Ayr 28% 7.7 -28
Ifatfirst (IRE) 19:45 Ayr 14% 2.8 -14
Orkney (IRE) 19:45 Ayr 14% 16 -14
Cote dArgent 19:45 Ayr 3% 6.1 -3
Kyber 19:45 Ayr 0% 12.5 0

Returns +493 over 11 races (1100 pts turned over)
Deviation -73 to +360
.. these returns are worked to betfair SP and not bookies prices, no-one should be betting with the bookies unless they are thieving.

.. any thoughts?

#235415
Artemis
Participant
• Total Posts 1736

Hi Dave

Looks a decent return from a very small sample. If you could keep that going, it would be goodbye to wage slavery.

The first step is the crucial one. If you can beat the market prices consistently you can win using just about any method. Unfortunately, arriving at an algorithm to determine your prices has to be subjective and success at creating a successful model has perplexed the best mathematical and racing brains, not least yours and mine, for many a long year. I’ve just about given up on this quest, although a few embers are still flickering as the fire slowly dies.

If you could, by some miracle, produce a successful model, the best way to proceed is the way you suggest either as a backer or layer. In the latter case, you would need to set up some kind of Betfair interactive account where you layed bets only at your designated odds adjusted to account for your profit margin, perhaps 5 to 10%.

Nice to see you are still active in the field.

#235466
Formath
Member
• Total Posts 1451

Let me see, so what you are saying is if I go through the card at Royal Ascot today in the following manner I could be in with a chance? Couldn’t be easier could it?

Royal Ascot June 20
2.30 Meglio Ancora – 10 points
3.05 Enroller – 17 points
3.45 Kingsgate Native – 13 points
4.25 Evens And Odds – 6 points
5.00 Hatton Flight – 9 points
5.30 Tasheba – 20 points

#235503
Formath
Member
• Total Posts 1451

My previous post on this thread was a bit tongue in cheek, just trying to bring the discussion more down to ground level where the rest of us are
However, I do support the principle involved so I inspected this afternoon’s cards in a more serious manner to see if there were any form possibles worth considering.

Asc 2.30
Beethoven form price calculated 10/1 so 10 points stakes (Betfair 11/2)
Emperor Claudius 11/1 9 points (Betfair 5/1)
Bikini Babe 11/1 9 points (Betfair 10/1)
Result 28 lost[/color:n0rfntw5]

Asc 3.45
JJ The Jet Plane 5/1 17 points (Betfair 11/4)
Kingsgate Native 11/2 16 points (Betfair 12/1)
Sacred Kingdom 5/1 17 points (Betfair 3/1)
Result 50 lost[/color:n0rfntw5]

Red 4.45
Elna Bright 9/2 19 points (Betfair 9/4)
Darbers Ridge 9/2 19 points (Betfair 8/1) won 7/1 return 152[/color:n0rfntw5]

Sam’s Secret 9/2 19 points (Betfair 9/4)
Result 152 – 135 staked leaves 17 profit = 12%[/color:n0rfntw5]

The bets are made at the form price so backing at SP if the short odds selections come the outlay should be about covered, hopefully one of the longer-priced one’s will win. You can’t expect to make more than about 7% overall.

#235521
dave jay
Member
• Total Posts 3386

Artemis .. I’m a bit sorry to hear that you have given up the ghost. I’ve never taken racing anything more serious than a hobby, albeit a marginally self financing one. Hopefully, this might get you to get your thinking cap out of the cupboard ..

This method of staking works on the principal that you have produced a rating that has an edge, or a difference between the top half and bottom half of the rating that the market hasn’t taken into account. This method also allows you to stake on every horse and protect your capital.

Jackform .. if you look through the results you will notice that the winners (that made the difference) were considerably longer on Bf to SP.
(Trouble Mountain (USA) 15:15 Redcar 13% 35 447 .. was 20/1 @ SP returns 273)
(Yes Mr President (IRE) 20:30 Goodwood 19% 12.5 231 .. was 8/1 @ SP returns 171 )
This won’t work, in fact nothing will, when you are up against a (2 – 3% per runner) bookies overround.

