Forums

General Betting

Welcome to Live View – Take the tour to learn more
Start Tour
There is currently 1 person viewing this thread.
THE-FOX
03 Nov 11 16:07
Joined:
Date Joined: 22 Oct 11
| Topic/replies: 3,395 | Blogger: THE-FOX's blog
i know the answer is yes  but can someone confirmed is it ok to back 1.05s that should be 1.02s
Pause Switch to Standard View IS PRICE IRREVELANT IF THERES VALUE
Show More
Loading...
Report catflappo November 15, 2011 4:30 PM GMT
Archimedes wouldn't have needed a bath to understand value Wink
Report jimmy69 November 15, 2011 5:58 PM GMT
I think you have misunderstood me. There is no point getting so called "value" when your selection loses...
Report Cooee November 15, 2011 6:21 PM GMT
Jimmy, if I let you back Man City to beat QPR at the odds of 3.0, you'd snap my hand off. Right?   

Now, if Man City hit the bar five times and missed two penalty and ended up not scoring, the bet would have lost.       You could play that same game ten times and have Man City win virtually every time. With odds of 3.0, you'd be making a fortune.  It just so happens that on this one occasion, you were unlucky and the bet went against you. Are you really trying to tell us that backing Man City at 3.0 was a bad bet, simply because it lost on that particular occasion?
Report TheInvestor2 November 15, 2011 6:26 PM GMT
If you lay 'no corner' at 1000 for every single match where there is a market for this in football, you will be getting fantastic value, as fair odds avg about 15000.

If you happened to lay 'no corner' in Wigan v Chelsea last year though, you'd have lost, as it ended 0-6 with zero corners.

If you repeatedly back heads at evens on a coin toss that has a 50.5% chance of landing on heads, you can easily be down after 10,000 tosses despite getting value.
Report TheInvestor2 November 15, 2011 6:27 PM GMT
Cooee, I hope you won't be faced with the usual argument that this is 'unrealistic', despite the fact that bets of that kind of value are available every day.
Report jimmy69 November 15, 2011 6:48 PM GMT
Of course it was a bad bet. It lostSad
Report Cooee November 15, 2011 6:51 PM GMT
Well Investor, getting a bet like that is most definitely unrealistic for mere mortals like myself and Jimmy. However, I was using an extreme example to show how stupid it is to think that long-term success has anything to do with a single event - if the luck favours you, even a minnow can slay a giant over one match, even though they might get slaughtered on every other occasion if the match was to be replayed 20 times.

Regardless of whether you can find the equivalent of the Man C bet on a regular basis, every day there are cases of events going off where, for example, one participant has a genuine 50% chance of winning, but is priced as though it only has a 30% chance of winning.  If you back the 50% participant again and again, you'll sometimes lose and you'll sometimes win. But overall, you'll make money because you're getting value.

Jimmy's point is an utterly stupid one that shouldn't really need refuting. The fact that he honestly believes his is the right way to think tells you why, by his own admission, he's been betting for many many years and still struggles to be successful.  DrWhom's point, though, is of rather more relevance. Unless we're crooked, we can't 'know' the actual winning probability of any participant in a future event. All we can do is use our skills and tools to come to as accurate an assessment of the actual winning odds as possible. 

Some events will be easier for us to work out than others. But assuming we only bet when we feel we've got a good handle on that event, and also assuming we've been able to come up with estimates that are reasonably close to the actual probabilities of winning, and ALSO assuming that we believe the market is significantly overpricing or underpricing one of the participants, then we have a potential bet.  A lot of assumptions in there, but that, I believe, is what we should be aiming for...
Report Cooee November 15, 2011 6:52 PM GMT
Of course it was a bad bet. It lostSad


CryCryCryCry

I give up!
Report jimmy69 November 15, 2011 6:53 PM GMT
So a losing bet is a good bet...interesting
Report Cooee November 15, 2011 7:01 PM GMT
JIMMY YES!!!!     Assume you were able to play the equivalent of the Man City bet 20 times in a row. It loses the first time. At that point it's looking like a poor bet, but that was simply because you were unlucky on that one occasion - you can always be unlucky in a single sporting event.  But over the course of the remaining 19 bets, the odds should right themselves.

The end result is that you lose 3 times but win 17 times. You've won 17 times with odds of 3.0. Hence you've made a very big profit.  The fact that you lost that first time is pretty much irrelevant by the time you get to Bet 20.

