usmobo13
NewbieNewbie
Joined: 14 May 2011
Posts: 2
Thanks received: 1
14 May 2011 - 2:13am

Suspicious "randomness" at Silver Oak and English Harbour

7 replies • Last post

Yesterday I was playing European Roulette and was betting on dozens and red/black to see what's coming out. Before I placed any actual bets the dozens were switching within expected parameters. As soon as I placed a bet the sequence got wacky and did not produce any winning combo for the next 18 rounds and also threw in 3 zeroes almost in a row. I lost almost $140 in under 2 minutes. Same thing happened at English Harbour - I lost $100 under IDENTICAL conditions. It feels like the software will play truly randomly UNTIL you place a bet. Then it will detect your bet and start throwing loosing combos at you. I am programming Artificial Intelligence, so I know for a fact how easy it is to program the software like that. It does not do it all the time, though. It will let you win for a little bit but after certain number of your logon sessions is up - get ready to loose.

auCL-Ed
StaffStaff
CL-Ed's picture
Location: Sydney
Joined: 7 Sep 2007
Posts: 10400
Thanks given: 6286
Thanks received: 5350
15 May 2011 - 7:22am
#1

Hi there mobo13, it sounds like you are steaming. I'll see if I can help you out.

Quote:
Before I placed any actual bets the dozens were switching within expected parameters.

What exactly do you mean by this? I mean, I think I know what you are trying to say - the dozen columns that the ball was landing in were changing on each spin as often as you were expecting? But how often is that and what exactly are your "expected parameters"?

Quote:
the sequence got wacky and did not produce any winning combo for the next 18 rounds

Please forgive me if my assumptions here are wrong, but I have a really bad feeling that you were trying to use a roulette betting system and have found out the hard way that they do not work. There is an inherent house edge in roulette that does not change, no matter how you arrange your bets. There is no such thing as a "wacky sequence" - that is just human interpretation of the numbers. For example, spinning the number 17 five times in a row is exactly as likely as spinning the numbers 32, 3, 0, 16, and 27 in that order. But if five 17s happened in a row a lot of people would think it was almost impossible, while no-one would bat an eyelid at the second set of numbers.

You do not mention your bet size or the system I assume you were using. For example if you were starting with a $10 bet and your bankroll was $140 and you're doubling after every loss then it only takes 4 losses in a row and you are down to your last $20 and unable to recoup your losses in one spin. Even if you started with $1 you'll still reach that point in seven spins. So it is very hard to judge how meaningful the numbers you have quoted are.

Re the 3 zeroes "almost in a row", again its a little vague as to whether you mean 4 spins or 10 or the 18 you mentioned. Here are the probabilities of spinning the same number 3 times during "n" spins:

  • 3 times in 3 spins: 1 in 50,653
  • 3 times in 4 spins: 1 in 13,015
  • 3 times in 5 spins: 1 in 5351
  • 3 times in 8 spins: 1 in 1037
  • 3 times in 10 spins: 1 in 511
  • 3 times in 18 spins: 1 in 94

You mention you're a programmer so I'll give more detail. This is a binomial distribution problem and I am assuming that you are using a European roulette wheel with only 1 zero, and the order of spins does not matter.

If you want to check this yourself in Excel then use this formula where a = appearances, n = number of spins:

p = combin(n,a) * (1/37)^a * (36/37)^(n-a)

So for example, for the same number appearing 3 times in 8 spins you would get:

p = combin(8,3) * (1/37)^3 * (36/37)^5
p = 0.000964022
1/p = 1037

For a comparison of how likely or unlikely this is, Microgaming has a jackpot called Roulette Royale that is paid if the same number is spun 5 times in a row. The odds of that happening for any of the 37 numbers is 1 in 1,874,161 (a single specific number is over 1 in 69 million) but it has happened before as the jackpot has been won (a couple of times before we started tracking it - €977,000+ in November 2008 is one I have found, and it was won again in 2009 too).

There is not much else I can say other than without any detailed data of all your spins and the results, there's no way anyone can tell if the roulette is cheating you, is buggy, or you have just hit a bad run which is to be expected in the long run. In my opinion there is very little need for the casino to cheat you - after all its a high house edge game that cannot be beaten by a betting system. And if they were to cheat you, I would expect them to do something more sophisticated than just favouring the zero. But if you do have a log of your play, then by all means post it here and I'll take a look at it for you and try to work out how likely or unlikely your results were.

