Mathematics and the "Edge" in Punting

[Rock Steady]

A question for Mr J:

If my long term strike rate is 29% and my POT is 15.6%, what info do I need to calculate my percentage edge?

[moeee]

I'd like to comment on your question until Mr J arrives.
You'ld think,at first glance,that your edge is the POT of 15%.
If you were playing an even money game then that would be true.But you are taking various prices about your selections.
Supposing you tested at 1000 events,then at level stakes you would have invested $1000 but not necessarily got back $1150.
Only if you were betting level stakes in your declared POT would it be true.

[La Mer]

On 2004-10-08 08:40, Rock Steady wrote:
A question for Mr J:

If my long term strike rate is 29% and my POT is 15.6%, what info do I need to calculate my percentage edge?


All of what Mr. J has written on the issue of roulette is true - it's a negative game and therefore can not be beaten in the long term by any system or staking plan, although it might be possible to have some sort-term success, eg. having a single bet by placing your whole bank on an evens payout and walk away a winner if successful or walk away broke if not.

Don't laugh, it's probably the best chance of success that roulette offers.

Rock Steady it's not so simple to work out your advantage, this has to be taken on a bet-by-bet basis, eg. if you think you have a genuine 2/1 chance of winning and you can get odds of 4/1 then you should bet 13.3% of your bank: if you think you have a genuine 1/1 chance of winning and can get odds of 2/1 then you should bet 16.7% of your bank etc.

Now betting to such levels of your bank is normally too much for most people so a lesser amount should be bet, eg, 25% of the advantage which in the two examples above would be 3.3% and 4.2%, but if your prices are 100% spot on then to maximise your returns your stake should be equal to your advantage.

The problem is of course in non-casino games (in most other cases) the price is nothing more than an opinion and therein lies the problem. It all comes down to being very, very accurate with your assessment/opinion.



<font size=-1>[ This Message was edited by: La Mer on 2004-10-08 11:47 ]</font>

<font size=-1>[ This Message was edited by: La Mer on 2004-10-08 11:48 ]</font>

[moeee]

No offence to you La Mer,But I have been involved in a fair bit of friction on this Forum.
And I can see why,after reading your post.
You talk about roulette in answer to Rocksteady's question,but I don't see any reference in his question to that form of betting.
And this is why I think we really need to read what it is people are writing their thoughts about and not presume there is something sinister in between the lines.
Again,not having a go at you,La Mer,and I found your comments very interesting nevertheless.
But to the people who wish to criticise others for their poor efforts,well it seems over 95% of punters end up losing.
And that is the purpose I use of this forum.To share what I know and to learn from others in a way that can be used as a guide.
Well done La Mer for giving something,not necessarily the question poster,but to others that read the forum.

[La Mer]

On 2004-10-08 13:20, moeee wrote:
No offence to you La Mer,But I have been involved in a fair bit of friction on this Forum.
And I can see why,after reading your post.
You talk about roulette in answer to Rocksteady's question,but I don't see any reference in his question to that form of betting.


You are correct, there is no direct reference to roulette in Rock Steady's question. However, if you read the roulette system thread, Mr. J makes a number of points, one being that to win you need an advantage (gambling with an edge), which is what Rock Steady asked about (the edge). My response was based on that.

But it really matters little, for Rock Steady didn't actually state what game (casino, sports betting, horse racing etc) that his stats referred to, for in casino games such as roulette there is always a negative math, while unless arbitraging in sports betting or horse racing there is always only an opinion price.

[Mr J]

This is for sports/horses??

La Mer is right in that you can really only estimate your edge in sports/horses, since the odds are based on opinions (not fixed like in casino games).

Your longterm POT is generally a good indication of the size of your advantage.

It'd help if we knew how many bets your've placed.

