Doing doubles with value bets

[mick bones]

This should be a simple maths question on probabilities so just checking that I have this right before I go down the wrong road.  I have never been a fan of doubles till I sat and looked at results of my approx 10,000 singles bets

With average profit of 30% I'm doing better than I expected when starting almost 1yr ago.  Problem is I am running out of options to get any sort of money on as have lost most accounts.  I only do £5 or £10 bets ... but I do a lot of them as you can see and I record everything.

I have noticed that some bookies will allow me do doubles for up to £5 but singles are limited to less than £1. 

So if you have a proven method of finding value bets ( and I believe I have that, with avg odds of 5.0 ) then should 2 of these bets in a double produce profits as follows

Bet A , avg value of 30%

Bet B, avg value of 30%

So with doubles avg profit after 10,000 bets would be   1.3 * 1.3  = 1.69  ie 69% profit?    For £1 bet, avg win of first leg = £1.30 and this goes on 2nd leg to compound the % profit .

 

I can probably get the answer by modifying my spread sheet to pick 10,000 random doubles and check results but as my first guess is you just compound the % profit then I get a 69% profit .. which I like the sound of.

I understand why bookies like doubles because when you have two -% negative value bets the a double just compounds the % loss long term,  BUT ....... when you have value to the tune of 30% then the bookie is on the wrong side of a 69% value bet when doing doubles ?

[Snoopdog]

Mick B,

In general, I'm very much a singles punter. I have convinced myself the worst value bets are those between 4/6 and 6/4 (the range that the public and I seem to love). Geoff Banks once said: "the public love a shorty, but generally call it a day at 4/6".

I am now starting to trial doubles with approx 1/2 chances. (A pair of 1/2 double pays 5/4). The strike rate is looking higher than might be mathematically expected comparing 1/2 with the 4/6 - 6/4 chances. Early prices essential as the 1/2 frequently become 1/3 or shorter later on.

Early days.

[mick bones]

What I'm trying to get clear in my head is if you have a data set of 10,000 or so singles which proves your single bets have Y% value,  then should rolling those value bets into doubles compound the value to give (2xY)% value?  My maths head tells me it does but just looking someones take on it who is better than me at probabilities and stuff.   Have plenty of selections crop up each day,  just need another way for the fox to get into the hen house as the bookies have the front door closed on me effectively.

[BillyHills]

Quote

With average profit of 30% I'm doing better than I expected when starting almost 1yr ago.  Problem is I am running out of options to get any sort of money on as have lost most accounts.  I only do £5 or £10 bets ... but I do a lot of them as you can see and I record everything.

Surely you could be accommodated on the exchanges for this sort of money?

I have to say Singles is the best way forward for any serious punter, as you are well aware by the sounds of it, especially if you have a decent strike rate. Good luck with the doubles but i'm guessing it may be a little frustrating along the way.

[mick bones]

Nah,  exchanges are way way too sharp. Looking at some data I have from the exchanges after comm I would make no more than 2 or 2.5% profit at the available prices.  This is less than 1/10 of what I can make with the bookies ( Basically snapping up value prices with them,  hence losing accounts as they don't like people taking value )  So to make the same with the exchange I would need to be staking in the region of £100+ for each £10 I get a bookie to take and in majority of cases this money is not available.  But my original question is more about the maths than the logistics of the betting as I still have the option of going on foot into bookie shops and this is where I will be upping the stakes to compensate for lower prices and time taken.

[arsenalfh]

I'm sure you can make money this way as it sounds you're making the right selections. The big problem here is going to be variance I would say but if you have the bankroll to deal with it this could be a new approach for you. By combining selections the probability is reduced and you're going to have increased variance in your results, ie. longer losing runs. A long losing run could wipe out your betting bank even if you're making good bets so be careful there. Another issue is how you pair your selections together as you may end up doubling the wrong bets together or if you tie them all up have more money on some selections than others. That could turn good bets into bad ones.

Like you suggested in your first post your best approach would be to simulate random doubles from all your selections and see what comes up. Simulate a big sample of doubles and if possible post the results up here as I would be very interested in what happens. Most of the time you hear multiples are bad and singles are good but if you can deal with the variance in results maybe this is a valid approach?

On another note it's ridiculous that bookies are refusing £5 and £10 bets off you. Only interested in mug punters these days rather than taking in money on all outcomes and making profit based on their inflated books. 