In the long term, the only thing that matters is that you get a good price to begin with. If your starting price is good, the rest will take care of itself over the long term.

Honestly Jimmy. How can you not see this rather basic point????
Report catflappo November 15, 2011 8:11 PM GMT
jimmy winding everyone up I reckon.  When I clearly had him earlier he just made a joke - quite a good one so he's not as daft as he's making out.
Report Cooee November 15, 2011 8:16 PM GMT
For his sake, I hope you're right Cat.  There do seem to be quite a few Betfairers who have the same blind spot though.
Report catflappo November 15, 2011 8:28 PM GMT
Long may they continue to pay my daughters school fees. Laugh
Report Cooee November 15, 2011 8:29 PM GMT
Laugh
Report jimmy69 November 15, 2011 9:03 PM GMT
I wouldn't dream of winding anyone upTongue Out
Report catflappo November 15, 2011 9:09 PM GMT
Wink
Report FINE AS FROG HAIR November 15, 2011 10:08 PM GMT
As I said before Jimmy you won't win in the long run unless you're getting value.
But you won't know till well into the long run as to whether you are getting value bets, and then it will be too late to do anything about it if you're not.
What a bugger this betting game is.
Report Cooee November 15, 2011 10:58 PM GMT
Frog, if you're placing two, three, four bets a day (and I'm sure many here place far more than that!), you'll probably know within a few weeks if you're not getting value!   I don't think long term needs to be *that* long term...
Report DrWhom November 15, 2011 11:14 PM GMT
"There is only one true probability of any given outcome"

Is that true ? You may be right but I'd love to see a mathematical proof.
It may however be that probability is like velocity and therefore relative. It could be that probability depends on how events and observers are related.

Perhaps someone clever, like Bayes, could give us the benefit of their thoughts (not that you're not clever, of course).
Report FINE AS FROG HAIR November 15, 2011 11:36 PM GMT
Cooee
Let's be honest, for most on here a couple of weeks would be extremely long term.
Most probably give up on a strategy if it doesn't work on its first time outing.
Dr Whom
An interesting philosophical question for sure. But of any other potential value to us " ordinary " punters on here ?
Report jimmy69 November 15, 2011 11:48 PM GMT
I've been on BF for over 10 years, turned over goodness knows how much  (100K per week on average), and am currently about 2K down overall...is it worth it I ask myself?Excited
Report FINE AS FROG HAIR November 15, 2011 11:54 PM GMT
That's a bit of a stretch jimmy, even for the likes of me.
Report jimmy69 November 15, 2011 11:59 PM GMT
I've enjoyed every moment of it FroggyGrin
Report FINE AS FROG HAIR November 16, 2011 12:04 AM GMT
Your life would have been much emptier without it all ?
Gonna write a book about it all one day ?
Perhaps titled along the lines " A decade in the life of an unsuccessful, but enormously happy, gambler "
Report jimmy69 November 16, 2011 12:07 AM GMT
That's a thoughtExcited
Report Cooee November 16, 2011 12:29 AM GMT
Long Post Alert...


DrWhom, I think you're missing the point by asking for 'mathematical proof'. If you want to know how many times the number 10 will come up on a roulette table, you have two options. You can mathematically calculate the chances of that 10 coming up. Alternatively, you can spin a roulette wheel several hundred thousand times. Both options should give you pretty much the same outcome.

But if you want to calculate the winning probability of an outcome in a sporting contest, you have something of a problem.  Mathematically, a roulette table is relatively simple. The factors that affect the chances of something occurring are essentially known and fixed. But a horse race, for example?  How many factors are involved there? The answer is thousands if not millions. You have to be able to work out the various factors that revolve around pace, class, speed, trainer, weather, surface, distance etc.  And those are just vague terms. Actually breaking them down into what they really mean and how they affect the probability of a particular horse winning is difficult, if not impossible.

And the experimentation approach isn't open to us either. If we want to know the probability of the 10 coming up, we could just spin the roulette wheel hundreds of thousands of times. But we can't have the same horse race run thousands of times. And that's the essential problem. There's no real way of testing the outcome of something happening in a horse race. Not with total precision. We can't physically run the same race more than once. And since the number of factors that determine a horse winning are huge (and many will remain unknown to even hardened horse players), we have little chance of mathematically calculating the chances of something happening either.

