Roulette New Thread

Read luckyfella post.

From what i read he play quad / dozen.

 

BUT still remains a question:" does he play 1 unit per spin or can be more units per spin????"

 

I only made the first statement, the other stuff is from someone else.

I wouldn't say forget everything. You have to anchor to math, house edge is real and can have a corrosive effect so work out how to deal with that and stay within the square root of the variance. But the biggest mistake people make is taking the long term average from the infinity view and applying that to the short game. Hence the faulty conclusions and the constant pot shots by the intellectual light weights who got schooled on WoV.

It's like in the long term we all die so everything is pointless. Don't bother going on vacation, living a fullfilling life and enjoying time with friends and family because the long term it matters not, death is certain.

But, but, but I can enjoy life you say. Sir Anyone appears and says nonsense, there are too many days and not enough heatbeats you're going to die, everyone dies, you think you can escape death? why are are you disputing this fact? We are all laughing at you over here on the WoV as we know it's pointless trying to live, after a million heart beats the reaper wins, here's a picture of him. (Insert grim reaper postcard).

They lead you to thinking a number will show 1 in 37 on average and that the house edge means you can never win. The fact is the probability is theoretical, what we witness is not perfect. With a very large sample, probability will get close to its theoretical but never perfectly close, even a million spins there will still be differences from expectation but the deviation does get smaller and smaller. There's an important concept there, the deviation gets smaller and smaller so you need to make a net profit in a session before it gets too small to beat the house edge. Variance is what beats the house edge and the deviation is the square root of variance.

Probability is actually an average on an imaginary infinite sample. Remember that. Variance will make more a fool out of you than anything else.

That is why I use methods to understand the limits of variance in the short game, that's why I know winning at least 1 unit in 69 spins from two hits (and more are possible) has a 55% probability of success.

And at least 1 hit in 26 spins has 50.95% probability to net at least 8 units. Whats not to like?

But then the rabble will say, well you might win more than 50% of the time but what about all the times you lose, They will say you lose more than you win. But they are wrong because they haven't done the math.

I really encourage anyone to do the math as the house edge is not to be seen anywhere until you lock up the win, and then who cares, you won, and it's cheaper than the tax man.

Sir Anyone will focus on the infinity view saying there are 37 pockets and the true odds are 1/37 and you only get paid 35. But he misses the fact that the average return is positive and already better than 50% at spin 26 to win at least 8 units, I don't need to go to infinity, ever.

 

FYI: I've posted recently about PRNG limitations.

 

This is mathematically correct. A few questions pop up:

1) Theoretically, a number can go missing for like 600 spins. How can we avoid this? Do we need to avoid it at all?

2) Is the payout enough? I believe it's not. So either we need a progression, which converges to a losing game again, because the progressed bets can be classified as identical bets, only with higher amounts, or we need a method to get better hit rates, which can be done by A) avoiding negative variance, and B) taking advantage of the current imbalance in the game, until it's there.

3) Is there any way to catch the imbalance in the game before it disappears?

PS: Thank you, @TwoUp , for the crystal clear posts. You are appreciated.

 

I understand the limitations when we talk about how the numbers are generated (seeds). The question is, if it acts differently, why the results are identical? You will get similar hit rates. Although variances may get bigger than physical RNGs. I can't say. I personally have not witnessed any difference between real wheels and PRNGs.

 

No point worrying about the events outside your envelope of profitability.

A number can theoretically be lost for a million spins or more but it's unlikely (the infinity view isn't useful).

But as a useful upper limit a number can be lost up to 757 spins which exceeds a 1 in a billion event (1,017,916,227).

Again I suggest get a spreadsheet and do the math for each spin to calculate the probability of a streak of losses 1,2,3 ... 300.

To compute the probability of a loss streak for n spins, it's q = (36/37)^n. That says that we will lose 36 out of 37 times betting on a single number. For example the probability of losing twice in a row is (36/37)*(36/37) and the probability for losing n spins in a row is (36/37) to the power n (n being the number of spins). In excel its =POWER(36/37,n).

Now compute the expected cumulative loss for a spin level: cum_loss = -(q * n) because we are flat betting 1 unit and at spin n we have lost n units. This is always negative value.

Now to compute the probability for the cumulative wins. We simply subtract the loss probability from 1. So p = 1-q This provides the cumulative probability of all the non losing events up to spin n. Sometimes you will win on spin-1, spin-2, spin-17 etc and this provides that cumulative probability of all wins up to spin n.