 

[mick bones]

I have cobbled together some random doubles samples from my singles data set, like every other single paired with the one directly after it and another sample set of each bet with the next bet which is 2 bets down the list.  Not very scientific but gives me two random data samples of approx 10,000 doubles bets.  Yes,  varience plays a big part in the results.  With singles alone I was just shy of 30% profit. However first data set of doubles shows profit of 52% with obviously a far inferior strike rate and I do like a good strike rate as part of any betting system. Avg odds for these new data sets is around 22.0   However the other dataset was the other end of the spectrum with profit of 85% !!   So I figure if I were to generate a couple dozen random data sets the returns would converge towards my original estimate.  But with such % margins the skeptic in me tells me I have forgotten to carry the one or some other silly mathematical error.  Bank is going in the predicted direction so that is main thing anyways

[mick bones]

PS.  I was wrong to say Paddy and Ladbrokes don't take my bets.  Paddy allows me 15p per bet while Ladbrokes will take the vast sum of 23p per bet. Surely you would think that for the purpose of getting their prices in line they would let me have my grain of corn and then tap the guy on the shoulder whos job it is to keep prices in line. But I'm guessing these prices are all taken from 3rd party services now with little or no human interaction so they don't want any value thieves in their customer list.

[arsenalfh]

2 hours ago, mick bones said:

I have cobbled together some random doubles samples from my singles data set, like every other single paired with the one directly after it and another sample set of each bet with the next bet which is 2 bets down the list.  Not very scientific but gives me two random data samples of approx 10,000 doubles bets.  Yes,  varience plays a big part in the results.  With singles alone I was just shy of 30% profit. However first data set of doubles shows profit of 52% with obviously a far inferior strike rate and I do like a good strike rate as part of any betting system. Avg odds for these new data sets is around 22.0   However the other dataset was the other end of the spectrum with profit of 85% !!   So I figure if I were to generate a couple dozen random data sets the returns would converge towards my original estimate.  But with such % margins the skeptic in me tells me I have forgotten to carry the one or some other silly mathematical error.  Bank is going in the predicted direction so that is main thing anyways

Interesting Mick! How would you pair selections together going forward? Would you back them all in doubles? That way you're going to have more money on certain bets which would be an issue.

[arsenalfh]

2 hours ago, mick bones said:

PS.  I was wrong to say Paddy and Ladbrokes don't take my bets.  Paddy allows me 15p per bet while Ladbrokes will take the vast sum of 23p per bet. Surely you would think that for the purpose of getting their prices in line they would let me have my grain of corn and then tap the guy on the shoulder whos job it is to keep prices in line. But I'm guessing these prices are all taken from 3rd party services now with little or no human interaction so they don't want any value thieves in their customer list.

I thought the same but they don't need good punters to adjust their prices anymore as they can just use the exchanges. Easier to use Betfair to set the prices and allow mugs to bet which is a shame.

The minimum liability bookies must offer in NSW Australia seems to have had a positive effect over there so would be nice to see something similar up here.

[mick bones]

I intend doing the ones with max value ( as I calculate it ) as singles and cobble these along ones which have lesser value, but still +ive value in doubles. I don't like stakes which have a return higher than £200 or £300 as that seems to be the glass ceiling, from my experience,  above which you run the risk on being added to the "additional scrutiny" pile.  Using a few friends to get a few quid on is more viable with higher % returns as I have my £ per hr effort target and don't want to waste time farting about if only getting a few quid per hr of effort.  It makes me laugh when I hear of punters going on about getting £100s staked on what ever.  Thats code for I'm making a loss cause its only since I turned the profit corner that I have realised the hardest part of this game is staying under the radar when you have a system in place.  If a book allows you to drop £100 on anything then they have very good reason for allowing that IMHO and its not cause you have been beating the odds for the previous 9 months. 

 

Edited May 29, 2017 by mick bones

[mick bones]

Thanks Paddy .... Ballinrobe,  this evening 21:00,  Princetonian, @ 2.50.  He's allowing me the hansome wager of £4.50.  Brave man.  Just to sicken em I took it and will buy myself a bag of chips with my winnings lol

[DanV89]

@mick bones on a separate note to your doubles discussion, would you be prepared to divulge on your system?!

[mick bones]

You'll probably understand but not going to divulge too much about it. Lot of effort went into the start of it. As old as the hills I'm sure but works for me. In short its based on monitoring the BF exchange prices using a simple Java app I wrote and spotting ponies which are likely to shorten based on monitoring different markets.  The edge is in having found the correct or 'most accurate' data to monitor which indicates a win price is about to move. That particular data/market shall remain a secret,  gives me about 1 min lead time on the exchange to grab a few clicks profit but the real profit is in the bookies side who generally lag the exchanges and have lot of value in their place prices as a result.  9 months later with real money on it is still working but like everything else I'm sure it will fade away with time. Have one other dude on board , sort of as a beta tester, just to help with stats and recording actual bookie prices when required as this is a manual step and getting money on at an actual price is not the same as scraping Joe Bookies site and having your software or odds checker  tell you such and such a price is available. Would never work without software as the time and effort required to manually search for possible horses and monitor the different markets would not be viable time wise