And yet, the chances of something winning must exist. The outcome of essentially everything can theoretically be boiled down to a figure. If we hold a coin with the obverse side facing us and let it drop three feet, a decent mathematician/physicist would be able to work out how often it'll land on a particular side. If you let it drop from 300 feet, a good mathematician/physicist would probably be able to calculate that too. But what about if we drop it from the Moon? There's no reason to suppose there isn't an answer to how many times it'll land with the obverse side facing up. But how many factors would go into calculating the answer?   

Everything must theoretically have a chance of being so. But in some cases the chances of that thing happening are easy to calculate, and in some cases they're almost impossible due to the sheer weight of factors.   I would say, with sporting events, you can be accurate enough to get to within a few percent of the actual probability under some circumstances. The very best players will get closer than that. However, as long as you have a reasonably effective means of calculating the probability using important and quantifiable factors, and you only take a bet when the prices are out by 20% or more, you should be able to make money.
Report FINE AS FROG HAIR November 16, 2011 1:19 AM GMT
A key factor in pricing a specific sports event, as compared to the analogy of pricing games of chance ( such as roulette), is that huge amounts of unpredictable one off items come into play in the former. Sending offs, in game injuries, bad referee decisions to name a few in football.
And so I'd say that it is mathematically impossible to price one stand alone sports event.
The best one can do is to try to be able to estimate the outcomes  of a reasonably large sample of such events to a satisfactory degree of accuracy, allowing for any and all unexpected or potential eventualities to occur.
Report Trevh November 16, 2011 1:30 AM GMT
20% is quite a large buffer Cooee, I like to play if I think I can get even 5%. (You're not RTTF are you?)


Jimmy69 : I've been on BF for over 10 years, turned over goodness knows how much  (100K per week on average), and am currently about 2K down overall...is it worth it I ask myself?

That's not bad Jimmy, you've paid thousands in commission but are only slightly down, so you've beat the markets on the whole and nearly beat commission too, but not quite. Just imagine how much profit that would equate to if you had managed to get just one extra tick on all your bets, if you ever bother to work it out it would be thousands.
Report Cooee November 16, 2011 1:35 AM GMT
But you could theoretically price in, for example, the probability of a red card happening and how that would be likely to affect the game if it did happen. That would be just one extra factor that went into creating 'the figure' that you got at the end.  Not easy to calculate, but theoretically possible..

In some ways the 'unpredictable' aspect is higher in roulette, where the actual run of the ball is nearly all luck.
Report FINE AS FROG HAIR November 16, 2011 1:36 AM GMT
But even you are not saying trevh that you are always sure that the price you can get on any one particular event outcome is 5% better than the " actual " probability , or are you ?.
You're talking average prices over a series of like events aren't you ?
Report FINE AS FROG HAIR November 16, 2011 1:41 AM GMT
Cooee
Not sure I know what you're saying here.
To me the ball spin unpredictability in roulette is not analgous to the unpredictability of unforeseen circumstances unfolded within a game of sport.
For example how do you price in a  " red card" into match price odds ?. Is that even really possible, in the sense of ending up with a price that would ever be achievable pre- match ?.
Report Cooee November 16, 2011 1:42 AM GMT
Trev, some day I'd like to be able to look at 5%.  However, I don't think my system is so accurate yet that I'd be confident of getting that close.  To be honest, I tend to look more for 12-15% rather than 20%. But I do see a few horse racing bets every week where I believe the value is in excess of 20%, and I believe most decent players ought to be able to come up with a system that gets them close enough to the real figure to make a 20% 'edge' profitable.
Report Trevh November 16, 2011 1:51 AM GMT
Yes average prices Frog, reckoning that if I can get that price it will pay 5% profit over time, and confirmed by reviewing long term results.

You're a football frog Frog, so for example have you noticed the way mad backers are backing the short odds correct score 0-0's this year, not with small amounts of money either. Matches that were kicking off between 9.0 and 10.0 last season are all being hammered in between 7.0 and 9.0 this season. That makes them a value lay to me at just over evens for the entire first half -  a far better price than laying the half time 0-0 market at say 2.6.
Report Cooee November 16, 2011 1:52 AM GMT
Well as a crass method, from looking at lots of past results you could calculate how frequently a football side got a red card. You could then calculate what difference it tended to make to how well the team fared.  Using those two figures, you could generate a figure for today's game that took into account the chance there might be a red card and what effect it might have.

