Percentage staking plan, thoughts?

[Lardonio]

Re: Percentage staking plan, thoughts?

But with stakes it is often what a punter is comfortable with that matters. :ok

Agreed. :ok

In fact' date=' in extreme cases it is clearly wrong. If I could identify bets at odds of 2.00 for which I was confident that the expected ROI was around +10% then I would be very comfortable to stake 2% or more of my bank. If I identified bets at odds of 101.00 with ROI of +10% then I would eventually go broke if I staked 2% of my bank on each bet. [/quote'] Of course, but you would be mathematically very incompetent to bet 2% of your bank on 101-shots and therefore the example is not really relevant. It's a bit like saying that it would be wrong to stake 10% of your bank on 1001-shots. However, if a punters betting habits include everything from 1.1-shots to 101-shots then he would have to consider his staking strategy very carefully. Mathematically, for profit, assuming the same ROI on all bets, it would be wrong to vary the stakes, but of course if you bet 100 101-shots for every 1.1-shot then risk of going broke becomes an important factor. Personally, if I were regularly betting both big favourites and extreme longshots I would make a separate bank for each and vary the stakes so that the risk of going broke would be as low as possible for each bank while still getting enough expected ROI to keep me interested.

[slapdash]

Re: Percentage staking plan, thoughts?

Of course' date=' but you would be mathematically very incompetent to bet 2% of your bank on 101-shots and therefore the example is not really relevant. It's a bit like saying that it would be wrong to stake 10% of your bank on 1001-shots.[/quote'] But I thought that what you were saying was precisely that you should stake the same percentage on 101-shots as on 2.00 shots (with the same yield)? Maybe I'm just misunderstanding what you meant. I still don't understand what you mean by this, I'm afraid. If you're using a percentage staking plan, and all your bets have the same expected ROI (+10%, let's say), then even if most of your bets are around odds of 2.00 (on which using 2% of your current bank would be quite conservative) and you only occasionally have bets at odds of 101.00, then your bank will grow more slowly if you also bet 2% on those than it would if you didn't bet on them at all.

[Gingertipster]

Re: Percentage staking plan, thoughts?

Agreed. :ok However, if a punters betting habits include everything from 1.1-shots to 101-shots then he would have to consider his staking strategy very carefully. Mathematically, for profit, assuming the same ROI on all bets, it would be wrong to vary the stakes, but of course if you bet 100 101-shots for every 1.1-shot then risk of going broke becomes an important factor. Personally, if I were regularly betting both big favourites and extreme longshots I would make a separate bank for each and vary the stakes so that the risk of going broke would be as low as possible for each bank while still getting enough expected ROI to keep me interested.

So if a punter has the occasional bet at 101.00 it is o.k. to put the same stake as he would a 1.01 shot? And if you have two seperate banks for normal prices and "extreme longshots", where would you draw the line? What counts as an "extreme long shot"? 12.00, 20.00, 33.00 or 66.00? If it is say 20.00, then what mathematics suggests you should have the normal stakes for 19.00 and then a completely different stake for 20.00? Unless you split it even further in to 3, 4, 5, 25 or 55 different staking plans. Which in effect is what I have done in one staking plan; with stakes decreasing the less chance a selection has of winning. :ok

[Gingertipster]

Re: Percentage staking plan, thoughts?

you would be mathematically very incompetent to bet 2% of your bank on 101-shots and therefore the example is not really relevant. It's a bit like saying that it would be wrong to stake 10% of your bank on 1001-shots. Mathematically, for profit, assuming the same ROI on all bets, it would be wrong to vary the stakes, but of course if you bet 100 101-shots for every 1.1-shot then risk of going broke becomes an important factor.

Those two sentences mean the opposite, you can't have it both ways Lardonio.

[slapdash]

Re: Percentage staking plan, thoughts? Another example, in terms of level stakes rather than percentage staking. And although I was only using extreme differences in odds before because that made the principle more obvious, I'll stick to odds of 2.00 and 3.00. Suppose half your bets are at odds of 2.00 and half are at odds of 3.00, and suppose that your expected ROI for both types is +10%. Consider the following two strategies: (A) Bet 3 units on all bets. (B) Bet 4 units on the 2.00 shots and 2 units on the 3.00 shots. Then both strategies give you exactly the same expected winnings. In (A) you make an average of 0.3 units on each bet. In (B) you make an average of 0.4 units on the 2.00 shots and an average of 0.2 units on the 3.00 shots, but since it's 50/50 whether each bet is 2.00 or 3.00, then on average you make the average of 0.4 and 0.2 units, or 0.3 units. So which of the strategies you choose makes no difference at all to your expected winnings. But still, strategy (B) is better, because it reduces the variance of your winnings. In terms of risk of ruin, suppose you start with a bank of 30 units. Then using (A) you have about a 26.4% chance of losing it all. But using (B) you only have about a 22.4% chance of losing it all.

