General Betting
There is currently 1 person viewing this thread.
These 157 comments are related to the topic:
how do u measure ur edge?
how do u measure ur edge?
sort by:
Show
per page
Replies: 157
|
By:
If you lay 1 horse to a liability of 100 then you can lose 100.
But you are assuming liability on bets that never had that liability. Assume I lay 100 horses at odds of 100 So I layed $1.00 on each which had liability of $100.00 There is one winner, which is one loss. Forgetting commission, I broke even. My way: Total Stakes $100 Total Return $100 Total Profit $0 Profit on turnover 0% Or the other way Total liability $10,000 Total return $100 Loss looks like $9,900 on paper. But it isn't. This shows the full futility of using liability. |
|
By:
If you were just to use liability of $100
in the case of including it in the turnover. You'll have the same result of 0%, but only because it's 0% Anything less or greater will be minimalised from the true picture. To each their own, but I think total liability diminishes your true and correct edge percentage. Just as including turning over $100,000 in stock when you only outlayed $50,000 (but the truck could have turned over, and does one time in a thousand. |
|
By:
BTO: You're assuming you lay multiple horses.
If you laid 100 horses IN THE SAME RACE at 100 then you'd HAVE AN OVERROUND. Not only that but your liability is REALLY easy to work out, it's ZERO. Your liability isn't your liability on EVERY BET if you lay multiple in the same race it's your MAX LIABILITY IN EACH MARKET. |
|
By:
I need to stop arguing with you BTO, you never understand what I am saying, to the point where I wonder if it is wilful misunderstanding.
If you lay 2 horses in a 3 horse race at 4/1, then your liability is 3 on each horse is 3. since that is the most you can lose. Your book would look like: +2 -3 -3 That's a liability of 3. That's the same as backing the favourite for 3 units, of course, but we've been though that one  . |
|
By:
BTO: You're assuming you lay multiple horses.
No, not in the example I just gave. If you laid 100 horses IN THE SAME RACE at 100 then you'd HAVE AN OVERROUND. Not only that but your liability is REALLY easy to work out, it's ZERO. Your liability isn't your liability on EVERY BET if you lay multiple in the same race it's your MAX LIABILITY IN EACH MARKET. Is it in practice? No it's not, you can never lose the $100. The stake on every other horse is taken off your loss on the horse that won (you lost on) So you can never lose the full $100 in one race. |
|
By:
If your liability is 100, you can lose 100, that is what liability means.
What are you talking about "the stake on every other horse"? The only stake is the one you accepted, and you return that to the winner when the horse you laid wins, along with the liability. |
|
By:
If you lay multiple horses in the same race then of course you reduce your liabilty as you do, but if you lay to a given liability then you will be dutching such that your liability is large, so the liability on any one horse is even, and if you only laid that one horse would be larger than your overall liability.
|
|
By:
I need to stop arguing with you BTO, you never understand what I am saying, to the point where I wonder if it is wilful misunderstanding.
I'm not trying to wind you up, perhaps it's a communication thing. Please forget multiple horses in one race for a while, that's just going to confuse the issue. |
|
By:
OK, forgetting multiple horses.
If you lay a horse to a liability of £10, then if that horse comes in you lose £10. Do you agree/disagree with this? |
|
By:
Where was I when a thread needed me?
 I think the fact that henok is just another super chimp has been dismissed far too quickly. |
|
By:
Agreed Ror.
|
|
By:
Glad I ducked out of this whole pointless debate when I did.
|
|
By:
Not like you to shirk a pointless debate,frog.
|
|
By:
I know GJ , but I'm slowly learning.
And I do now have the experience of knowing one when I see one. But I was in it for quite a while as you might have noticed. So still a way to go. |
|
By:
i think turnover is a term too confusing to use it in betting. a more appropriate measure can be average profit over average stake which is simple to use and reveling enough for both traders and gamblers in both back and lay fields. that said when calculating the average profit it is better to separate the lay profit and back profit as they are at the opposite ends of an event definition.
|
|
By:
Isn't laying just betting dutch ?.
What's the difference ?. In other words what does the phrase "--they are at the opposite ends of an event definition" actually mean ? |
|
By:
FAFH , if we define a team win as an event the back is for the event happening and the lay is a back of the event not happening as u said it dutching all other possible outcomes. this two probalities should add up to 1. as a result as long as the probality of the event is not around 0.5(the odds being even shots) ur win and stake vary inversly.
therefore when considering average profit per average stake to measure ur edge this may lead to a suppression of possible insight u may get from ur analysis. eg if u back short odds ur profit is small and ur stake is large. for lay the opposite. when u take ur average profit the wins from the lay of the short odds will be more expressed, and in the average stake the stakes from the figures from the "back" of the short odd will be more expressed. this may be especially very important when considering a small sample size. hope i explaind it well |
|
By:
While I'm not sure I completely follow henok's language with "opposite ends of the market", I would agree that when measuring your edge, if you start with a known edge then it shouldn't matter whether that edge is against long or short odds.
That is to say that roulette has the same edge against someone backing 20 numbers compared to those backing 1. This is another reason why I prefer comparing profit to losses incurred rather than total staked. Risking a grand at 1% isn't as risky as risking a grand at 10%. |
|
By:
henok
sorry clear as mud. A lay is a bet and vice versa. If you consolidate them all in your figures you will get no distortions, as long as you do it correctly in proportion. That if I bet for i unit liability, then the lays I compare it with must be also for 1 unit liability. |
|
By:
oky, was on the turkish game. i will explain my self with an example, and will try to define clearly some things that i came to understand from the discussion here and my thoughts about measuring edge. my last comment to separate lays and backs.
first to clarify when i say separating lay i am not meaning laying all other possibltie s which in a sense means the same as backing. What i mean is back and lay of the same event as in trading. eg event: manchester winning: possible bets: back manchester win, lay manchester win this two should be separated when calculating average profit and average stake fitst i have came to understand that a better and easy way to measure ur edge or profitablity is by dividing average profit/ average stake(liablity) this tool will clearly cover the risk factor associated with ur bets. |
|
By:
err laying the same event as you have backed is the same as backing the other outcomes.