As I say, any individual game has a fair degree of luck and unpredictability attached to it. Things like the likelihood of red cards and their probable effect would be factors you would have to take into account.   Either that or you make sure that the 'value' you accept is wide enough to allow for unpredictable factors. That's why I might prefer to look for bets where I estimate the value to be 12%+ rather than the 5% that a more talented player like Trev can settle for.
Report FINE AS FROG HAIR November 16, 2011 1:54 AM GMT
Yep trevh there are quite a few other type " seemingly " wrong long term odds.
I wonder why ?
Pretty easy to spot if you have a reasonable data base tracking program comparing results to pre-match off prices.
Report Cooee November 16, 2011 1:57 AM GMT
And remember Frog, those factors will often work in the favour of your team as well. Sometimes your team will go down to ten men and lose. Another day your opposition might have two of their key players taken out with injuries. If you're looking at bets in sequences of 10 or 20 or 30, those things shouldn't be too destructive to the final figure.  Any individual game is a bit of a lottery, but nowhere near as much as one spin of a roulette table.
Report FINE AS FROG HAIR November 16, 2011 1:57 AM GMT
Cooee
We could debate all day on these specific matters eg the timing of a red card etc, but I actually think we are in general agreement on this whole question of value.
So why bother. Just say good luck to each other ?.
Report FINE AS FROG HAIR November 16, 2011 1:58 AM GMT
I agree, nothing is more unpredictable than just one spin of a roulette wheel.
After all, no way would I ever play russian roulette with a gun. Would you ?
Report Cooee November 16, 2011 2:00 AM GMT
Depends whose head the gun was being held up to!
Report Cooee November 16, 2011 2:02 AM GMT
And I'm happy just to say good luck to you and leave it at that.  I'm not sure how else you're going to spend the early hours of the morning??  (I know you don't ever sleep!)

But I'm fine with agreeing to, erm, well, whatever it is that we've agreed to do..  Farewell
Report FINE AS FROG HAIR November 16, 2011 2:02 AM GMT
And what result you wanted of course.
Report FINE AS FROG HAIR November 16, 2011 2:04 AM GMT
Farewell ?
Sounds so serious and permanent.
You're just off to sleep aren't you ?
As you say, I'm a napper not a sleeper.
Report Cooee November 16, 2011 2:08 AM GMT
Well farewell is something you generally say to somebody you may well not speak to again for some time. This is the first time I've spoken to you in months, so who knows when we'll speak again? Until that shining moment, farewell!


(And no, I'm not going to sleep. Sleep is for wimps!)
Report FINE AS FROG HAIR November 16, 2011 2:10 AM GMT
We've spoken before then ?
And it was a shining moment then too ?
Well I never.
Report Cooee November 16, 2011 2:17 AM GMT
Well no, actually it was a pretty grubby moment. We spoke once or twice before when I was allowed to boast the title 'Lord Bobbin'. To be honest, I think you only ever spoke to me because I was the only other person who tends to be up at 3,4 in the morning...  I find being the only person available to talk to boosts my popularity considerably!
Report FINE AS FROG HAIR November 16, 2011 3:17 AM GMT
Yeah I find that things do often seem to get quite grubby at 3,4 in the morning.
Almost could perhaps be called  " the grubbing hour ".
Report Methane_Magnet November 16, 2011 9:36 AM GMT
Having read this thread, I can only conclude that my old friend Jimmy has hit hard times and has sold his forum name to Paulme Cry

Either way - the net is bulging to breaking point! - a fine haul Jimmy.
Report Contrarian November 16, 2011 9:40 AM GMT
Dr Whom,

I think you can reasonable define the probability of an event in terms of a potential infinite number of instances of that event: to say that an event E has a probability p of occurring is to say that as the number of trials of E-like events tends towards infinity, the ratio of E occurring to E not occurring will tend towards p : (1 - p). 

This gives the notion an objective base, although it still makes it difficult to substantiate it fully, mainly because it is unclear what constitutes a 'trial of an E-like event'. In the case of a coin toss, it is fairly clear. An E-like event is just another 'random' coin toss. In the case of a sports event, it is much more difficult, because the trials of E-like events would presumably be (theoretical) reruns of the event with exactly the same starting conditions but allowing conditions from that point to vary.

I think the standard way to flesh out this notion of multi theoretical reruns of E-like events would be in terms of possible worlds. So, to say, for instance that Man U have a probability of 0.7 of beating Sunderland at home next week is to say that in 70% of the nearest possible worlds in which Man U play Sunderland at home next week, they will win.
Report Cooee November 16, 2011 10:09 AM GMT
Is Contrarian clever enough for you, DrWhom?
Report askari1 November 16, 2011 12:51 PM GMT
I agree about poss. worlds in the sense that I can't have any truck w/ people who say after a horse has won that it had 100% chance of winning.