[Lardonio]

Re: Percentage staking plan, thoughts?

Those two sentences mean the opposite' date=' you can't have it both ways Lardonio.[/quote']

Originally Posted by Lardonio you would be mathematically very incompetent to bet 2% of your bank on 101-shots and therefore the example is not really relevant. It's a bit like saying that it would be wrong to stake 10% of your bank on 1001-shots. Mathematically, for profit, assuming the same ROI on all bets, it would be wrong to vary the stakes, but of course if you bet 100 101-shots for every 1.1-shot then risk of going broke becomes an important factor.

No, they don't. "you would be mathematically very incompetent to bet 2% of your bank on 101-shots and therefore the example is not really relevant. It's a bit like saying that it would be wrong to stake 10% of your bank on 1001-shots. " This sentence says that you must expect to go broke if you place 2% of your bank on 101-shots. RISK. "Mathematically, for profit, assuming the same ROI on all bets, it would be wrong to vary the stakes, but of course if you bet 100 101-shots for every 1.1-shot then risk of going broke becomes an important factor." This sentence says what I have been saying all along; assuming the same ROI on all bets it is wrong to vary the stakes. YIELD. When planning your staking strategy, consider expected RISK and YIELD.

[Lardonio]

Re: Percentage staking plan, thoughts?

So if a punter has the occasional bet at 101.00 it is o.k. to put the same stake as he would a 1.01 shot?

That's entirely up to the punter. It would of course depend on the stake. Mathematically when talking ROI he should put the same stake. Mathematically when talking risk it would be stupid to put 10% of the bank if that is what he usually places on his 1.01 shots. Yield and risk needs to be considered. The punter makes the decision.

And if you have two seperate banks for normal prices and "extreme longshots"' date=' where would you draw the line? What counts as an "extreme long shot"? 12.00, 20.00, 33.00 or 66.00? [/quote'] That is entirely up to the punter. We're not consulting a lawyer here or playing a board game. It's your money, do whatever you want. Personally I would not draw a line, but if I for example have a correct scores project, which I am fond of, where average odds of bets are around 12-15 and stakes around 0,5% of bank I would not hesitate to play a 66 or higher with those stakes. It would depend on my feeling for the bet, the frequency of those odds, my mood that day etc.

[Gingertipster]

Re: Percentage staking plan, thoughts?

Personally, if I were regularly betting both big favourites and extreme longshots I would make a separate bank for each and vary the stakes so that the risk of going broke would be as low as possible for each bank while still getting enough expected ROI to keep me interested.

I have "paid attention". This sentence means there must be some point where the "favourites" staking plan ends and the "extreme longshots" staking plan begins. If it is not 19.00 to 20.00 then where? It makes no mathematical sense. Your writing on this thread Lardonio has not been up to your usual standard. Full of contradictions.

[nkdz]

Re: Percentage staking plan, thoughts?

But still, strategy (B) is better, because it reduces the variance of your winnings. In terms of risk of ruin, suppose you start with a bank of 30 units. Then using (A) you have about a 26.4% chance of losing it all. But using (B) you only have about a 22.4% chance of losing it all.

Interesting thread lads although it's a bit complicated though. :unsure slap, can you explain how to calculate those chances of losing entire bank?

[Hooloovoo]

Re: Percentage staking plan, thoughts? Thank you very much slapdash for explaining this so even i can understand :) I have been thinking about staking vs risk of ruin for a week now. I have so far been using a reduced stake where odds are 3.00 - 4.00, and it seems that would be the right way to go.

Another example, in terms of level stakes rather than percentage staking. And although I was only using extreme differences in odds before because that made the principle more obvious, I'll stick to odds of 2.00 and 3.00. Suppose half your bets are at odds of 2.00 and half are at odds of 3.00, and suppose that your expected ROI for both types is +10%. Consider the following two strategies: (A) Bet 3 units on all bets. (B) Bet 4 units on the 2.00 shots and 2 units on the 3.00 shots. Then both strategies give you exactly the same expected winnings. In (A) you make an average of 0.3 units on each bet. In (B) you make an average of 0.4 units on the 2.00 shots and an average of 0.2 units on the 3.00 shots, but since it's 50/50 whether each bet is 2.00 or 3.00, then on average you make the average of 0.4 and 0.2 units, or 0.3 units. So which of the strategies you choose makes no difference at all to your expected winnings. But still, strategy (B) is better, because it reduces the variance of your winnings. In terms of risk of ruin, suppose you start with a bank of 30 units. Then using (A) you have about a 26.4% chance of losing it all. But using (B) you only have about a 22.4% chance of losing it all.

