In my experience, the only occassions where I've found overlaying to be of any use is when you receive a bonus if you get a winner. For instance 365's 4+1 and when ladbrokes had a similar offer that ended during cheltenham. It was especially useful in the laddies offer, as if you had mulitple runners and none of them won you at least won your overlays. However if you did have a winner, you got a matching snr, and your other overlays compensated your overlay on your winner.
The bonus in underlaying at high odds is that it doesn't cost you that much to get a potentially really good bonus.
Originally Posted by pinkyexcel
For every pound you don't lay you will get that pound x whatever the high odds are as extra winnings if the bet does win.
(not sure if the above is exactly correct, but it is something like that).
I use ultimatcher which is great for me personally but again it's personal preference. There will be a link for it on here i'm sure if you pop ultimatcher into the search function.
Do you use a spreadsheet to work out what you would be comfortable laying? The spreadsheet I'm using is just a basic mb one which isn't happy to allow you to enter a different lay than the one suggested for the same gain/loss either way.
Yup, it's been a while since I did any of those kinds of offers but certainly on "free bet if win" offers then overlaying produces profit if book bet loses and if it wins then you lose more from the exchange but are rewarded with a free bet. Beyond that, my concern with those is if the book decides not to give the free bet but that's just a general risk anyway.
Originally Posted by hippocrates
Nice post, Andy... those two are often overlooked points. Also, with your description of the 2-2 scenario, it reminded me of the particular 2-2 refund offer from Betchronicle where you could often either match everything fully before kickoff and take £5-£10 profit max normally or you could underlay to break even but potentially be able to make £100-£200 or lock in £50-£100 nearer the end of the game if choosing to lay off.
Originally Posted by Andy
this is an interesting thread, which has really got me thinking, as i have rarely underlaid before - the only time being the bet365 4/1 horses offer as mentioned above.
at first i thought that the arb has to be fairly decent to make underlaying worthwhile, so as to make the percentiles work in your favour over the medium - long run. however having looked into it it's now occured to me that, without even considering any externalities such as how much you value your time placing the bets, risk associated with running up balances at bookies etc, underlaying should be pursued (mathmatically speaking) almost all of the time.
assume that you back crystal palace to win the championship at 10.0, and that you have the oppurtunity to lay at 9.5. this can be a win/win, straightforward matched arb with a profit of 0.53% (assuming 5% commission - obviously the profit % will be higher the lower your commission rate). so, with a back of £100, you can lay it off at £105.52 for a profit of £0.53, whether palace win the league or not.
alternatively, i can underlay so that i equal my stake lost if palace do not win the league. at 5% commission, the lay would need to be £105.26 (as £105.26 x the commision rate = £100). if palace do win the league however, i make it that i am up by £5.26.
so the question becomes, do i take the guaranteed £0.53, or risk getting nothing for the longer shot of the event selected winning and getting £5.26? i believe this is answered very simply by the odds offered on the event. in actual fact in the example above, the market dictates that palace have a 9/1 shot of league glory, i.e. using the odds as a guide to probability, if the league were played out in full over 9 different season scenarios, palace would win the league once, or 11.1 times over 100 different season scenarios.
as a logical extension of this, if i placed this exact bet 9 times, i could expect palace to win it once, and lose the rest. therefore, if i continually opted to underlay, i would expect to break even 8 times, and benefit from the event result going my way once. in total i would make (£0.00 x 8) + (£5.26 x 1) = £5.26. conversely if i wanted guaranteed profit and did not underlay, i would receive £0.53 x 9 = £4.77.
what i find interesting about this is that it the benefit of underlaying is negatively related to the size of the arb, i.e. the smaller the arb, the better it is to underlay. if you get a really juicy arb, say back 10.0, lay 8.5, it actually makes more sense to keep arbing for guaranteed profit as you would generate (£12.43 x 9) = 111.87 as opposed to (£0.00 x 8) + (£110.53) = 110.53 if you persist in underlaying. i think this should be able to be incorporated into a spreadsheet, but i'm crap with excel!
at least, this is how it appears to me.
Interesting, but don't agree.
