A case for betting fewer numbers

[Scarface]

While fighting negative variance, betting fewer numbers is a better option.  Let's compare betting even bets (18 numbers) to betting dozens (12 numbers).

Look at even bets first.   I'll leave zeros out of the equation to make it more simple.  To bet a total of 36 numbers, that is betting an even bet 2 times in a row, since there are 18 numbers in even bets.  Let's say our target is Red.  What are the odds that in 2 bets we will beat expectation?  First, look at the possibilities in 2 bets....they are RR, RB, BB, BR.  There is only 1 option out of the 4 that wins, which is RR.  So, the probability of winning is 25% (1/4).

So, if our total wager for an even cycle is $6.  That would be 2 bets at $3 a piece.  If RR hits, our total profit would be $6.

Now, let's look at a 12 number bet.  The cycle for a dozen bet is 3.  We need to hit more than once to beat expectation.  Let's say our target is dozen 1.  What are the odds that dozen 1 will hit more than once?  First, you have to look at all the combinations dozens can make in 3 spins....without listing them all, I'll go ahead and tell you there are 27 different combinations.  So how many of these combinations will show dozen 1 winning in 3 spins?  There are 7 winners.  They are 111, 112, 113, 211, 311, 121, 131.  So 7 winners out of 27 gives a percentage of 25.92%

Both the dozen cycle and even cycle bets a total of 36 numbers.  But to make it a true comparison, the total wagers also have to be the same.  On even bets, our total wagers was $6, so it needs to be $6 for dozens as well.  Since there are 3 bets in a dozen cycle, each individual bet will bet $2 a piece.

All of the 7 winning dozen cycles will profit $6 each...except 1.  The cycle 111 will profit $12.   So, when you average the winners out you have an average win of $6.85 (6+6+6+6+6+6+12) / (7).

Final Results between even bets and dozens.  In both cases, we bet a cycle for 36 numbers.  In both cases, our total wager for the cycle was $6.  For even bets (18 numbers), chance to hit above expectation is 25% with a profit of $6.  For dozens (12 numbers), chance of a hit above expectation was 25.92%, with an average profit of $6.85.  Very easy to see which is the best option.  Playing fewer numbers gave us better results, not only on hit rate but payout too.  My personal preference is to play only 3 to 4 numbers.

[Blood Angel]

Great post, thank you.

[Mako]

Nice scarface, well written. 

[Bigbroben]

Consider the ratio risk/reward when betting on different number of numbers.

Playing 1 number: risking 1 to gain 35 (yes, 1u reimbursed and 35u won, so 36 total).  Probs of losing: 36 in 37.
So, with the lowest possible variance, losing 36 times 1u and winning 35 times played units.
Ratio of 0,027 (36/37 x 35/1).

Playing 35 numbers: risking 35 to gain 1.  Probs of losing: 2 in 37.  So 1/18.5 chances of losing to win 1/35th of played units.  Ratio of 1.89 (2/37 x 1/35).

Playing 18 nrs: risking 18u to gain 18u, so, doubling up, or + 1x bankroll.  Probs of losing: 19/37. Ratio of 0.51 (19/37 x 18/18)


My feeling: the more played numbers, the more severe is the house edge.  The proportions are against the player as the amount of played nrs increases.

[Gandhi]

  My personal preference is to play only 3 to 4 numbers.

Hey Scarface, just curious how do you like to play 3 to 4 numbers? Any special way you feel is best?

[Scarface]

My feeling: the more played numbers, the more severe is the house edge.  The proportions are against the player as the amount of played nrs increases.

Yes, agree.  Nothing wrong with starting out playing many numbers.  Just not indefinitely.  Alot better to reduce amount of numbers played, than to drastically raise wagers covering more numbers.

Some will say the house edge, or unfair payout, is the same for all bets.  Yes, this is technically true.  But, that doesn't mean all bets are equal.  As I shown above, you can get almost a 1% better hit rate as well as better payout betting dozens vs evens.  Mathematical evidence shows that comparing 2 different bets, there better options....not all are equally bad 

[Scarface]

Hey Scarface, just curious how do you like to play 3 to 4 numbers? Any special way you feel is best?

Can't say what's best.  But my preference is to play the most recent repeaters.  Pretty easy, with no tracking.  Just look at the marquee, pick last 3 numbers that repeated, and play until hit or 12 spins.  If no hit in 12 spins, change number to latest 12 spins again, and repeat. 