[Rock Steady]

Mr J,
I am talking greyhounds. I have detailed records for the last 7 years.
No of plays = 6484
No of winners = 1880
Units returned = 7545

[Mr J]

Your sample is very large. Hitting 15.6%POT over that many bets means your edge is probally around 15%. Also says that you are definately profitable longterm.

97.5% chance that you will hit 8.7% going forward as well. You're also pretty safe betting 8.7% of your bankroll on your picks.

Out of interest, what are the betting limits on the dogs?

[knowledge]

Mr J I have read what you've been saying since I came here and it sounds like you have a lot of wisdom in gambling. I have a lot to learn and I think you could help me.

I'm interested in knowing how you approach gambling. What sports do you bet on, how do you decide what amount to bet, what systems do you use to choose your winners, etc.. I'd appreciate it a lot if you could answer those questions.

Thanks.

[Rock Steady]

Mr J,

In Australia, there are virtually no limits on dogs BUT if you are betting dogs with IAS, Sportodds, Sportsbet you have to keep the bets low otherwise they will bet them back on the tote. I keep my bets with these three down to $150 Maximum. With the three totes (Unitab, NSWTAB and Tabcorp) the bets really have to be kept down in relation to the size of the pool. I never bet more than 150 on Unitab, $200 on the other two as their pools are much larger than Unitab.
I generally bet no more than $400 per dog so I spread it evenly over the six organizations.
Also, 8.7% of the BR per bet seems too high.
My longest run of outs is 17 and I have had other losing runs of 12 and 13. If I had bet this high on Friday, I would have lost over half my bankroll. My start bank each year is $50,000 and I draw $6000 per month for total living expenses.

[woof43]

Hi, I just ran your figures thru my sheets, it appears you avg payoff is just under $3.00 working on your figures the kelly optimum is 5.21% so to allow for some risk an some gain its wise to stick to approx half of this figure.

[Rock Steady]

woof,
My average dividend is $4.01.
I am betting roughly $400 on a BR of $50K, so I am currently betting less than 1%.
Should I be betting a percentage of the increasing bank rather than a flat stake?

[thefan]

How old is everyone here. 50k is alotta money.

[Mr J]

Yep sorry, I just realized 8.7% is full kelly for a 50/50 bet.

Average price is actually $3.99, with a stikerate of 28.99%. Fully kelly suggests betting 5.24% of your bankroll on each bet.

It is better to bet a % of your bankroll, you experience exponential growth.

Like Woof said, betting Full Kelly is a bit dangerous. My opinion is you should be betting 3.59%. With your record, there is a 97.5% chance you will hit 10.7%POT or greater going forward, meaning 3.59% would be the full kelly figure. You are very safe betting this %.

So to answer your original question, there is a 97.5% chance that your edge going forward is at least 10.7%.

As to bet size, you should choose a size you are comfortable with, as long as it's not overbetting. You might not feel comfortable with 3.59%, but you should definately be betting more than you currently are. Maybe 2 to 2.5%???

Betting a larger % will lead to larger swings, but your BR will also grow alot quicker. Taking 6k out of a 150k BR hurts less than taking it out of a 50k BR.

[Mr J]

Knowledge.

First of all, I'm in this to make money, not enjoy myself. I still enjoy betting, but if I wanted just pleasure I'd be betting more of my own picks.

I make my own picks on AFL/NRL/Super 12s using a system I created. This system is based of statistics, not historical trends (sportznut :wink:) Has performed pretty well (10%+) and gives around 300-350 bets. I really did this to see if I could beat sports myself, off my own numbers etc.

Most of my bets come from this site and handicappers (guys who analyse and create picks) I know. E.g, I get tennis, baseball and soon probally NBA from this site (2000 bets right there). I also bet on US college sports (bball and football), NFL, NHL and golf from guys I know.

There are around 4000-5000 bets per year there, all with an edge of at least 4%. Most of these are around 10%.

I figure why create my own picks when I can just use someone elses. The amount of time it takes to create my own picks is just better used finding other guys to follow. I do alot of research when I find someone I'm interested in following.

