How much cash back would make a "value bet" ?

[Human123]

In light of the recent willhill vegas 50% cashback offer which could be used on roulette..

How much cashback is enough to make the bet a value bet?

[Landprofits]

Depends how much value you want sir. Depending on the version of roulette (which makes a neither here nor there differnece to the percentage really), your chances of success on a single bet on either red or black are 47.3%. The odds offered are evens.

BUT, if you are geting cashback on losses, you are only risking the amount of money you actually lose. To work out the effective odds you're getting, divide the potential returns by the amount risked.

For instance, if it were just a straight bet with no cashback, and you bet £100, you are risking £100 to return £200. 200/100 is 2.0 decimal odds, which is evens. That implies a 50% chance of success. But since the actual chance of success is 47.3%, you have a 2.7% disadvantage.

If you are eligible for 10% cashback, you are risking £90 to return £200. 200/90 = 2.22 Decimal odds. Expressed as a percentage that is 100/2.22 = 45.04%. So now you are getting paid 47.3% to take on a 45.04% chance of success. Thus, you have value. But not a lot!

There is no hard and fast answer to this because it is all down to the risk profile of the individual. You have to decide for yourself whether it is "worth it" or not. I wouldn't advise taking on such a small edge though, certainly not with only 2 bets to do it in.

If it were 20% cashback, then yes, it's probably worth a punt but only if you are actively looking for long term value. Have a read of Zenmaster's thread about Casino variance to see how variance can affect the player. If you're only going to make a couple of value bets, you're at the whim of negative variance if it kicks in.

I'm happy to make value bets, because i've built up a float that can absorb any negative variance, and i'm in this for the long haul. It all depends on how much risk you wish to expose yourself to. You may wish to consider this carefully before you pile in.

[dulence]

I'd just correct you if I may about the chances when playing on red/black. There are 18 blacks and just as many reds, so putting your money on either of them gives you 18/37 = 48.6% chance of winning.
As for the most valuable way to play it would be to select a game with a combination of maximum variance and minimum house edge, so putting all your money on a number at roulette would be a good long term strategy. For the sake of simplicity, imagine that you were doing 37 separate promotions, all of which are doing the 50% cashback. Say you bet a tenner on a number at each one of them. You'll lose 36 times (-360), only getting the cashback which would be 36x5 = +180, and you'll win one time getting paid +350, all in all leaving you with a +170 profit, or 170/37 = 4.59 profit per spin. Now, if you were to go for black/red, you'd win 18/37 times (+180) and you'll lose the rest (-190) but get 50% cashback on those (+85), leaving you with a profit of 75, or 2.02 per spin. All this assumed you'd be betting your whole balance each time.

[nixonite]

Originally Posted by dulence

I'd just correct you if I may about the chances when playing on red/black. There are 18 blacks and just as many reds, so putting your money on either of them gives you 18/37 = 48.6% chance of winning.
As for the most valuable way to play it would be to select a game with a combination of maximum variance and minimum house edge, so putting all your money on a number at roulette would be a good long term strategy. For the sake of simplicity, imagine that you were doing 37 separate promotions, all of which are doing the 50% cashback. Say you bet a tenner on a number at each one of them. You'll lose 36 times (-360), only getting the cashback which would be 36x5 = +180, and you'll win one time getting paid +350, all in all leaving you with a +170 profit, or 170/37 = 4.59 profit per spin. Now, if you were to go for black/red, you'd win 18/37 times (+180) and you'll lose the rest (-190) but get 50% cashback on those (+85), leaving you with a profit of 75, or 2.02 per spin. All this assumed you'd be betting your whole balance each time.

And assuming variance isn't a bitch....

[Human123]

Nice posts^^

Am i right in thinking the correct odds of red/black should then be 2.10?

If so, then getting even 10% cashback should make this a valuebet? or at least getting the "true" odds...

[Landprofits]

I'd just correct you if I may about the chances when playing on red/black. There are 18 blacks and just as many reds, so putting your money on either of them gives you 18/37 = 48.6% chance of winning.

Yes you may. To be honest I pulled that figure of 47.3% out of my head from memory without working it out, and my figure of course comes from American roulette where there is the double zero, thus; 18/38 = 47.37%.

