Punting - why value is the key

Colonel Sanders · 20-01-2013 4:02pm

so to summarise:

without the law of large numbers the statement "you cannot win money in the long run without obtaining prices that under estimate your selection" would be a conjecture, a subjective topic to be debated

with it the statement becomes a fact grounded in mathematical rigour

Colonel Sanders · 20-01-2013 4:10pm

RoverJames wrote: »

I've mentioned already anyway but I'd consider a lifetime punting total of 750 bets over 50 years as very small due to there being 10,000 + ish races in the UK and Ireland per annum.

Small in what regard? You're focusing on bets (s)he hasn't had

If this person "only" has 750 bets over the course of 50 years I'd be very confident (in fact I'd say the probability is close to 1) that there have been value bets this punter has failed to identify

By me saying 750 was "large" I meant that if someone can show a +ive RoI over the course of 750 bets I'd assign a reasonably high probability that the +ive RoI is due to the punter having obtained value over the course of those 750 bets rather than mere chance

The sample to concentrate is the bets the punter has actually placed, not the number of bets that could have been placed

RichieLawlor · 20-01-2013 4:27pm

RichieLawless wrote: »

Went to number 43 and saw the words 'coin toss'. So didn't bother with the other two.

Lol

Colonel Sanders · 20-01-2013 4:42pm

Colonel Sanders wrote: »

By me saying 750 was "large" I meant that if someone can show a +ive RoI over the course of 750 bets I'd assign a reasonably high probability that the +ive RoI is due to the punter having obtained value over the course of those 750 bets rather than mere chance

The exception would be if this punter told me they regularly bet on:

1. BAGs
2. Cartoon sports
3. Numbers

However showing a +ive RoI over 750 bets in these "disciplines" would be very unlikely

20-01-2013 5:19pm

RoverJames wrote: »

Any post relying on the law of large numbers can be disregarded as its not relevent to selective punting. Throws of the dice and tossing coins are fair enough, someone knowing their beans can conceptually pick events where a single selection has what they consider an extraordinary chance of success, as its their view no one could actually comment on if the selections are value or not, people might interpret them as being value due to high ROI but that's a tenuous link with a small sample size.

Some people try and find value, using methods that are in their view sound, others buy into the bookies being able to do their job most of the time so they look for horses with extraordinary chances of winning.

As these people might get on before the bandwagon they might well be getting value, that wasn't what brought the selection to their attention though.

Colonel Sanders wrote: »

Small in what regard? You're focusing on bets (s)he hasn't had

If this person "only" has 750 bets over the course of 50 years I'd be very confident (in fact I'd say the probability is close to 1) that there have been value bets this punter has failed to identify

By me saying 750 was "large" I meant that if someone can show a +ive RoI over the course of 750 bets I'd assign a reasonably high probability that the +ive RoI is due to the punter having obtained value over the course of those 750 bets rather than mere chance

The sample to concentrate is the bets the punter has actually placed, not the number of bets that could have been placed

You've just highlighted the point you keep missing, if a punter spots a horse with an extraordinary chance of winning and backs them, the positive ROI can be interpreted as value, sprinter sacres last few races are nice examples

In the long run doesn't come into selective punting either imo, not enough bets, you can't appreciate this either.

Failing to identify any value is immaterial, they buy into the bookies being consistent. As you've mentioned for the last time above we can look forward to you not repeating your not applicable views on selective punting, which is what Huntley practises, successfully too.

Colonel Sanders · 20-01-2013 6:01pm

RoverJames wrote: »

You've just highlighted the point you keep missing

I'm not missing anything

RoverJames wrote: »

if a punter spots a horse with an extraordinary chance of winning and backs them, the positive ROI can be interpreted as value

No it can not. Just because a bet wins does not mean the punter received value. This is not an opinion this is fact. In the coin toss example the punter betting 4/6 has never, not once, got value no matter how many times he collects and no matter what his RoI is (even if that RoI is +ive)

1/50 is probably an example of "an extraordinary chance of winning" (just over 98% probability) but if the punter is betting 1/100 they are not getting value. This is the case whether the horse wins or loses

You seem to be saying:

Punter shows a +ive RoI => that punter has got value.

This is not necessarily the case

What I am saying is that the larger the number of bets a +ive RoI is based on the higher the probability the +ive RoI is due to the punter being a "good" bettor and not simply enjoying a run of pure luck

RoverJames wrote: »

In the long run doesn't come into selective punting either imo, not enough bets, you can't appreciate this either.