Sometimes, if a super cold number just landed it, I might add it too

I usually have success with this at a B&M casino.  Always seems to be that 1 or 2 numbers that just stay hot. 

Of course, if somethings not working that day I'll switch over to another strategy.  But it works pretty well most of time 

[Joe]

Nice analysis. 

[Tinsoldiers]

This is good analysis scarface, but only partial.  We don’t lose that 6$ wager in all combinations. For your wagers on red, there is a 50% chance that you will get back your wager and 25% chance that your wager is gone. But if you take dozens, there is more than 25% chance that the wager is gone completely.  Take that as well into consideration on the time when you lose your wager completely and you might get a different picture. 

I don’t have an analysis like you, but I think if there is a winning system it is directly proportitional to the more numbers you are wagering. 

[nottophammer]

I don’t have an analysis like you, but I think if there is a winning system it is directly proportitional to the more numbers you are wagering. 

Well said Tins

[Scarface]

Thanks for the input.  And I'm not saying that playing more numbers is always wrong.  Playing more numbers is great for hit and run.  Also great for grinding out small steady profits, as long as you're winning.

I'm thinking more about when variance starts to turn against you, and losses come.  In this case, I believe less numbers are better to deal with variance.

Let's say 2 players are both down 36 units each.  Player 1 has been playing even bets and decides to up his numbers to double dozens.  If he bets $36 on double dozens and wins 2 bets, he breaks even.  If he losses both bets, hes now down 108 units.  Then what?  Triple the wager?  Bet 30 numbers? 

Ok player 2 is also down 36 units.  He decides to play only $1 unit on 3 numbers  for 12 spins.  If he gets only 2 hits, he back in profit.  Even if he gets 0 hits in 12 spins, he can still remain at $1 bets for another 12 spins...he'll just need to get 3 hits to win.

Look at the 2 bets, and imagine how this will play out if both players hit some negative variance.  The way player 1 is betting more numbers, wagers will have to go alot higher than player 2.  Even though both players start out with a -$36 balance, in just a short amount of spins player 1 wagers could be approaching the triple digits while player 2 is still wagering in the single digits. 

[Gandhi]

Can't say what's best.  But my preference is to play the most recent repeaters.  Pretty easy, with no tracking.  Just look at the marquee, pick last 3 numbers that repeated, and play until hit or 12 spins.  If no hit in 12 spins, change number to latest 12 spins again, and repeat. 

Sometimes, if a super cold number just landed it, I might add it too

I usually have success with this at a B&M casino.  Always seems to be that 1 or 2 numbers that just stay hot. 

Of course, if somethings not working that day I'll switch over to another strategy.  But it works pretty well most of time 

Awesome, I like it, nice and simple, thanks for sharing scarface.

[Steve]

Generally in the long run, betting fewer numbers is better. But in the short run, more numbers is better. It does depend on your method though. You cant say conclusive which is best without considering the method.

[Bebediktus3]

You cant say conclusive which is best without considering the method.

Maybe you that simply not know. But that not means that we can't calculate.
Realy on every spin, every number has its expectation to be hit. So the best solution is to cover - all numbers with positive expectation.
Here are only one small nuance - usually we not know enough exact which numbers are positive - which no. Now say accordingly some experiments we know, that usually we have 15 positive numbers - that again not give us nothing worth because first, we must know, or they are distributed in one spot or in two, or other.
Say are some ideal wheel and all numbers are in one zone and say we from somewhere know that are 15 positive numbers - we still not know where is the center of that zone - or calculations can say it is number 17, but can be say +/-4 pockets.
In such way, we must cover 15-4-4=7 pockets. This way we minimize the ability to cover negative numbers. In game covering too many numbers with negative expectation is worst what we can do....

[Steve]

If youre talking ap, then betting every positive expectation number gives you highest profit per hour. But that may also make your ap transparent to surveillance.

Generally if you are betting sectors, betting fewer numbers means higher edge, but lower profit per hour, and higher chance of having lots of near wins, but too few actual wins.

Again you cant say which is best without considering the method.

If youre talking typical system play, then it doesnt make much difference. More numbers can mean more stable results although still loss. Ie more numbers means smaller wins but with possibility of a bigger loss when numbers dont hit. The reverse is true for fewer numbers but then you can go a long time without a win. It is more a personal preference than a statistical advantage.