Yep, more gamblers fallacy for you. GOOD TIMES!

[Bayes]

An oldie but a goodie but I have to post it again. I love the different answers.

Gamblers fallacy will never help you? Hmmm

I'll pick 3 numbers for you and 3 numbers for myself, all flat betting.

I have tracked the last 600 spins. The 3 numbers I choose for you, are the 3 with the fewest hits. (by the way, there is no darn bias in the wheel)

For myself, whatever 3 numbers have three hits on it, most recently.....in the last 'X' number of spins. In other words, presently hot.

We keep betting until 2 of our 3 numbers gets a hit, that's the winner. If we did this experiment 500 times, it SHOULD come close to you winning 250 times and I win 250 times, correct?

(Close to 250, I'm not saying it would be perfect)

Who would say 'YES Ken', it would be around 250 and 250 over and over again?

Ken

Ken, can you give me a number for X? I'd like to do some tests on this.

[MrJ]

It would depend.....meaning, which ever are the most CURRENT last 3 numbers with three hits on it. The third number of the group could go as far back as 47 spins (thats the extreme end).

Ken

[MrJ]

A: No.

B: Depends. I think if you expect a hit in less than 36 spins, you're trying to be lucky. And past 36 spins, you're betting because it's "due to happen".

At least that's what maths and the gamblers "fallacy" is based upon...

It's just theory, I've yet to see a working system based on one or the other. people also like to name it positive progression & negative progression, from your expectation of making a profit in a given time. What it really is is flat betting a number..

I dunno. Its all the same In My Humble Opinion.

Different words. Different thinking.

You bring up a good point, I could do another thread just on this.

NEW question: We know if you win on the 36th try, you break even. We're gonna bet on that #17 (pulled the number from a hat). Let say one guy is flat betting but only up to 36 attempts, then stop. Another guy flat bets until the 35th attempt but then adds one unit on for the 36th and last attempt, then stop. Is one guy using gamblers fallacy and the other not? Are they both using it or are neither of the guys using gamblers fallacy? (lol) This s**t drives me crazy.

Ken

[nitrix]

Ken I'm glad to know I'm not alone! I usually do the braistorming alone though.

I say nobody wants to stop at the break-even point, don't we all want to make a profit? So I think the legetimate way to do it is by adding an unit one spin before the break-even point. Thus 35th in our case.

But is it gambler fallacy? Every system is based on an observation... it's your call. I prefer terms like "calculated risk". Sometimes you can risk less or more money depending on your bankroll.

The way I play at the moment; I take a simple system like "Tier et tout", and if it loses more than the break even point, a start over and multiply my bets by a number... say fibonnaci, marty, any neg. progression.

It's like.. a system playing sessions of another system. Inception much.

[Bayes]

It would depend.....meaning, which ever are the most CURRENT last 3 numbers with three hits on it. The third number of the group could go as far back as 47 spins (that's the extreme end).
Ken

So it has to be 3 numbers with EXACTLY 3  hits up to a max of 47 spins back, right?

It may happen that the 2 numbers which hit belong to both groups, in which case I would call it a draw. 

[Bayes]

Ok, I've done a simulation.

Get 600 spins, and pick the 3 numbers which have hit the least number of times, then pick 3 numbers which have hit exactly 3 times going back no more than 47 spins. Continue to get spins, and the winner from the 2 groups of 3 numbers is that group which is the first to get 2 hits.

Here are some results using 500 sessions at a time (note that the numbers in each pair don't add up to 500 because in 20%+ of sessions there were less than 3 numbers in the last 47 which hit exactly 3 times. If the there were MORE than 3 numbers which hit 3 times, 3 of them were selected randomly):

Hot   Cold

183, 181
189, 198
175, 187
184, 192
193, 166
178, 197
189, 172
196, 189
208, 166
170, 203

The numbers in red are those sessions in which the hottest numbers hit twice first. So it seems to be a draw. However, I noticed that there is a slightly larger gap between the pair when the hot numbers won, so I ran a longer test of 100,000 sessions. The result was:

Hot          Cold
38,192    36,340

A percentage difference of around 5%.

I was surprised at this, so I ran a few more tests and got similar results each time. So on the basis of this test, there does seem to be some support for the claim that hot numbers are a better bet.