I appreciate you think this but I can't appreciate it is mathematically sound because it isn't.

I'm not sure how how can come onto a thread where I am discussing winning in the long run and not consider the long run

RoverJames wrote: »

Failing to identify any value is immaterial

No it is not. Despite your refusal to acknowledge it the law of large numbers proves (not suggests) otherwise

Colonel Sanders · 20-01-2013 6:33pm

I just want to reiterate, a small sample size does not invalidate the law of large numbers

It merely makes it harder to ascertain if a +ive (-ive) RoI is due to "good" ("bad") punting or down to luck (or variance)

This is precisely what the law of large numbers states

20-01-2013 7:29pm

Colonel Sanders wrote: »

.............

You seem to be saying:

Punter shows a +ive RoI => that punter has got value.

.......................
.

I'm not saying that at all, I said it could be interpreted as that, which is what you are doing, by assigning a high probability that the +ive ROI was due to the punter having obtained value. I'd see that as an incorrect interpretation, but let's be clear, its your interpretation, not mine, as illustrated in the piece I've quoted (below in bold). I'd put it down to the individual punter being quite likely knowledgable and good at picking winners.

Colonel Sanders wrote: »

..............

By me saying 750 was "large" I meant that if someone can show a +ive RoI over the course of 750 bets I'd assign a reasonably high probability that the +ive RoI is due to the punter having obtained value over the course of those 750 bets rather than mere chance.................

RoverJames wrote: »

You've just highlighted the point you keep missing, if a punter spots a horse with an extraordinary chance of winning and backs them, the positive ROI can be interpreted as value, sprinter sacres last few races are nice examples

..............

Colonel Sanders wrote: »

.............
I'm not sure how how can come onto a thread where I am discussing winning in the long run and not consider the long run

No it is not. Despite your refusal to acknowledge it the law of large numbers proves (not suggests) otherwise

I agreed with your opening post as to me it seemed to be about betting in general, the lads who punt away all day in the betting office racking up 10s of bets every day illustrate your point perfectly. If they don't chase losses and bet for level stakes they'll lose up to 20% of what they gamble, we all know that.

But you started the thread to continue an attempt to make Huntley look a muppet, I wasn't aware of that when I posted my first post in the thread.

So I added more views that are relevant to the real purpose of the thread.

You've taken his posts in a 100% literal sense and applied them to betting, you seem unable to acknowledge that some people can pick a higher percentage of winners than most of us resulting in a +ive ROI, you maintain any +ive ROI would be most likely due to the punter having obtained value inadvertently. Some of us can appreciate that the +ive ROI is due to the punter picking more winners than would be considered average.

As per the quotes in this post, you contradict yourself, mention I seem to be implying x,y,z which I clearly haven't and in fact you have been !

Having mentioned that you have completed a pure maths degree doesn't make any of your points more valid than they appear in black and white on the screen.

If you actually know what you are talking about you'd have to concede that someone backing at typical bookies prices (ie the bookies have taken their 15%) can show a +ive ROI due to enough of their selections winning.

This is admittedly not as likely as them losing but it is a definite mathematical possibility, it being rare does not make it as impossible as you would like your theories to imply.

There's multiple reasons why the majority of punters make a loss, the inherent lack of value is a huge one but no less a factor is that most people don't really have a clue about what they are betting on.

Getting value gives people an edge, so too does knowing your beans. Some people know more than the rest of us so value isn't as important to them.

Others know just as much as the average Joe so price enhancements etc offer them perceived value.

Colonel Sanders · 20-01-2013 7:53pm

RoverJames wrote: »

I'm not saying that at all, I said it could be interpreted as that, which is what you are doing, by assigning a high probability that the +ive ROI was due to the punter having obtained value

It's a direct application of the law of large numbers, I'd be content to assign a reasonable high probability to the statement "punter X is a "good" punter" based on a positive RoI over 750 bets. You cannot definitively say he "is" or "isnt" based on 750 bets

RoverJames wrote: »

I'd put it down to the individual punter being quite likely knowledgable and good at picking winners.

But picking winners isn't "enough", you need to pick winners where the price offered is too big.

RoverJames wrote: »

But you started the thread to continue an attempt to make Huntley look a muppet, I wasn't aware of that when I posted my first post in the thread.

No I did not, a number of threads had become derailed over this topic. I only replied to Huntley directly when he replied to a post of mine in the Grandouet thread. I replied here rather than in that thread.