It doesn't matter a bit when or why you do it imo. |
|
By:
to give an example
measure of edge =average profit/average stake eg: backing once in four outcome at 4.4 possible bets: put 1, win 3.4: 1, loss 1, put 1 lose 1, put 1 lose 1 average profit over a long run (3.4-3)/4=0.4/4 average stake over a long run 1 edge= 0.4/4/1=0.4/4, the interpretation using this formula in the long run u are risking 4 to win 0.4 lay once in four event at 4.4 ur edge is -ve of the above. but for this example let us take a lay at big price where u will win 0.4 in 4 runs eg, let us say u lay once in 4 outcome at 3.6 possible bets and results, put 2.6 win 1, put 2.6 win 1, put 2.6 win 1, put 2.6 lose 2.6 average profit=0.4/4 average stake=2.6 edge=0.4/(2.6*4) risking about 10 to win 0.4 ur edge measured with the profit per stake formula shows that the back bet at long odds is bigger than the lays. during analysis this formula may interprets ur back and lay bets diferntly with difernt outcomes. this doesnt make sense. yes? it is because this formula ignores the propablity of 4 consecative losing bets in the back and lay bets. u can do the maths . this probalities are difernt. if u consider a 1:4 back and lay the probality of 16 consecative backs losing is equal to 4 consecative lay loses. thereforre to correct the ignorance of this fact either somebody has to use profit per stake formula at difernt probality levels or he has to normalize the stakes using the probality of the stake losing . i hopw i have explained it well, but thanks for the siscussion it was well informative. i would continue discussing but got to go the CL games. sorry about the languge and grammer |
|
By:
Well that should cause a grinding halt to the thread.
|
|
By:
well didnt mean to stop the discussion.
|
|
By:
British black sense of humour.
Ignore. Hopefully somebody can work their way through your last post and give it the answer it deserves. Unfortunately ( or fortunately ?) it won't be me. |
|
By:
|
|
By:
Measuring your edge using stake size in some way is flawed as demonatrated by the amount of discussion above. You are trying to take into account if the bet is back or lay when the economics might be the same, and how do you treat a position that is traded out?
Given that you have a betting history to look at the best test of the edge is to look at some measure of the P&L variation over time. To come back to the same example I used earlier, suppose you have two guys: A. Is up £365 for the year, having been up and down approximately £1,000 each day. B. Is up £365 for the year, having made £1 profit each day. They might both have the same stake but guy A could easily wipe out his profit this afternoon whereas B has clearly beat the market - you'd probably only want to know what's stopping him scale up. I have a very good track record of ending threads myself but hope you're still with me henok. |
|
By:
Divide etc
Many people argue that trading out should be treated just as separate positional bets. Their only nexus being the commission offset ? A good or bad argument ? |
|
By:
I haven't read much of this thread because it looks too boring but I will add, in case it hasn't already been said, that I think it is a huge red herring to try and shoehorn gambling activity into definitions and terminology that really don't fit or add any value to its analysis.
|
|
By:
Well I have to say Cat that for someone who has not read the thread in detail, you have amazingly summed up its content extremely accurately.
But if we applied such exacting and perceptive standards to all the threads on here, then we would end up with nothing being posted. And no, I am not for once being facetious at all. You are spot on re this particular thread. |
|
By:
OOh Frog, someone summed up a thread and it wasn't you. That's got to hurt. Still good recovery by summing up the summing up.
|
|
By:
I would measure your average edge (per bet) as the compounding growth effect the bet should have on your bank, this alleviated all the measurement problems with the differences between back and lays and favourites and outsiders, i.e.
Backing 100 (and the results coming up like a 20 shot) or Laying 1.01 (and the results coming up like a 1.05 shot) will roughly equate to a similar edge of around 4.2% You can then average this calc for every bet to get your average edge. Any long-term positive edge you have in specific circumstances can then be optimised by staking. |
|
By:
Jabmast
Sounds both extremely commensensical and mathematically rigorous to me. I trust all others accept it as so ? UA I'll accept any compliment at all I can dredge out of you. A first even maybe ? Certainly in recent memory. |
|
By:
I think you're right jabmast, I've been thinking about this and I'm increasingly convinced that stake amounts shouldn't affect your measure of 'edge'.
|
|
By:
Stake amounts could affect your average edge but they should first be converted into banks (amount the growth %age is related to) and then used as purely as weighting factors.
|
|
By:
It still all boils down to how really useful developing all this info is.
Theoretically interesting and perhaps even useful up to a certain point, but beyond that just an exercise in curiosity ? |
|
By:
I suppose if it gives one the confidence to really start using an " edge" to a greater level of liability, and hence potential profit, then it is useful to develop it as far as it is mathematically possible and sound to do so.
But for most it is not really the most important factor in improving their gambling. |
|
By:
Not at all, I've never actually calculated my average edge but would be fairly trivial to do (likely figure would be very low), but have used actual/expected edge as an incredibly important figure in my gambling.
|
|
By:
Jabmast
Yes I forgot the " Kelly " factor. I don't follow the so called " value " mdethod of betting, where staking is correlated to " edge" levels. |
|
By:
BANK SIZE is irrelevant to whether you have an "edge" or not.....
ROI is far better measurement................. |