As for the most valuable way to play it would be to select a game with a combination of maximum variance and minimum house edge, so putting all your money on a number at roulette would be a good long term strategy. For the sake of simplicity, imagine that you were doing 37 separate promotions, all of which are doing the 50% cashback. Say you bet a tenner on a number at each one of them. You'll lose 36 times (-360), only getting the cashback which would be 36x5 = +180, and you'll win one time getting paid +350, all in all leaving you with a +170 profit, or 170/37 = 4.59 profit per spin. Now, if you were to go for black/red, you'd win 18/37 times (+180) and you'll lose the rest (-190) but get 50% cashback on those (+85), leaving you with a profit of 75, or 2.02 per spin. All this assumed you'd be betting your whole balance each time.

Yep, I absolutely agree with your math here, but of course you're exposing yourself to variance and also following this strategy would need a strong psychological approach, in other words you would need to keep your nerve whilst you were waiting for that 35/1 payout.

Of course, on the particular offer we are talking about we only have a couple of bets really, so personally I would opt for the lower variance startegy which has the best chance of extracting profit out of this offer.

Now, of course, this is where the debate could start, with a value bettor saying that I would be following a false economy here. As I pointed out above though, there are other external factors other than the math, such as the pychological impact on you. However, the main thing that sticks out in my mind is this;

If you did follow the High Variance strategy, you would make more profit in the long run from whatever value bets you were doing. That is a mathematical fact. However, it's fairly safe to say that you're probably going to go through a period where you're down a substantial amount before you get a payout. Whilst you are in a loss position from a strategy, that money is not availabe for you to use elsewhere. On the other hand, if you follow the less profitable but lower variance strategy, you can get a ROI on your funds by arbing them, matched betting or whatever.

I'm not disagreeing with Dulence, he is absolutely correct and when probability is worked out properly the results are often suprising. But what I am saying is that there are other factors to consider.

By the way I put £50 on red last night and it won.. I don't know why I chose red.......isn't black paint heavier than red paint?

[Landprofits]

Originally Posted by Human123

Nice posts^^

Am i right in thinking the correct odds of red/black should then be 2.10?

If so, then getting even 10% cashback should make this a valuebet? or at least getting the "true" odds...

18/37 = 48.6%.

100/48.6 = 2.06 (ish) your true odds are. So yeah, at 2.1 isn't far out.

With 10% cashback, say if you bet £10, you would be risking £9 to return £20. So the effective odds you would be getting are 2.22.

The effective odds you are being offered exceed your true chances of winning, so yes, you have value, but not enough for me to want to take that bet. Not unless I knew I could get that value over a long haul.

I'm calculating your edge though at 3.194%, which would be a luxury for a card counter! Don't forget though that card counter would play 1000's of hands.....we only have a couple of bets!

[Human123]

very interesting.

you say not enough to make you take that bet, so going back to the original question, how much would be enough for you to take it?

Im pretty sure i was reading somewhere that if you can get above the true odds of something you should ALWAYS take the bet as you will win in the long run... so surely it should mean the same for roulette if your doing it somewhere where you get 10%+ cashback?

[Andy]

Originally Posted by Human123

Im pretty sure i was reading somewhere that if you can get above the true odds of something you should ALWAYS take the bet as you will win in the long run

It still depends on the chances of it happening and the cost of entry as to whether it is right for you to bet on it. Just because you will always win in the long run doesnt mean you will get there, in the long run we are all dead.

The odds of winning the National Lottery are 1 in 13,983,816. If the Jackpot was £14m then buying a ticket is a value bet. Im stil not sure I would buy a ticket expecting to make money from it, it is so likely I will lose £1 I would rather keep the £1 and forgoe the EV.

Coming from another angle somebody may offer me odds of 2.2 on the toss of a coin, which is a 50/50 bet (Odds of 2). So that is value there and odds above true odds. If it cost £1 to play I would all day long without thinking, £10 I would play it, I would probably play at £100 a time. If the guy said I could play that game at those odds but I HAD to play at £10,000 a throw, I would not play. It is the same +EV bet, but I could not afford to lose £10k, so again I would have to forgoe the EV. The first rule of gambling club is dont gamble with more than you can afford to lose.