In ny case Huntley called me an "idiot" more than once. He i well able to give posters on here a roasting if he believes they are talking rubbish. I'm sure he's big and bad enough to do the same to me if he believes it necessary

RoverJames wrote: »

You've taken his posts in a 100% literal sense

No I've taken posts where he (and others) have not stated opinions on the subject but have incorrectly spouted rubbish as "facts"

RoverJames wrote: »

you seem unable to acknowledge that some people can pick a higher percentage of winners than most of us resulting in a +ive ROI

I acknowledge that some people are better at picking winners than others. However the ability to pick winners is unlikely to result in a +ive RoI in the long run unless as well as being able to pick winners they only bet on these "likely winners" when they obtain incorrect odds

RoverJames wrote: »

Some of us can appreciate that the +ive ROI is due to the punter picking more winners than would be considered average

Define "average". If you mean can identify more horses where odds X/1 are available but the probability of winning is greater than 1/(X+1) then I actually agree

RoverJames wrote: »

you contradict yourself

where, I have been quite consistent

RoverJames wrote: »

Having mentioned that you have completed a pure maths degree doesn't make any of your points more valid than they appear in black and white on the screen.

I never said they did, however I have backed up my points with mathematically rigourous results. my posts being consistent with these rigourous results make them more valid than someone simply saying "you're wrong"

RoverJames wrote: »

If you actually know what you are talking about you'd have to concede that someone backing at typical bookies prices (ie the bookies have taken their 15%) can show a +ive ROI due to enough of their selections winning.

I have repeatedly said that over a finite numbers of bets a "bad" punter can show a +ive RoI.

RoverJames wrote: »

This is admittedly not as likely as them losing but it is a definite mathematical possibility, it being rare does not make it as impossible as you would like your theories to imply.

It is impossible over an infinite number of bets. This is a fact Nobody can place an infinite number of bets (pretty sure I've stated that a few times too). It is possible over a finite number of bets, I have stated this numerous times but less likely the more "bad" bets that are placed (the old law of large numbers again which you claim is "irrelevant")

RoverJames wrote: »

Some people know more than the rest of us so value isn't as important to them

If by knowing "more" this enables them to assign a more accurate probability they are immediately at an advantage, however they still must obtain value prices to win in the long run

It's no point knowing the Willie Mullins hotpot in the bumper is a 4/6 shot if 1/2 is the best price out there (and you've no Betfair account

)

RichieLawlor · 20-01-2013 8:10pm

Is say the Colonel has some pain in his head arguing with this guy. It's not even an argument, it's a complete nonsense from RJ.

The worst kind of poster is one who acts like they know what there are on about when they clearly have no clue.

Im not an expert on it by any means, but I don't think it's that difficult to comprehend. CS isn't even giving an opinion he is stating facts which certain geniuses are trying to dispute.

Colonel Sanders · 20-01-2013 8:19pm

RichieLawlor wrote: »

Im not an expert on it by any means, but I don't think it's that difficult to comprehend. CS isn't even giving an opinion he is stating facts which certain geniuses are trying to dispute.

I have no problem with people arguing my opinions. Opinions are like arse holes, everyone has one

I have a problem with people arguing facts, not "my" facts (I did not invent any of the theories discussed here) but generally accepted rigourously derived mathematical facts

20-01-2013 9:10pm

Colonel Sanders wrote: »

.............

It is impossible over an infinite number of bets. This is a fact ............

We all know this, I am repeatedly referring to a selective punter who places a small number of bets.

RichieLawlor wrote: »

............. it's a complete nonsense from RJ.

The worst kind of poster is one who acts like they know what there are on about when they clearly have no clue.

Im not an expert on it by any means, but I don't think it's that difficult to comprehend. CS isn't even giving an opinion he is stating facts which certain geniuses are trying to dispute.

As well as being no expert on maths, stats or horse racing you are highlighting an inability to comprehend posts.

I have consistently agreed that over a large number of bets anyone will lose as most prices are inherently short of value, the bookies have taken their cut.

When there are 10,000 + events per year (horse races in UK and Ireland) it is quite conceivable a selective punter can pick a small number of horses that have a great chance of winning, back them at 85% of the price they should be in a round book, the majority of them win and he has a +ive ROI.