I used RNG for this, I'll run another test using actuals and report back. I only have 1M actuals so it won't be so conclusive, but interesting to see nevertheless.

[Bayes]

NEW question: We know if you win on the 36th try, you break even. We're gonna bet on that #17 (pulled the number from a hat). Let say one guy is flat betting but only up to 36 attempts, then stop. Another guy flat bets until the 35th attempt but then adds one unit on for the 36th and last attempt, then stop. Is one guy using gamblers fallacy and the other not? Are they both using it or are neither of the guys using gamblers fallacy? (LoL) This s**t drives me crazy.

Ken

You can't really answer these types of questions on the basis of observed behaviour, because GF is a BELIEF. If the guy believes that the number is "due" on the 36th attempt and that's WHY he increases the stake, then it's GF, but you can't say whether or not it's GF just by observing him increasing the stake, because he may have other reasons for doing it.

[superman]

there does seem to be some support for the claim that hot numbers are a better bet

Well, with my tracker I have now added a bit of code to print out every 250 spins, 250,500,750,1000 for the auto spin on BV, I'm asking it to print out which number has hit most at those points, I haven't done many tests yet but, whichever number is ahead at 500 seems to stay ahead at the 1000 spin point. The number ahead at 250 doesn't seem to be the one to win the race.

Was going to run a few more tests thi smorning but BV is down for maintenance.

[Bayes]

Just tested my file of actuals, there were 1239 sessions of 600 spins -

Hot   Cold
638, 601

Again hot numbers are the winner in the race (around 6% difference).

[MrJ]

Thanks for all your hard work Bayes.

(I deleted that thread of yours last night)

Ken

[Bayes]

Well, with my tracker I have now added a bit of code to print out every 250 spins, 250,500,750,1000 for the auto spin on BV, I'm asking it to print out which number has hit most at those points, I haven't done many tests yet but, whichever number is ahead at 500 seems to stay ahead at the 1000 spin point. The number ahead at 250 doesn't seem to be the one to win the race.

Was going to run a few more tests thi smorning but BV is down for maintenance.

There's some interesting stats for ECs on this. You'd think that there would be many times R/B would "cross over" in a few hundred spins, meaning R would overtake B and vice-versa, but it turns out that's not the case (there's even a formula which tells you the actual probability), so if R starts out winning, the chances are that after a couple of 100 spins it will still be in the lead. Maybe that applies to single numbers too.

[Bayes]

Thanks for all your hard work Bayes.

(I deleted that thread of yours last night)
Ken

No worries, Ken. And thanks for deleting the thread. 

[superman]

You'd think that there would be many times R/B would "cross over" in a few hundred spins

Yes you would, but I have to agree more often than not there is no cross over as I have tried chasing (flat betting) the colour/EC behind and yes you make a few chips but the percentages don't switch over most of the time.

Side not: BV finished the maint so I looged in and started auto spinning then placed a few chips on numbers unhit after 70 spins and the game told me to log out, so I did then they went into maint mode again for about 45 mins, now, I log back in and my balance is missing the 8 chips, I've emailed them to see what they say, I never got a result so how can my chips vanish LOL

[Bayes]

Side not: BV finished the maint so I looged in and started auto spinning then placed a few chips on numbers unhit after 70 spins and the game told me to log out, so I did then they went into maint mode again for about 45 mins, now, I log back in and my balance is missing the 8 chips, I've emailed them to see what they say, I never got a result so how can my chips vanish LoL

That happens if you close the game window without exiting from it in the normal way (like if you have to quit because of an error). Your balance might say zero but if you login and start another game it will tell you that there's a game in progress (the game you didn't quit from properly), that's when you'll see your balance. It is a bit alarming when that happens for the first time though. 

[superman]

That happens if you close the game window without exiting from it in the normal way (like if you have to quit because of an error). Your balance might say zero but if you login and start another game it will tell you that there's a game in progress (the game you didn't quit from properly), that's when you'll see your balance. It is a bit alarming when that happens for the first time though

Yes I understand all that, the balance did say zero so I resumed the game and the chips were missing still, they haven't replied yet either, I WANT MY 8 CENTS BACK LOL