ty Trev, misread his post . If all the bets are going to be matched IR only then yes, depends how long he is going to leave them up for.

I can see how it could be value in the first 10 seconds of a race - unless a 5f race i suppose - as a backer who really wanted to get on at a price but didn't quite get matched before the off may dive in for fear of 'missing a winner'.

BTO was more concerned with laying pre off, and most of his matched bets would be pre off, he did leave them up briefly IR but then would cancel quickly if not matched (how quick depending on the race type i expect).

I just picked what I thought was the best written, didn't really look into the numbers as I know what the broad idea is.

whether maria lied or not her staking system was solid and her selections had AN 82%+ strike rate...so she would have made money either way.

No, wether she made money depended entirely on what prices she was matched at .She posted the prices she claimed to have been matched at. After racing had finished.

She stated on here thread that 2/3 of her lays were matched by 12pm, and posted these Lays .But she never said at that time what price. She waited til the race was over and then claimed to have been matched at the lower end.

If she was genuine, surely easy to say 'i've layed Dobbin at 7.8', then people can check to confirm money has been matched at those odds at that time.

If she was dishonest, but wanted her thread to show a profit, then she would wait to see if Dobbin wins ,then see what was matched , and if the lowest price was ,say, 7.4, claim to have layed @7.8, even if most of the matched money was between 8.6 and 10.0.

If she did it the latter way then people following her on the thread would have been losing while she was winning, even with the same selections.

Which is what happened.

kef4o Joined: 05 Sep 07
Replies: 274 05 Jan 11 13:48 
If you are betting flat you are essentially giving very imilar importance to every bet you make. With Kelly, the way I see it regardless of how good your judgement is, you are done if a big value election fails to win. On a different note, essentially, only your big bets would really matter for the overall bank balance, so you don't even need to bet small.

In other words, using Kelly, you end up betting flat but too big of a percentage on all your big value plays and you mix it up with a few smaller bets.
==================================================================


If you check the results shown on my last post on the KELLY HORSE BACKING thread , you will note that, of the 18 selections using the KELLY CALCULATOR on thread, there were ONLY 3 bets where STAKING was over 3% of bank ( reflecting the "edge" I believe I hold on those races) AND ALL WON.

However, there were also 3 return bets where the staking risked was under 2% of bank ( reflecting the competitive nature of race.....even though "edge" held).....and they ALL WON.

So are you saying I should NOT have bet on those 3 bets ( under 2% risked)......even though the contribution to profits was high due to high odds re the 20-1 SP horse being placed and there being 2 winners out of the 3....?

There will be days where the bulk of my returns will come from the under 2% risk bets. As many of such bets are at high odds eg BLAZING BOLT at average odds bet of 6.91 in PLACE market, I don't believe what you suggest would be beneficial to profits.( ie ignore doing the "smaller" stakes bets)

kef4o -Have you ever tried using KELLY CALCULATOR ?

YES - you have to be reasonably accurate on weighing up the probability of a horse winning/placing to use KELLY, but IF YOU HAVE "EDGES" THIS WULD ASSIST IN SUCH WEIGHING UP........

Of course, if you DO NOT HAVE AN "EDGE", you will lose anyway.......

REM   


kenilworth

REM
You need a system that can identify value coupled with a sensible staking plan to optimise winnings and protect against losing runs.

The underlined doesn't exist IMO. If it does, perhaps you can divulge ..?

Tobermory : BTO was more concerned with laying pre off, and most of his matched bets would be pre off, he did leave them up briefly IR but then would cancel quickly if not matched (how quick depending on the race type i expect)

The method he was using in 2007 was purely in-running lays and cancelling at a pre determined point. He lives in Oz so couldn't actually bet in-running, only cancel.

If she did it the latter way then people following her on the thread would have been losing while she was winning, even with the same selections.

Which is what happened.

I haven't read the Maria story, but yes I can believe that. It's a similar story to the Daily Donkeys, if you google their website their results are fantastic but cannot be replicated here on BF, simply because they calculate their winings to industry SP, so the whole thing is a con in that respect.

Re. the Maria staking plan

I've always been puzzled as to why the highest stakes (1% of bank) are used for an odds range where statistically most winners come from and similarly the lowest stakes (0.4% of the bank) are used for an odds range where fewer winners are likely to be encountered.

I realise that her selection method may have been able to identify bad value bets at odds below 3.5 hence the higher stake but my first impression was that it was more of a backer's staking plan than a layer's.

i hope its for "sectionals" my friend !

ykicketc
I'm interested in why you differentiate between laying and betting ?
After all , laying is just betting on more than one horse in the same race instead of just the one horse.
Surely value factors apply equally no matter if you're laying or betting.
What am I missing here that's so important ?

Don't be so sure K.
After all it's only your opinion, not fact.
Overstaking or overbetting is widely regarded by most experts as the single biggest fault in gambling and in fact in investing in general.

Ah! The all important edge How do we define it? How is it found? What do we do with it when we A. find it, B. When it becomes clear how much we have of it what is the best way to exploit it?

Does everyone agree that you have to have an edge to make a profit (long term)?

Is everyone in agreement that level stakes is equal liability on each selection whether it is betting to win or laying?

There are many more questions that need to be answered before risking money foolishly.

Staking, when the discussion of a staking plan comes up, most immediately think of some sort of recovery plan. There are a lot of staking plans
that do not involve chasing, just sensible optimization of capital.

Any thoughts?

I am going to try an answer some of these questions. I have 'one note' open and as I don't have the gift of the gab will write and re-write
my answers till they make sense and are accurate as far as I can tell. (Could take a while)

kenilworth
I disagree with your opinion ~ which IMO doesn't, in isolation, exist.