It's not surprising you come along though with a genuinely useless post that is simply you wading in behind CS while admitting you don't actually have a clue at the same time

I'm not arguing any facts by the way, I'm suggesting they are being applied incorrectly.
Between the lines CS is admitting this as he's blowing on about an infinite number of bets etc.

20-01-2013 9:13pm

Colonel Sanders wrote: »

.......

The law of large numbers formally shows this. If you can tell me how it doesn't be my guest

It is relevant to selective punting if by "selective" one means only betting when the odds are in one's favour, whether such an opportunity comes along once every hour or once every year

I don't mean that.

So the law of large numbers is not relevant.

Colonel Sanders · 20-01-2013 9:15pm

RoverJames wrote: »

When there are 10,000 + events per year (horse races in UK and Ireland) it is quite conceivable a selective punter can pick a small number of horses that have a great chance of winning, back them at 85% of the price they should be in a round book, the majority of them win and he has a +ive ROI.

I actually think this probability (over the repeated example of 750 bets) is quite small

In any case the number of possible events isn't a factor, the numbers of bets placed is

I do think we are both actually saying almost the same thing, we are just arguing over some semantics

20-01-2013 9:22pm

Colonel Sanders wrote: »

I actually think this probability (over the repeated example of 750 bets) is quite small

.....

Of course the probability is quite small. Only a tiny percentage of punters make money, only a tiny amount of Betfair accounts are in profit (despite the inflated prices).

I'm not claiming it's easy for anyone to make money backing horses, but when someone comes along and illustrates they can do so and also shares their views (largely independent of the price) publicly on a forum I'm all for it

I'd hate to think that you reckon I'm arguing with you to be honest.

I'm all for an argument with the lad who waded in accusing me of talking non sense when he has no idea what I'm talking about.

20-01-2013 10:51pm

RichieLawlor wrote: »

.............It's not even an argument, it's a complete nonsense from RJ.

The worst kind of poster is one who acts like they know what there are on about when they clearly have no clue.

Im not an expert on it by any means, but I don't think it's that difficult to comprehend. CS isn't even giving an opinion he is stating facts which certain geniuses are trying to dispute.

Colonel Sanders wrote: »

...................
I do think we are both actually saying almost the same thing.............

Poor ole Richie :pac:

Huntley · 21-01-2013 3:20am

Colonel Sanders wrote: »

In ny case Huntley called me an "idiot" more than once.

I never said that you were an idiot.

I believe that your insistence that finding the subjective term 'value' in selections is the most fundamental aspect of successful gambling is idiotic. I haven't changed my view in this regard and all your arguments (or more aptly described as your continual repetition of theories and statistics) have done is simply offer a different opinion on the issue.

Your belief that that one must adhere to statistics and theories to turn a positive RoI in the long term is flawed. I don't do this yet have already demonstrated my ability to have a positive RoI. That is merely down to luck according to you, I would see it as being able to distinguish the most likely winner when reading race form and thus have continuously stated that I see that as the most fundamental aspect of successful gambling.

You have failed to recognise this on many occasions and merely stated that I am wrong. I have the evidence to suggest that I am not wrong and the ideal of following theories and statistics is essential for successful gambling is far too literal.

RichieLawless wrote: »

Huntley, anything people have against you here is purely personal. I have yet to see you proven wrong. Ireland is full of begrudger's and this forum is no different, why you're still posting here is beyond me? They won't be happy till you're gone, and listen to Bull**** opinions they agree with

I skim over the dross/pointless posts these days ft9 and did the same when I saw that username :pac: Good thing I saw your message and realised my mistake.

I get things wrong ft9. I have no problem making pittance of opinions on here but when I do so it is generally with good reason. This grates many users and when they think there is an opportunity to return the favour there is an arousal like gathering amongst those that have been made look very foolish in the past by myself. If my opinions were as worthless as people state than nobody would bother even taking account of them.

There are a select few knowledgeable people that provide decent insight and its always good to get a different angle, the problem is that the majority of users here are unable to do that. I don't get too caught up in it anyway after all the majority of us have our own lives outside of some forum.

RoverJames wrote: »

Poor ole Richie :pac:

I always found that Dick's continual use of 'Lol' as responses acts as a masquerade. No substance or credible response in the post but it gives others the perception that he isn't as clueless as we know.

In this case, there is no more apt response to that comment than his trademark 'Lol'.