You get the idea, just because something is +EV that means you should always take it given an infinite bankroll and an infinite length of time. If you dont have those, there are times and reasons to turn +EV offers down. I think so anyway.

It is true that if the factors are all right for you then you should do +EV offers because in the long run you cant lose. You may mean this thread and it is pretty much shown that you will never get considerably over or under the EV line:
Tracking my Variance - Is Andy right?

[Human123]

Some very interesting thoughts.

I agree with you andy, reading one gambling forum recently someone worked out that leeds were at good "value" odds to beat man united.

But would any sane person really want to put money on something like that?


(Of course, the result came in but thats not the point im getting at lol)

[Landprofits]

Originally Posted by Andy

If the guy said I could play that game at those odds but I HAD to play at £10,000 a throw, I would not play. It is the same +EV bet, but I could not afford to lose £10k, so again I would have to forgoe the EV. The first rule of gambling club is dont gamble with more than you can afford to lose.

Hence why I would not take the bet.....

You get the idea, just because something is +EV that means you should always take it given an infinite bankroll and an infinite length of time. If you dont have those, there are times and reasons to turn +EV offers down. I think so anyway.

Exactly. And in this case we only have two bets, and if you're only playing with about 3% edge, you haven't got long for the odds too turn in your favour!

Originally Posted by Human123

very interesting.

you say not enough to make you take that bet, so going back to the original question, how much would be enough for you to take it?

At 50% cashback, I can't really ignore it, because there is so much value in that. Had I got the 20% offer, I probably wouldn't have bothered but that attitude may change as I bring more casino type offers into my activity.

[Carson]

Used to do this up to a £100 a time when sporting index ran their double profits offers on games. Was a good earner but they've not done it for months now

[advanta]

Originally Posted by Andy

It still depends on the chances of it happening and the cost of entry as to whether it is right for you to bet on it. Just because you will always win in the long run doesnt mean you will get there, in the long run we are all dead.

The odds of winning the National Lottery are 1 in 13,983,816. If the Jackpot was £14m then buying a ticket is a value bet. Im stil not sure I would buy a ticket expecting to make money from it, it is so likely I will lose £1 I would rather keep the £1 and forgoe the EV.

Coming from another angle somebody may offer me odds of 2.2 on the toss of a coin, which is a 50/50 bet (Odds of 2). So that is value there and odds above true odds. If it cost £1 to play I would all day long without thinking, £10 I would play it, I would probably play at £100 a time. If the guy said I could play that game at those odds but I HAD to play at £10,000 a throw, I would not play. It is the same +EV bet, but I could not afford to lose £10k, so again I would have to forgoe the EV. The first rule of gambling club is dont gamble with more than you can afford to lose.

You get the idea, just because something is +EV that means you should always take it given an infinite bankroll and an infinite length of time. If you dont have those, there are times and reasons to turn +EV offers down. I think so anyway.

It is true that if the factors are all right for you then you should do +EV offers because in the long run you cant lose. You may mean this thread and it is pretty much shown that you will never get considerably over or under the EV line:
Tracking my Variance - Is Andy right?

The concept of risk and return has been much discussed in gambling literature. For those who are interested it is worth googling the term "Kelly criterion".
Essentially Kelly bettors bet the % of their bankroll relative to their edge. So, if a rebate gives you a 10% edge (which would be roughly a 20% rebate for a one-big-bet approach), you'd bet 10% of your bankroll. Even professional gamblers don't have the balls for pure Kelly and usually use a Kelly fraction (say, a quarter) to make the bad streaks psychologically bearable.

Although I agree with Andy in general, you'd be surprised how aggressive your bet sizing can be with advantageous situations like rebates whilst maintaining an acceptably low level of going broke.

[Grandthrax]

Very interesting...

[TomYom]

Originally Posted by Andy

The odds of winning the National Lottery are 1 in 13,983,816. If the Jackpot was £14m then buying a ticket is a value bet. Im stil not sure I would buy a ticket expecting to make money from it, it is so likely I will lose £1 I would rather keep the £1 and forgoe the EV.

assuming that there can only be one winner?