The compiler's difficulty is in determining what is the same in the world in which the current betting event is about to start and the world that hosted previous similar events. Is Torres in the form he was for Liverpool? Is his head right? Is the horse fit? Even if we had an exhaustive list of all the variables involved in the outcome, and an accurate formula for their weighting, we would have problems assigning the correct valuables for each of our factors.
Report Sandown November 16, 2011 2:07 PM GMT
Theoretical parallel worlds, true probabilities, multi-variables - all good fun but essentially useless in the practical world. The importants points for me are:

1. Every sports event is unique and occurs only the once; no two events are ever the same.

2. Randomness is a major factor in sports events which gives rise to the unpredictable nature of results.

3. More information does not necessarily result in better forecasting

4.Every handicapper carries a different forecasting model (in his head or in his computer) and every forecast of odds is therefore individual.

5. The sum of these different forecasts is reflected in the market which is more accurate over the totality of events than any one forecaster can ever hope to achieve

6. Everybody is 100% correct with hindsight.Nobody is or can be 100% correct with foresight.
Report DrWhom November 16, 2011 8:50 PM GMT
Thanks for the answers, and yes, cooeee, Contrarian is quite clever enough I'm sure. However, when I said 'mathematical proof', I meant that there is only one probability of an event happening. Contrarian's answer is the standard concept of probability but I'm not convinced its necessarily the best way for probability to be expressed. I'm not certain what an alternative could be but I'm working on it. I still like the idea of probability as a relative term, rather than an absolute.
Report FINE AS FROG HAIR November 16, 2011 8:52 PM GMT
In the meantime we'll all get on with making some winning ( and losing) bets eh ?
I hope you do too.
Report DrWhom November 16, 2011 8:54 PM GMT
"Theoretical parallel worlds, true probabilities, multi-variables - all good fun but essentially useless in the practical world"

All good science starts from  theory and is then brought into the real world after testing and analysis. Therefore its short sighted to say its simply 'fun but essentially useless'.
Report FINE AS FROG HAIR November 16, 2011 8:57 PM GMT
Btw I've had a terrible football betting week.
What's the probability of that ?
Of course it's all relative.
Oh no it isn't, it's absolute.
See what a pickle you've put me into DrWhom.
Report DrWhom November 16, 2011 8:57 PM GMT
"I hope you do too. "

I sure do and have been for over 20 years. But all my algos have started from a theory of mine which I've tested, modified and then either abandoned or made use of.
Report FINE AS FROG HAIR November 16, 2011 9:00 PM GMT
As has everybody else on here presumably, that is those who are succesfull.
I'm all for creative thinkning and endorse any efforts you are making at such.
Report frimpong November 16, 2011 9:37 PM GMT
Cooee no longer using LB username?
Report FINE AS FROG HAIR November 16, 2011 9:43 PM GMT
Must be catching all this name changing .
Does it change your fortunes somehow ?
Report jimmy69 November 16, 2011 10:20 PM GMT
Why do people change their names? What is the benefit?
Report frimpong November 16, 2011 10:47 PM GMT
can be necessity rather than benefit
Report TheInvestor2 November 16, 2011 10:48 PM GMT
Sandown
16 Nov 11 14:07 Joined: 06 Dec 01 | Topic/replies: 954 | Blogger: Sandown's blog
Theoretical parallel worlds, true probabilities, multi-variables - all good fun but essentially useless in the practical world. The importants points for me are:

1. Every sports event is unique and occurs only the once; no two events are ever the same.

2. Randomness is a major factor in sports events which gives rise to the unpredictable nature of results.

3. More information does not necessarily result in better forecasting

4.Every handicapper carries a different forecasting model (in his head or in his computer) and every forecast of odds is therefore individual.

5. The sum of these different forecasts is reflected in the market which is more accurate over the totality of events than any one forecaster can ever hope to achieve

6. Everybody is 100% correct with hindsight.Nobody is or can be 100% correct with foresight.



I don't get point 5. Isn't anyone that beats the market by skill (as opposed to luck) more accurate than the market over the totality of events?

If we were to play a prediction game, where a group of people has to guess the probabilities of various events occurring for which there are markets, I would obviously expect the results of my 'guesses' to be more accurate than the market. The reason for this is that I would continuously agree with the market (take the market price as my guess), except for those occasions where I had a high level of confidence that the market price is wrong.