2 yobes, barbadosslim93

Always play it safe! Consult our list of rogue casinos and warnings before depositing at a new casino.
Post in our forums to earn CLchips which can be used to buy real prizes in our CLchips shop.

usbarbadosslim93
Forum AngelForum Angel
barbadosslim93's picture
Location: Michigan
Joined: 28 Jan 2011
Posts: 15132
Thanks given: 847
Thanks received: 2154
15 May 2011 - 7:20pm
#2

Good explanation Ed. Hopefully we can get the full details as to what happened on the player's end here, but your initial diagnosis is interesting.

usmobo13
NewbieNewbie
Joined: 14 May 2011
Posts: 2
Thanks received: 1
16 May 2011 - 7:30am
#3

Thanks for your explanation Ed...I am working on getting my logs together. I don't have the log from the "bad run" but I am will try to produce another one. I play European Roulette and start with $1 bet. I don't question the random numbers, I just have the feeling that the "randomness" is changing as soon as the bet is placed. I am doing 100 spin runs. My expected parameters - start betting after 6 spins if only two out of three dozens came out (min 5 spins - max 13 spins). In my test run the long sequence happens once, maybe twice, most of the time it's 5-7 spins before I start betting. Now, that applies to "fun money" and real money (betting only on red+black). As soon as I bet with real money on the "coming" dozen this changes - in the "bad run game" I placed my bet after 7 spins and nothing came out for the next 10 or 11 spins. It never happened in fun mode or (black+red bets). I've ran hundreds of spins to test this thing and the longest loosing round was only 14 spins. I just find it "convenient" for a casino to extend that number as soon as there are money on the table. As for zeroes - I had them coming out 2 times, 4 times or 6 times in a 100 spins runs. Then again, in my "bad run" they came out 4 times AFTER I placed the bet. I think I'll try to record longer runs - maybe 300 spins or so to see if that makes any difference.

CL-Ed

auCL-Ed
StaffStaff
CL-Ed's picture
Location: Sydney
Joined: 7 Sep 2007
Posts: 10400
Thanks given: 6286
Thanks received: 5350
16 May 2011 - 11:28pm
#4
mobo13 wrote:
My expected parameters - start betting after 6 spins if only two out of three dozens came out (min 5 spins - max 13 spins). In my test run the long sequence happens once, maybe twice, most of the time it's 5-7 spins before I start betting.

Ok, what you are doing here doesn't make you any more likely to win because the probability of things that we know have happened in the past is always 1 or 100%.

For example, if you were to predict that the next 10 spins in a row would be red, the chance of that happening would be (18/37)^10 = 0.000742519 or 1 in 1347. But if you decide to jump in after the wheel has already spun 9 reds in a row, the chance of the next spin being red is 18/37 - just like it always is.

The other thing to remember is that although the chance of spinning 10 reds or 10 blacks in a row might be 1 in 1347, the chances of any other specific set of 10 red or black results is also 1 in 1347. This is because there are 1347 possible combinations, so 10 reds or blacks are no more or less likely than any other red/black combination. Its only our human propensity for pattern recognition that interprets such a result as extraordinary.

Past random events (like a roulette wheel spin) do not make future random events more or less likely to occur. The belief that they do is known as the gambler's fallacy. This does not always apply when gambling. For example when you are playing blackjack and a ten is dealt, this makes it less likely that another ten will be dealt next as the cards are dealt from a deck and are not replaced. This means the events are related. Roulette spins, however, are not.

Here are some specific numbers related to what you are doing:

Chance of the same 2 specific dozens (24 numbers) being hit on 6 consecutive spins:

p = (24/37)^6
p = .074482991
1/p = 13.43

So you should expect that to happen around 7.5% of the time. But once it has happened and then you start betting, the chance of it occurring was 100%. The chance of the same 2 dozens being hit again on the next spin are now 24/37, just as always.

You can't factor in the past results when predicting the next spin, unless you have reason to believe the results are biased - eg an unbalanced real-world roulette wheel. In such a case you would bet with the series of results, not against it. Please don't interpret that as a recommendation to change your strategy to betting with the 2 dozens, it will not make a difference online!

Seeing 2 (25% chance), 4 (15%), or 6 (3.5%) zeroes (or any other number) in a hundred spin series is not extraordinary. And remember, if the casino wanted to cheat you they would not need to pick the zero all the time.

I recommend you try a lot more practice mode games at a few different software providers (we have a few here: Free Casino Games) and you will see that if you are relying on specific sequences to win at roulette, you will always get hit with a run that busts you sooner or later. It is guaranteed. If the game could be beaten that easily, casinos would not be offering it!

yobes

Always play it safe! Consult our list of rogue casinos and warnings before depositing at a new casino.
Post in our forums to earn CLchips which can be used to buy real prizes in our CLchips shop.

aukabean
GamblerGambler
kabean's picture
Joined: 14 Oct 2009
Posts: 79
Thanks given: 9
Thanks received: 8
19 May 2011 - 11:18pm
#5

its like a maths lesson in here! Ed you come across as a genius...

all i took from this is that you can't beat roulette with a system? is that right?

auCL-Ed
StaffStaff
CL-Ed's picture
Location: Sydney
Joined: 7 Sep 2007
Posts: 10400
Thanks given: 6286
Thanks received: 5350
20 May 2011 - 4:29am
#6

If that is all you took from it then I am happy enough because that's the most important thing!

(not a genius either)

Always play it safe! Consult our list of rogue casinos and warnings before depositing at a new casino.
Post in our forums to earn CLchips which can be used to buy real prizes in our CLchips shop.

nzyobes
GamblerGambler
Joined: 26 Nov 2007
Posts: 97
Thanks given: 9
Thanks received: 11
26 May 2011 - 6:48am
#7

The old gambler's fallacy, it gets them every time. I am suprised an "artificial intelligence programmer" would fall for it though (no offence intended mobo13).