[Lardonio]

Re: Percentage staking plan, thoughts?

I have "paid attention". This sentence means there must be some point where the "favourites" staking plan ends and the "extreme longshots" staking plan begins. If it is not 19.00 to 20.00 then where? It makes no mathematical sense. Your writing on this thread Lardonio has not been up to your usual standard. Full of contradictions.

I disagree. There are no contradictions. If my writing is bad then I blame it on not having English as my native language :tongue2 Edit: Though I can see why it gets confusing, because most of the examples here (including my own) are done with a known ROI. A known ROI does not exist in sports gambling. With an unknown edge (or ROI), like we have in sports gambling, it is mathematically correct to bet the same stake regardless of odds size. With a known edge (or ROI) it is mathematically correct to vary the stake IF that maximises profits. Collini out

[Hooloovoo]

Re: Percentage staking plan, thoughts? But isn't an estimated edge more close to known than unknown, if backed up by historical results using the same selection process?

[Lardonio]

Re: Percentage staking plan, thoughts? The answer could be yes and the answer could be no :D Only the future will tell.

[Datapunter]

Re: Percentage staking plan, thoughts? The point of money management is to manage the bank over time REGARDLESS of whatever your expectation might be. If it comes out then great, plenty of profit. But if reality turns out different than expected, however you've arrived at that expectation, your staking plan allows for you to sustain a period of betting sufficient long to determine if the strategy has a merit over time. In short, if you need 1000 bets before you can say with some degree of certainty the strategy is profitable then your staking plan should allow for you to place 1000 bets given a neutral expectation of yield. The rest will be history.

[slapdash]

Re: Percentage staking plan, thoughts?

slap, can you explain how to calculate those chances of losing entire bank?

OK. First, look at a simple situation. Suppose you're betting one unit at time on 3.00 shots with each bet having a 40% (or 0.4) probability of winning. And suppose you start with a bank of N units. To lose the entire bank, you have to first lose one unit, then lose another unit, ... N times. So if the probability of ever being one unit down is X, then the probability of going broke (being N units down) is X^N. So we just need to know what X is. Starting at N units, we have an X^N chance of going broke, but after the fisrt bet we have a 0.4 chance of having N+2 units (and so an X^(N+2)) chance of going broke, and a 0.6 chance of having N-1 units (and so an X^(N-1) chance of going broke). So we get an equation for X: X^N = 0.4*X^(N+2) + 0.6*X^(N-1) or, dividing by X^N: 1 = 0.4*X^2 + 0.6*X^(-1) The equation you get will always have X=1 as a solution, but assuming your bets have positive expectation, there is always another solution between 0 and 1. In this case it's X=0.822875653 (approximately). So your chance of going broke starting with a bank of N units is 0.822875653^N. It's a bit more complicated in situations where you could end up with less than a whole bet at some stage (do you stop there? do you stake it all on the next bet even though that's less than your normal stake? or do you "borrow" enough to make it up to a whole bet?), but the method still gives an answer which is reasonably accurate so long as the size of your bank is fairly large compared to your bet size. So, in my previous example where half your bets are at odds of 2.00 and half are at odds of 3.00, all with +10% ROI (so the 2.00 bets have an 11/20 chance of winning and your 3.00 bets have an 11/30 chance of winning) and you stake 3 units on each bet, then you list all the possible outcomes of a bet, with their probabilities: The bet is odds of 2.00 and wins. You win 3 units. Probability 11/40. The bet is odds of 3.00 and wins. You win 6 units. Probability 11/60. The bet loses the rest of the time. You lose 3 units. Probability 65/120. You add up the probability times X to the power of the profit/loss and set that equal to 1: (11/40)*X^3 + (11/60)*X^6 + (65/120)*X^(-3) = 1 which has a solution of X=0.956587349 (approx.). So if you start with a bank of 30 units, you have a chance of 0.956587349^30 of going broke, which is about 0.264, or 26.4%. I bet you're glad you asked! :D

[aldric]

Re: Percentage staking plan, thoughts?

I bet you're glad you asked! :D

I don't think that explanation was detailed enough, can you try again! :lol :ok

[Hooloovoo]

Re: Percentage staking plan, thoughts?

Wall of text

Impressive and interesting! :notworthy