It's important to take your time when trying to find cappers to follow. I'm not talking 'touts' (who are garbage, they're either frauds or in denial), I'm talking handicappers. Guys who use skills and information to come up with bets that actually have an edge. It takes time (as this site would know) to find someone who is genuinely profitable.

I try to find a handicappers in major sports (sports with high betting limits and a large volume of games...). I look for guys who have at least a longterm 5% edge over a statistically significant sample.

I also hit sportsbook promo's. Things like deposit bonuses, low vig (where the house edge is less than the standard 4.5%) free half points etc etc. You can make a TON of money of promo's.

If you want a good place to start, you can't go wrong with this site. There really aren't many genuine and profitable sites out there, but you are at one so take advantage of it.



thefan

21, and yeh 50k ain't bad at all. I'd be looking to turn that into at least 200k within a year :wink:

[knowledge]

alright thanks man. I appreciate your help. Hopefully I can learn from you and the others at these forums.

[Rock Steady]

the fan,
Age 44.
The $50K is my working bank - I actually have about $20K spread across the three totes and online books. The other $30K is in a Macquarie Bank 24hr Call A/C so if I need it, it's there.

Mr J,
Thankyou very much for the info. On the figures I should be betting around $1000 per play but on a bad week I could drop $12K.
Psychologically this is a tad much for me.
The most I've lost in a week over the past seven years is $5500 and that hurt although I had it back within 3 weeks.
I also have to be careful not to overbet the dogs as my cream comes when I back $8+ winners. If I increased the bets on those dogs, the dividends would drop slightly.

[woof43]

Hi, I would start with an Overall Bankroll, then set aside a Session Bankroll, this Session Bankroll would be based on your anticipated run of outs.
Not knowing you betting history but based on my own experiences an using your information, your Base Bet per betting series should be approx $80 then extending to a Max Bet size of $390. these too figures would be a % proportion of your Session Bankroll.
The best idea would be to look at your past history, find the avg Number of bets in each betting series before you have a drawdown.
Then work at what point in your betting series was your maximum bankroll an your minimum bankroll, eg i avg approx 120 races per week that invest i find that on avg my minimum bankroll is approx around the 30-45 races envelope an my maximum is later, with this knowledge you can then adjust your Kelly optimum % to determine your Base Bet an Maximum bet size.

[moeee]

Too many people try to make something simple difficult.
It is so simple to me to bet a given selection inproportion to the chance I give said selection.
If you like something in one race more than in another,then you put more on it!
Kelly,maths,POT's,.I don't know!
Well maths maybe.

[thefan]

Thanks for sharing your information with us. I usually back grey hounds only. Ive had my good and bad days. But the bad days out weight the good. Noting that I just back placing,the dogs still like to take the long route around the course.

Can anyone convince me the advantages of backing greyhounds.

[Mr J]

Mo, these things are pretty important if you are a serious bettor.

Rock, I get why you don't want to be that much. Variance can be a b!tch :wink: Thing is, it just doesn't make a difference longterm.

No pain no gain :wink:

[Mr J]

Can anyone convince me the advantages of backing greyhounds.


Inefficient market. Plenty of opportunities. Inside information is probally obtained pretty easily. Don't know anything bout the dogs though.

[knowledge]

Has anyone ever had inside information on a race being fixed?