Also, simply backing the horses that are value has it’s pitfalls too. There is a chance that they will all lose and blow your bank and there’s the chance that they will all win and you’ll be paying the Betfair Premium Tax, which is worse than SP .. ?!?!

I agree about making a profit of around 7%, you are going to need to have an edge of about 15% to do that, if you take the cost of betting into account. Does that not also mean that the more I can get on the more profit I can win .. someone backing every horse is going to turn over far more than someone selectively backing value selections?

.. as I said in the opening post, this is just a bit of food for thought and up for discussion amongst my friends on here.

#235594
Artemis
Participant
• Total Posts 1736

Dave,

Your comments regarding Betfair SP are very relevant. I’ve found that using Betfair SP is the difference between winning and losing(using SP).

I haven’t quite given up, more busy with other projects. I just thought I couldn’t take my methods much further without using more of my precious time. I’m now a senior citizen and there are a lot of things I want to do while my brain is still functioning normally- if it still is!

I believe the basic method I have used which is ratings plus about 10 additional positive factors is logically sound, but it takes up a lot of time and just about breaks even at SP and makes small profits with Betfair SP. I don’t have the resources(or the time these days) to make it worthwhile even though it is(as you mention) an extremely absorbing hobby – well worth doing for anyone at a loose end or bored.

I have no doubt that I will be back on the forum with more of the RP ratings + method when time allows.

#235777
Formath
Member
• Total Posts 1451

My record of assessing AW races by form has been very poor, so I thought why not try something different and rely on coupled odds to pull me through – it is Monday after all

In the late 70’s a letter published in the old Sporting Chron forum promised to provide every punter with the path to a fortune from following successful trainers. It was OK, but far to many bets and boring, boring. However, some years later the Raceform Handicap Book stated that in their opinion it was one of the two best systems ever published by the Sporting Chron, and that is some accolade – so today my selections are a modified update of that system, fill your boots

Lingfield 3.15
Peak District priced to 100% 10/3 stake 24 pts won 9/4 returned 78 pts[/color:1r1pkz93]

Sir Edwin Landseer 18/1 6 pts 2nd 11/1[/color:1r1pkz93]

Result returned 78 – 30 stake profit 48 pts[/color:1r1pkz93]

If anyone fancies a look at the original forum letter please PM with an email address and I will oblige, we might as well all be rich
(Before anyone inquires the other top system was to follow the selections of Beat The Book, Dick Shepherd, who was a starting price reporter).

#236040
Sean Rua
Member
• Total Posts 511

An extremely interesting thread, imo!
Thanks to all concerned.

My personal view is that working to SP ( betfair or otherwise) is a difficulty, but, we have to have some yardstick when testing methods and systems.

#236127
Formath
Member
• Total Posts 1451

A 17 runner hcap, can our top trainers work the oracle today?

Carlisle 4.10 priced to 100%
Yes Mr President 7/1 stake 13 pts to return 100
Joe Jo Star 5/1 17 pts

#241782
Gerald
Member
• Total Posts 4293

I’ve only just read this thread, Dave. Rather an off the wall idea, backing every horse in the race. I’ll see if I can intellectually/mathematically justify it in a couple of days time. Having difficulty getting my head around it at the moment, but looks like I have to look at it in conjunction with the Kelly criterion methodology.

I’ve become intrigued by several people over the past few weeks having ratings for all races and selections for all races. I’ve got a couple of days off work, so I’m going to see if I can become a thief, and learn how to take RP data into spreadsheets etc. I’ll work through that link to another forum that is in the private h’capping thread. Will be a fairly steep learning curve.

#242018
binocularman
Member
• Total Posts 12

Very interesting but the problem is if i gave each of my selections my own price it would be odds on ( because i think it will win obviously ! lol )
I mean why would i back a horse if i think it has a 10 percent chance of winning ? ……thats the problem and also only a very very small number of punters can keep to level staking plans and betting patterns ……im not one of them unfortunately ……good luck , Lance

Viewing 11 posts - 1 through 11 (of 11 total)
• You must be logged in to reply to this topic.