I would agree that it would be nigh on impossible to outperform the predictions of the market if you didn't have access to the odds in the first place, and had to do it completely from scratch.
Report FINE AS FROG HAIR November 16, 2011 11:41 PM GMT
In the world of pure economics, price reflects supply/demand for something and not inherent value. Correct or not ? I'm not an economist btw.
If the sports betting world is the same, then the price of an event at the off surely just reflects the sum of all conflicting predictions/opinions on the outcome of that event, but only as they exist in equilibrium at the actual start time.
It has no specific correlation with good value per se, whether on a one off basis or a long term series basis, does it ?
Report Cooee November 17, 2011 2:08 AM GMT
I didn't change to the new name by choice. I invented the alternative username 'Cooee' about a month ago to prove how easy it was to create a new username!     Cooee was on an alternative account that I didn't use and so I didn't give it any further thought. But then a couple of weeks ago my main forum username suddenly changed to Cooee.   I had jumped around between accounts so that I could carry on being LB, but a few days ago that name just disappeared altogether.   So, I guess I'm now Cooee forever!

I must say, I miss old Lord Bobbin. It was quite nice being titled for a while, but I suppose he just got bored with me and moved on. I did hear a rumour he was seen last week in Monte Carlo throwing everything on red. But somebody else assures me they saw him running along the Yorkshire Moors baying at the moon. So, who knows where he is...

Maybe some day he'll drop by and let us know how he is? Until then, Good Night LB. Wherever you are..
Report frimpong November 17, 2011 2:12 AM GMT
surely he is the forum name associated with the account you place bets with Confused
Report Cooee November 17, 2011 2:18 AM GMT
You would think that, wouldn't you?   But I assure you, I've been using the same main betting account for months, and it's changed of its own accord.   Life of their own these usernames..
Report frimpong November 17, 2011 2:22 AM GMT
Shocked
Report Trevh November 17, 2011 2:24 AM GMT
Yes I agree Frog, the price also being overly influenced by recreational punters with deep pockets, i.e. one punter with deep pockets can temporarily move a price to a good value position to be picked off, but which will simply look like the true price to many others who don't have specialist knowledge in that market.
Report Cooee November 17, 2011 2:29 AM GMT
Frog, you are right that the price of an event does reflect 'the sum of all conflicting predictions/opinions'. But the wisdom of crowds does appear to work quite well here - the consensus of opinion will often be quite close to the real figure. In the long term, the favourite in the US Pools wins more often than the 2nd favourite, who wins more often than the 3rd favourite etc. So clearly the markets do get it right a lot of the time.

However, the consensus won't *always* be right. If you have a decent means of forecasting the odds yourself, and can identify those races when the consensus opinion has got it wrong (or those individual selections in a race that the consensus has got wrong), then you're onto something. And that 'something', in my opinion, is what most successful punters are searching for.
Report Cooee November 17, 2011 2:31 AM GMT
Erk.. Trev seems to have jumped in ahead of me there...
Report Trevh November 17, 2011 2:34 AM GMT
Sorry Lord B :)
Report Cooee November 17, 2011 2:35 AM GMT
Don't mention his name. He abandoned me. LordB don't live here any more.
Report catflappo November 17, 2011 8:43 AM GMT
I think the forum login is now seperate from the site login. I have several accounts and several usernames. I have noticed that often I am posting under a different forum name than the one associated with the account I'm logged into.

I suspect that LordB is still surviving somewhere with all his b's intact and we'll see him back here sometime soon.
Report Sandown November 17, 2011 10:26 AM GMT
Investor2

I don't get point 5. Isn't anyone that beats the market by skill (as opposed to luck) more accurate than the market over the totality of events?


I may have been guilty of not allowing for the remote possibility that someone, somewhere,is capable of beating the market over all markets over the very long-term. I would guess that for us mortals, to beat the market over the long term requires us to be selective and to build up superior know-how in a finite number of markets. In a head-to-head with any individual across EVERY market I would put the market on-odds on to win comfortably.I suspect that most of us suffer from a certain degree of arrogance which leads us to believe that we, alone, can do this.

(I am limiting these comments to our ability to forecast events and I exclude strategies which involve rapid trading, technological edges, time edges or just plain cheating.)
Report kenilworth November 17, 2011 10:57 AM GMT
'''as long as you have a reasonably effective means of calculating the probability using important and quantifiable factors, and you only take a bet when the prices are out by 20% or more, you should be able to make money.