Colonel Sanders · 21-01-2013 12:55pm

Huntley wrote: »

I believe that your insistence that finding the subjective term 'value' in selections is the most fundamental aspect of successful gambling is idiotic

What I define the term "value" to be is not subjective. It is obtaining odds that under estimate the probability of an occurance. I do however agree that any prior estimate of the probability of a horse winning a race is subjective

It is backed up by rigourous mathematics

You have claimed there is plenty of evidence to the contrary yet have presented none apart from hear say and conjecture which I suppose are kinds of evidence

Huntley wrote: »

your arguments (or more aptly described as your continual repetition of theories and statistics) have done is simply offer a different opinion on the issue.

I have repeated the gist of the OP continually as it has been continually accued of being false or irrelevant. It is neither false, irrelevant nor not applicable to any example raised in this thread

It is also not opinion it is (here I go again repeating myself) fact

Huntley wrote: »

Your belief that that one must adhere to statistics and theories to turn a positive RoI in the long term is flawed

My "belief" is that you cannot win in the long run unless you obtain value prices, "value" being defined as +EV.

I am awaiting your proof that the law of large numbers is either false or flawed, as you have repeatedly implicily stated by saying the gist of my OP is flawed

Huntley wrote: »

I don't do this yet have already demonstrated my ability to have a positive RoI. That is merely down to luck according to you, I would see it as being able to distinguish the most likely winner when reading race form and thus have continuously stated that I see that as the most fundamental aspect of successful gambling.

If you have shown a +ive RoI over a long time period you have either
1. been lucky
2. have implicitly only backed what you deem the likeliest winner if you think the odds are in your favour (and have, on average, been proved correct)

Again this is not opinion

Huntley wrote: »

You have failed to recognise this on many occasions and merely stated that I am wrong

I have recognised what I have set out in points 1 & 2 above. You have repeatedly made statemenst that violate the law of large numbers

Huntley wrote: »

I have the evidence to suggest that I am not wrong and the ideal of following theories and statistics is essential for successful gambling is far too literal.

your "evidence" is nothing more than subjective opinion, my "evidence" is irrefutable, rigourous mathematical theory

If you want to prove (note i said prove not argue opinions) that what I have outlined in the OP to be incorrect I would recommend you don't post it here. Send it to any noted mathematical journal and enjoy your nobel prize

21-01-2013 1:29pm

So you are maintaining the law of large numbers is applicable to 750 events picked from over 500,000 events? Picked through knowledge and interpretation of form and racing conditions while ignoring the value side ?
The events I am referring to are horse races not spins of a roulette wheel or dice etc etc.

Colonel Sanders · 21-01-2013 2:14pm

RoverJames wrote: »

So you are maintaining the law of large numbers is applicable to 750 events picked from over 500,000 events? Picked through knowledge and interpretation of form and racing conditions while ignoring the value side ?
The events I am referring to are horse races not spins of a roulette wheel or dice etc etc.

Yes, yes I am. In that I think the probability a +ive RoI over 750 bets is due to the weighted EV being positive (and not luck) is quite high

I am also maintaining the events where there was no bet are irrelevant to its application to the 750 bets

Colonel Sanders · 21-01-2013 2:45pm

Colonel Sanders wrote: »

Yes, yes I am. In that I think the probability a +ive RoI over 750 bets is due to the weighted EV being positive (and not luck) is quite high

And lets be rigourous. Last time I posted a similar comment you said you wanted a simple Yes or No answer

You cannot give a simple yes or no answer unless the RoI is based on an infinite number of bets. The best you can do is:

1. Estimate the probability the +ive RoI is based on luck or being a "good" punter
2. Derive a number of bets that would give a stated probability p that the +ive RoI is based on luck or being a good punter

If this isn't acceptable to you then I'm sorry but that is the best applying the law of large numbers can do or was designed to do if the sample of bets is finite. And nowhere have I claimed otherwise

The only time I stated I was happy that 750 was "large" I caveated my "Yes" reply with what I have outlined above

21-01-2013 3:16pm

Colonel Sanders wrote: »

Yes, yes I am. In that I think the probability a +ive RoI over 750 bets is due to the weighted EV being positive (and not luck) is quite high.........

So in tandem with this you'd have to concede there is a chance (admittedly small) that a selective punter could have a +ive ROI even though is selections weren't value, simply a disproportionate amount of them win

?

Every mathemathician or statistican I have met always understands that the probability of x happening being quite high does of course imply the probability of y happening is no doubt possible, yet admittedly unlikely.

You genuinely seem to struggle with that mathematical fact.