Originally Posted by tasty_snacks
In your calculations, you have used the bookie odds as the 'fair price' wheras arbing tends to assume that the bookie has a value price. Although it's not as straight forward as that (as true fair price is difficult to determine), I think the exchange tends to be a better guide to what fair price is. For instance, in your 10.0/8.5 arb, I would reckon on somewhere between 8 and 8.5 (back and lay price) being fair odds
The assumption that the bookie is fair price in your last example, has led to the conclusion that the exchange must be value in the lay, and so it follows that betting a relatively larger amount on the value odds has a higher expected value. This is why it appears that underlaying is less ev. In fact, if this were the case, overlaying would be the best option as you would expect to make £14 profit 8 times, 0 profit once, and be £112.00 up rather than £111.87.
In fact, if the fair odds are (for instance) 8.25, you would expect to win 4 and lose 29 in every 33 races. The overall profits from this would be:
Overlay (4 * 0) + (29 * 14) = £406.00
Straight lay: 33 * 12.43 = £410.19
Underlay: (4 * 110.53) + (29 * 0) = £442.12
demonstrating that underlaying is the most ev method out of these three.
In fact, even more ev is not laying at all. Then you would have
No lay: (4 * 900) - (29 * 100) = 3600 - 2900 = £700.00 - a massive 70% on top of straightforward laying.
The benefit in laying comes with reduced variance and very low risk of ruin. If you're sure you won't lose money in an individual bet, you can place a higher stake on it safely, which should give a higher profit in real terms, which is in certain respects, more important than percentage profit.
As to putting it on a spreadsheet, I do have my spreadsheet to automatically calculate lay stakes which will break even at the bookie and at the exchange, depending on which is preferable.
If you have the original matcher spreadsheet from MSE, the lay stake field can be overwritten to
Which will break even at the book
[(Back Return) - (Back Stake)]/(Lay Odds - 1)
Which will break even at the exchange
(Back Stake)/(1 - (Lay Commission/100))
[note: neither of these have been tested with commission on the back, so use at your own risk]
Thanks Tom; I'd been trying to add the min and max lays to my spreadsheet and hadn't managed to get the right formulae, so this is very helpful.
Have to say at first I went with the MSE theory of equalizing profit, but now I do underlay to some degree depending on the situation, although not sticking to any hard and fast rules. Probably not maximising profit either
I think it's been said plenty of times on here before in various forms, but if you're taking value bets or arbs regularly, then the chances of you being in profit in the long term IF you didn't lay anything are certainly VERY positive.
This is a similar kind of scenario as with any 'regular' investment though - it's always nice to look at fancy charts that show the huge gains you would have made whilst invested in the stock market over a period of at least 5-10 years, but when you're just starting out it's always daunting and to some extent requires a leap of faith. Think of the small print on any investment product along the lines of 'past returns are no guarantee of future returns' etc.
The recent events on the global economic markets represent to some extent the kind of 'variance' that you can expect over a long period with any kind of investment and in a similar vein gambling has the same characteristics. If you choose good quality 'stock' to invest in (ie value bets), over time they will provide a return that is positive - the expected value inherent in those bets is positive. However in the shorter term you can expect a varying degree of variance depending largely on the prices at which you take your bets - obviously higher bets, even if they are 'value' will take some time to 'pay off' - but the key factor is that if you do consistently take value bets then because of their very nature as being 'good value' you will see a profit in the long term.
Personally I still don't have enough faith in statistics to go 'all in' with a bankroll and rely on the fact that I'm taking 'value' bets in an attempt to see a profit in the long term. It's really something that I'd love to pursue and have meant to do for a long time now.
As such I find the various 'tipping' threads that tod and the like post very interesting, and I really should make more of an effort to put a roll aside dedicated to just 'value' betting, just to see how big I can make the roll after a given period (cue another 'challenge' ala the trading challenge!!!). Having said that though I always swore that the next promotion I did on here would be a spot the ball comp.... so don't hold your breath on that one.
i see, so when we assume the exchange to be an efficient market (god, let's not go there ) hence reflecting fair price and thus true chances of probability, the benefit of underlaying increases exponentially until we reach the point of not laying at all.
Originally Posted by tom
i would still not be confident in playing the value bets game consistently, at least not with my own cash. i think munk is spot on, it's a preference choice that needs to be looked at in the same manner as a financial instrument. by laying, i'm effectively writing a call option that minimises my risk with the underlying investment (the back).
Tags for this Thread