I dont bet unless I have, IMO, a 20% margin, to allow for error. Unfortunately,
although I show a profit overall, I don't make make anything like 20% therefore
I need the margin.Sad
Report TheInvestor2 November 17, 2011 11:10 AM GMT
Sandown
17 Nov 11 10:26
Joined:
06 Dec 01
| Topic/replies: 955 | Blogger: Sandown's blog
Investor2

I don't get point 5. Isn't anyone that beats the market by skill (as opposed to luck) more accurate than the market over the totality of events?


I may have been guilty of not allowing for the remote possibility that someone, somewhere,is capable of beating the market over all markets over the very long-term. I would guess that for us mortals, to beat the market over the long term requires us to be selective and to build up superior know-how in a finite number of markets. In a head-to-head with any individual across EVERY market I would put the market on-odds on to win comfortably.I suspect that most of us suffer from a certain degree of arrogance which leads us to believe that we, alone, can do this.

(I am limiting these comments to our ability to forecast events and I exclude strategies which involve rapid trading, technological edges, time edges or just plain cheating.)


I still don't really get it. Do you mean that if you were forced to actually bet on every single market, the market would win?  What I meant was that default mode is: 'market is correct'. Each case where you agree with the market is effectively a 'draw'. You cherry pick those cases where the market is likely to be wrong, where you will expect to beat the market. Taken together you beat the market as a whole. I guess this is not what you mean though?
Report TheInvestor2 November 17, 2011 11:34 AM GMT
FINE AS FROG HAIR
16 Nov 11 23:41
Joined:
12 Mar 07
| Topic/replies: 4,180 | Blogger: FINE AS FROG HAIR's blog
In the world of pure economics, price reflects supply/demand for something and not inherent value. Correct or not ? I'm not an economist btw.
If the sports betting world is the same, then the price of an event at the off surely just reflects the sum of all conflicting predictions/opinions on the outcome of that event, but only as they exist in equilibrium at the actual start time.
It has no specific correlation with good value per se, whether on a one off basis or a long term series basis, does it ?


Yes, I'd agree with all of that, although I think this might be redundant: " but only as they exist in equilibrium at the actual start time."

The interesting point here is that for something like an apple, there is no objective inherent value. The best proxy for this is the market price. An apple might be worth 50p to you, and 80p to me.

I would be willing to pay 500 quid a month for my internet connection if I had to, and it would be rational for me to do so if that was the market price and all other conditions remained unchanged, that doesn't mean the price I'm currently paying is wrong (under 50 a month).

However, if one person is willing to pay 500 for a bet that another is willing to sell for 50, one of them is wrong (factoring in only profit incentive and ignoring things like entertainment value).

A bet has an objective inherent value (although it may not always be possible to find this value exactly).
Report Sandown November 17, 2011 11:37 AM GMT
Investor2

You cherry pick those cases where the market is likely to be wrong, where you will expect to beat the market. Taken together you beat the market as a whole. I guess this is not what you mean though?



Yes it is what I mean, but the cherry picking means being selective, concentrating on where one might have an edge. By combining the skills and knowledge of a small number of proven players would widen the spread of success, but of course, continual spreading must be finite anyway because otherwise it leads to market matching.
Report frimpong November 17, 2011 12:17 PM GMT
I dont bet unless I have, IMO, a 20% margin

can you give an example please ken?
Report DFCIRONMAN November 17, 2011 12:31 PM GMT
WTF are you .....FP.....asking such a daft question?
Report Cooee November 17, 2011 12:37 PM GMT
How would an example help?  He presumably means a bet where he estimates there's a difference of 20% or more between the winning percentage implied by the true price and the winning percentage implied by current price.
Report catflappo November 17, 2011 12:44 PM GMT
kennilworth

I dont bet unless I have, IMO, a 20% margin, to allow for error. Unfortunately,
although I show a profit overall, I don't make make anything like 20% therefore
I need the margin.


Do you factor commission into the 20%?
Report Sandown November 17, 2011 1:07 PM GMT
I dont bet unless I have, IMO, a 20% margin, to allow for error. Unfortunately,
although I show a profit overall, I don't make make anything like 20% therefore
I need the margin.