Your quantifier of 'in isolation' needs expanding on. There are such plans that 'optimises winnings and protects against losing runs'

Kelly protects against losing runs? And if you encounter a tough losing run of your high value bets?

Compound Magic - The first sentence of your post 09 Jan 11 00:11 may have opened up a can of worms.  

With the questions you've raised in mind, I'd be interested to read your observations on the statement below.

The key to optimising staking is to know the edge in favour of the punter.  More should be staked when the edge is greater.  With cards, the edge can be calculated with exact precision.  The card-counter knows how many of each valuable card have already shown and the number of cards left.  Therefore the odds of turning up a valuable card on the next deal can be calculated exactly and hence the edge calculated exactly.

Horse-racing simply is not like this. 

The probability of a certain horse winning can never be known with perfect accuracy.  This would entail knowing precisely the details of every variable with a bearing on its chance of winning - its blood white cell count, how it was feeling before the race, the “vibes” it was picking up from its jockey, whether it had digested its last meal adequately and so on , almost ad infinitum.  Calculating prior chances for every horse in the race is a very, very uncertain process.  It depends first on the quality of the information that is fed to the computer which calculates them, next on the ingenuity of the measures used to estimate various aspects of the horse’s fitness and ability and the connections’ will to win and finally on the sophistication of the calculating algorithm.  Thus a calculation can be made based simply on ability factors, but it will never reflect the actual chance of that horse winning accurately.  More factors can be added to the equation, but the estimation will still be inaccurate.  Even if figures are re-jigged at the last minute to account for odds and market movements, the figures will still fall short of perfection.

If you cannot know the true odds for your bet, then you cannot know the true edge.

i agree with sea the stars...the only difference is i learnt to pick losers as a layer..:)))
its all about risk management over a long period of time..the more people realise this the better..systems r just that systems and most people dont stick to them long enough to get the answers..one years results is not enough to gauge if u have an edge.i feel 3 years of running a so called system will highlight all u need to know..do u have the patience to run ur system for 3 years without changing tact..most ppl would say no,they will pull out on the first losing run...if u can model ur betting on that of a bookie then its a good start..an egde of just 1% makes a huge difference over time believe me...ask any bookie..if u made 1% on ur betting bank evryday it would add up to a nice bank at the end of the year...i have talked about the maria staking plan here(not the person who invented it) as its so simple and feel it has a lot of wat i look for in a staking plan over the long term..it uses progressive staking to small percentages of a bank..after a losing run it recovers and if u refine it even more its pretty solid over the long term..i have even run maria over a year and let my bot random pick selections to test its viability and it produced a nice profit.its on year 2 now and its running at an 85% strike rate already in 2011 and this is random picking selections...i have of course refined maria to be more solid in the liabilities department and have 5 price ranges to certain percentages but it works..i have got stick in here for the way i do things but i make a profit and thats all that counts...COMPOUND MAGIC i agree with wat u say on kelly wit regards sports betting and blackjack....the variables of  a horserace and a card game r far removed from each other..its laughable to even compare them...and yes as a layer i am a backer of every other horse in a race bar the one i layed..i like those odds overtime..
keep it simple keep it pro :-)))



between 7.

all this rubbish about kelly systems and other systems is in fact all rubbish

Another one wants to take on the laws of mathematics.

There are good reasons, already presented, why Kelly isn't for everyone and should be treated with care by nearly everyone. To call it rubbish however is to show your own lack of understanding of an irrefutable mathematical proof.

The good judges can still maximize what they win by correctly staking.

Feel free not to use maths to aid you, but the further away from kelly you bet (whether deliberately because you can't get accurate odds assessments for various reason or by accident because you're not good at assessing odds) the less optimally you're staking.

Thankfully it's not even open for debate, the only debate is how close you should attempt to get.

Of course, I'll be happy as ever to listen to a reasoned mathematical proof of why you think it doesn't work.

Many of the posts on this thread demonstrate the relative inexperience of the posters (this is not meant to be patronising but I realise that it might be interpreted as such.)

The longterm success at gambling requires the player to understand the effect of the most important variable involved which hardly gets a mention. It is not selction. It is not staking plan. It is not % edge. It is VOLATILITY.

Let me explain. Most players would probably agree that an edge that gets you a return on stakes of 10% whether as a backer or a layer represents a desirable and probably maximum return that one could hope for, at least as far as horse racing is concerned.

many speak of having achieved such returns and declare themselves as winners at the game.They would not accept that they have been lucky. they would argue that they have a proven edge.

I suggest that anyone who is able to, puts a simple model together on a spreadsheet and using the random number facility, looks at the effects of volatility on an edge of say 10% across say 1000 bets.

What you will see is that at a small win% say 10% with say an average SP of 11.0 you will be subject to a wide variety of potential returns after 1000 bets ranging from profit to loss involving many ups and downs. Using Kelly staking will only add to the volatility.

Then look at say an 80% strike rate with an average SP of 1.375 giving a return of 10%. The difference in volatility is huge. Using kelly is a much safer proposition.Using such models should demonstrate why a strategy with a low win% irate is not sensible for anyone who bets for a ling, certainly on the horses.

I do not intend to enter into a debate again about Kelly but imo its use in practice as far as the horses is concerned is too risky. It is not possible to be sufficiently precise and consistent in pricing, especially outside the front two and especially when in horse racing too much apparent potential value is an indicator that all is not what it seems.Caveat emptor.

Fred

Assuming the conditions you state, and at 1000 bets p.a (20 per week possible) and returning 100% on stakes without commission (impossible), he would win an etsimated £200k so I guess he would be not so much a liar or genius but a fantasist.