Colonel Sanders · 21-01-2013 3:37pm

I have posted on numerous occasions that a "bad" punter (one that places a series of -EV bets) can show a +ive RoI over a finite series of bets

I understand that and have posted it several times (including in the OP) so I fail to see how you are interpreting that I'm "struggling" with anything.

Trust me the only thing I'm "struggling" with is:
1. Being asked the same question over and over again
2. People stating that some of what I'm posting is opinion

Colonel Sanders · 21-01-2013 3:42pm

RoverJames wrote: »

Every mathemathician or statistican I have met always understands that the probability of x happening being quite high does of course imply the probability of y happening is no doubt possible, yet admittedly unlikely.

I'd like you to indicate anything I've posted that suggests I'm "struggling" with this? In the part of my post you quoted I said the probability was "quite high". Nowhere did I say the probability was 1.

The only thing I've stated is impossible in this thread is a +ive RoI due to "luck"
over an infinite numbers of bets

I've repeatedly recognised a "fluke" may occur over a finite series.

the probability decreases as the number of bets increases (but only reaches 0 when the number of bets reaches infinity).

You either:
1. Don't understand what I am posting
2. Don't understand the concepts I'm posting about.

You definitely appear to be struggling with the inferences I've drawn from finite sample sizes

Colonel Sanders · 21-01-2013 4:03pm

So in a nutshell RJ where did I state that a "bad" punter could not show a +ive RoI over a series of 750 bets (or in the parlance I've been using where did I state that the probability a "bad" punter would show a +ive RoI over 750 bets was zero)

You seem to have implied that I did indeed state that.

Morgans · 21-01-2013 4:19pm

What I dont get is why people are criticising Colonel Sanders purely mathematical approach - disputing the law of large numbers and say that value isnt key. They are proposing that if you bet the best horse the value will take care of itself. Fine, I can see that point.

However, it also means that the best punted horses are those with the clearest chance to win. Meaning that the best way to punt is to have more on the 1/20 shot, the most sure of winning and should be looking at the 1.01 shots on betfair.

However, If some odds on shots are better than others, then automatically, you are saying the value is key.

Nobody is suggesting that you have to bet on the horse with most value in every race - there conceivably could be multiple horses priced wtih value in a single race.

The key is picking the horse you think will win the race AND that horse representing value for long term success.

It is why Pricewise is valued and Keiran O'Connell is not.

Colonel Sanders · 21-01-2013 4:23pm

RJ am I correct in saying you think the law of large numbers cannot be applied and that inferences cannot be drawn based on its result if the sample size is not infinite

If this is the case then wow, I've given you far too much respect in this thread.

Colonel Sanders · 21-01-2013 4:27pm

Morgans, I've repeatedly stated what you have but still been argued against

If you don't want to focus on value then fine. If you want to pick the likeliest winner with no regard for value then fine

However do not say that what I'm stating in the OP (which is not opinion btw) is incorrect

21-01-2013 5:23pm

Colonel Sanders wrote: »

I'd like you to indicate anything I've posted that suggests I'm "struggling" with this? In the part of my post you quoted I said the probability was "quite high". Nowhere did I say the probability was 1.

The only thing I've stated is impossible in this thread is a +ive RoI due to "luck"
over an infinite numbers of bets

I've repeatedly recognised a "fluke" may occur over a finite series.

the probability decreases as the number of bets increases (but only reaches 0 when the number of bets reaches infinity).

You either:
1. Don't understand what I am posting
2. Don't understand the concepts I'm posting about.

You definitely appear to be struggling with the inferences I've drawn from finite sample sizes

You're struggling with it as for you to accept it can happen you have to either label it a fluke or label the punter a bad punter.

This shows you're application of the theory you are spouting is flawed, plenty of folks with degrees don't know what they've studied fully, lots of them won't admit it.

750 bets is a tiny number, take winning SPs in lots of 750 and you'll see substantial variance.

Neither pricewise or koc are what I'd call overly selective btw, and to repeat I agree with your op.

21-01-2013 5:26pm

Colonel Sanders wrote: »

RJ am I correct in saying you think the law of large numbers cannot be applied and that inferences cannot be drawn based on its result if the sample size is not infinite

If this is the case then wow, I've given you far too much respect in this thread.

Inferences can of course be made with sample sizes of less than infinity, and for dice rolls etc 750 I would consider a large number, not for horseracing though as There are too many variables.

Punting - why value is the key

Comments