Unfortunately, the figure by itself is of little use unless you know what the % strike rate is. At 90%  strike rate, 20% is superb and there will be little volatility there. At 10% strike rate, you could easily lose over 1000 bets, 5 times out of 10. At this level of strike you really need to be looking for nearer 40% edge to be sure of winning over 1000 bets.
Report Johnny The Guesser November 17, 2011 1:14 PM GMT
How do you know when you have,or have not, got value as opposed to have just been lucky or unlucky?
Report Sandown November 17, 2011 1:15 PM GMT
The OP provides just a 3% return on stakes but because of the high strike rate (98%) the profit, although small across 1000 bets, would be a near certainty to provide that profit every sequence of 1000. Herein lies a truth about the way consistent traders win money on here.
Report Cooee November 17, 2011 1:15 PM GMT
Well I was the one who made the original 20% comment.  In my case, I meant that if the market was putting a value on the horse that suggested it should win that race 30% of the time, I'd only recommend backing it I believed it would really win at least 50% of the time. That way my estimate of the true winning figure could be 10-12% out and I'd still be able to make a profit.  I'm sure some people can work with a 5% figure, but most of us need to allow for a much larger degree of error.
Report Cooee November 17, 2011 1:20 PM GMT
Johnny, you don't really. But if you're not making money over a good string of bets then you probably haven't got it right.   

The market price is usually (but not always) somewhere close to the real figure, so if your estimates tend to be accurate, then they'll also be within a few percent of the market price on a regular basis. If you're getting huge disparities race after race (particularly if you're losing money long-term as well) then you probably need to rethink your method of estimating the price.
Report Sandown November 17, 2011 1:21 PM GMT
Cooee

That's a 20 PERCENTAGE POINTS edge. A 20% edge would be a win strike rate of 36%. Or 20 points equals a 66.7% edge. Just a question of definition.


Johnny TG

Staitisticians usually want a 95% or even a 99% chance of the results being significant ( more accuratle a 5% or 1% chance of NOT being significant). That means a large sample which can be calculated from the strike rate.
Report Cooee November 17, 2011 1:22 PM GMT
Yep Sandown. These threads do get very confusing since many of us use the same terms in different senses..  (Particularly amateurs like me!)
Report Johnny The Guesser November 17, 2011 1:31 PM GMT
OK.

Lets see if anybody can help me with this.

- 250,000 chimps play on betfair for a year having 10 x £100 bets a day into 101% markets. They pay 3% commission on winning bets. How many will be in profit at the end of the year?
Report Eddie the eagle November 17, 2011 1:53 PM GMT
Not enough information I think.  I think there would be a much bigger % in profit if they were betting on 100/1 shots rather than at evens every time.
Report jimmy69 November 17, 2011 2:15 PM GMT
Oh ffs...just back the winners...stop blathering on about this nonsenseExcited
Report Sandown November 17, 2011 2:20 PM GMT
jimmy

An excellent strategy.Guaranteed to make large consistent profits.

Just remind me again...how do you identify the winner BEFORE the event?
Report jimmy69 November 17, 2011 2:22 PM GMT
I use a pinGrin
Report Sandown November 17, 2011 2:23 PM GMT
Gosh. As simple as that. Where do you buy your pins?
Report jimmy69 November 17, 2011 2:25 PM GMT
Any hardware store should provide you with the necessary...Excited
Report kenilworth November 17, 2011 3:03 PM GMT
An example of 20% margin ? I thought it would
be obvious but here we go...
An 1/1 chance would be 2.20, the commission
covered by the 20%. If successful with
assessment, produces in theory 7% profit,
although tbh my overall profit is just under
5%, the reason why I only look at 20%+.
Report Sandown November 17, 2011 3:54 PM GMT
kenilworth

Easy mistake to make but 2.2 gives a 10% return on stakes. You need 2.4 for 20%.


johnny TG

Some of your 250,000 chimps/chumps would pay PC for sure
Report Johnny The Guesser November 17, 2011 4:11 PM GMT
Yep...and think they had 'cracked it'...and turn pro....and boast on forums about how much they make....and offer advice to others...
Report kenilworth November 17, 2011 6:25 PM GMT
Sandown, I'm talking about 20% on the price, not on the profit,
quite different. It becomes increasingly difficult to maintain a % profit
margin, the shorter the price, in fact backing, for instance 1.50 chances,
a near impossible 94% strike rate would be required.
Post Your Reply
<CTRL+Enter> to submit
Please login to post a reply.

Wonder

Instance ID: 13539
www.betfair.com