I once bet on a fixed soccer match but it was so obvious that it would end fixed. It was Sweden vs Denmark at EURO2004 and the only result that would get both Scandinavian countries through (by eliminating Italy) was a 2-2 draw. 2-2 was offering $6.00 and it was a good bet. Ended 2-2, what do you know. :wink:

(it may not have been fixed, just a major co-incidence)

[Rock Steady]

thefan,
Mr J is spot on!
Greyhounds in Australia provide the best opportunity for profits out of the three race codes because of INEFFICIENT MARKETS.
The favourites are overbet by the public.
Notice also that the selections given out on the radio or internet by the racecaller are always shorter in the market than they should be.
I love betting at Casino on Friday afternoons:
Big pools because lots of blue collar workers at the pub on Friday afternoon with their pay cheques.
Bad dogs which means if you can find the leader at the first turn, you will be on the winner on 85% of races.
Circular grass track suited to front runners. This is why the first sectional is so important. My strike rate at Casino is lower than other tracks BUT when I do back winners there I often get good odds. It is not uncommon to find a dog which has the best sectional that is well boxed that speeds to the front at $10 and wins by 5 lengths. Bad dogs can get pushed over the line at Casino. Lismore used to be the same until they changed the surface to sand.

[goldmember]

Knowledge, dont know about horses but i have had inside knowledge[ no pun intended] about a few dog fixes a few years and they were VERY profitable, but on the other side of it, pissed me off one night at maitland dogs, i had a dog in a race and it had NO hope of winning but was a quick beginner and would have led [box 5,and always led ]walked out of the boxes and was all over the track which was strange for him, found out later that the pro punters had taken 3 dogs in trifectas and collected,a trainer/punter that i knew,that was in the know told me as i was going home that he was GOT AT and i showed him that the dog could not open his eyes,the bastards had sprayed something in its eyes whilst in the kennels.

[Mr J]

Rock, what sort of money can you put down on the dogs, and are online odds generally pretty competitive?

[moeee]

Golmember.You were there and you would know better than me.
But that sort of thing is as bad as paedophilia.That no-one was brought to justice is the crime.
Although it might not have really happened as you describe it.
The trifecta bit maybe,but the eyes bit makes me sick.
Not because I am an animal lover,but to think that it may have been covered up by authorities worries me.
I believe it should have been brought to the attention of the stewards,Goldmember.
Of course no-one likes wearing concrete boots!

[goldmember]

Moeee, it happened exactly as i said it did, i was leaving to go home, noticed his closed eyes and got started talking to a friend thats when i was told the he was got at and told about the trifecta rigging, apparently he was not the only one,couldnt work it why HIM,as he wasnt much good an a 40/1 shot on the TAB, i expected him to lead and weaken in the straight, when he finished racing i gave him to my parents and they kept him as a pet for years.they still have his sister, she'll be 13 the day after the MELBOURNE CUP.

[moeee]

Don't really want to keep on with this,but 3 dogs in a trifecta leaves 5 dogs that need spraying in the eyes!

[knowledge]

Spraying the poor dogs' eyes is just wrong. If you have to fix a race do it the traditional way, over-feed the dogs.

[thefan]

Hmm no wonder I lose even on place. Pretty much every dog I chose has a great start then run off course because of another dog.

[Rock Steady]

Mr J,
Online books will let you have any amount on the dogs BUT they only pay tote odds (normally Home State).
You could have $5000 on a dog with IAS but you can be sure that IAS will lay off on the tote ensuring that the dog will pay $1.05 at best.
When I first started betting with online books, I got caught out on a quinella. I liked two dogs at Angle Park and took a quinella for $200. the dogs duly ran 1st and 2nd. The winner paid $7.20, the second dog was showing $6.00 for the win so you would expect the quinella to pay a bit more than $2.70!!! No doubt the book backed the quinella back on the relevant tote.
Practically, you can only bet small with online books otherwise the above will happen. I normally spread the action over the three totes and the three online books.

[Mr J]

How small, $500???

If I just used a couple of online accounts, what I be giving up much in the way of odds?

[knowledge]

Mr J, you're up late. :wink:

I'm just curious, how many bets will you make today? I know it's a random question but I'm interested (after hearing you make thousands of bets annually).

[Mr J]

I'm always up :wink: well not from around 4-5am till noon.

Around 7-8 today. A few of those are NFL bets though. October is pretty quiet.