Now compute the expected value of the cumulative wins to spin level n. It's cum_win = p * 35 as we are flat betting 1 unit and the bet only pays 35 units on a win and this is where the house edge is paid.

Now add cumulative wins and losses to get the net expected value at spin n. Its ev = cum_loss + cum_win.

Do the calculations in a spreadsheet for all spins 1 to say 300.

Notice anything interesting about the ev?

Now divide the spin ev by the spin number n to compute an average EV per spin: ev_per_spin = ev/n

You will see that playing up to a certain point provides the optimal average ev per spin.

Now go and do the math and prove it to yourself. Verify the calculations, double and triple check, don't believe anything I've said, study probability if you need to but not on gambling forums.

And then do it again for other bet combinations, splits, streets, quads, dozens/columns, and EC's.

You will get an insight into things you may have never realised before. It's not all the same-same.

 

Great post, @TwoUp ! I'm on it! It's funny that I have already done all the math mentioned in your post, but never considered a step-by-step analysis. I always expected it to give me the same old results. Let's see...

Another question: Let's say playing a straight-up up to x spins gives us positive EV. When we continuously do this, how is it different than playing one number for like x*100 spins? Since we are only flat-betting, all of our bets together construct a continuous betting of 1 unit on one number. Am I missing something here?

 

I have to make this important post for the readers.

This thread contains the solution you seek.

Read every word carefully.

Do not ignore or change anything.

Your current problem is,

You don't properly understand what's written. Mostly partial understanding.
That's the problem.

To solve this problem look up the topic, do research on it. That's the first thing to do to get proper and complete understanding.

You get proper and complete understanding, the answer on this thread becomes clear to you.

Good luck guys.

 

We have NO problems .We ignore B S camouflaged as advice .

 

Knowing the probability doesn't enable you to side step it and exploit it for obvious reasons. 1. The house payout is short of what the odds say is fair.
2. Each spin is an independent trial.

 

Knowing how long a number is likely too sleep 90% of the time is no more useful than knowing the probability of a number hitting on the next spin. Knowing the probability of likely winning only tells you that you're at a disadvantage.

 

I suggest that you do as suggested and if you do not find anything remarkable THEN you call it BS. But I doubt that you even understand what 2up wrote.

 

I went over all your posts and unfortunately I didn't find anything of substance. You keep on preaching that you are the only expert, that people should read and work and that the truth is all there out in the open. And you complain a lot about people bashing you.
Why don't you simply stop doing that? Where is the gain? For you and others?
If you were on one of these car forums they would have banned you in no time if you said: "I am the only one who knows how to change and set a timing belt on model XY but I won't tell. You are just too stupid to understand what's already written. Read and learn...". I am not sure why they put up with all this c..p here.

I will delete my account here. Does anyone know how to do that? According to GDPR this must be possible.

 

For my testing I use entropy. I've worked with random long enough that I want the truest picture I can get.

They may be good enough for some things but I know where the limitations are, random number generators have many internal factors, recovery from internal zero states, insufficient degrees of freedom and then fail some important statistical tests.

Fortuna is the best one I'm aware of and it does use real entropy as well, if your testing is not using that then you might want to reconsider validating your final results with something more robust.

 

I won't say I'm the only one.

Those on my list are TurboGenius, TwoUp and Raf(apologise I miss you because you seldom post).

These guys know what they're talking about.

Why this "expert" list?

To tell those still interested and doubters, on the fence that,

Yes you can design a systems betting strategy that win.

This is to counter the wrong naysaying message bombarded on forums.

None of us bother to brag.
What for?

You wrote a "you are too stupid..."

The truth is yes you are.

But I think my words sound more like "you are ignorant." I try not to be rude.

And the solution to ignorance is education.

Nathan Detroit call such advice to educate oneself as bs advice.

I notice there's this assumption that people have enough knowledge level when that's not the case.

TwoUp did a splendid job doing all the writing.

It's a waste if you don't educate yourself to the level to be able to use his posts to help you design your own systems betting strategy.

This is the best time to learn with TwoUp, TurboGenius forum and a little of Raf posting together at the same time.

The forum problem is the distracting naysayers posts. And guys like Nathan Detroit that clutter the thread.

It will be good for the forum if these naysayers keep their activity local to TurboGenius new thread.

If anyone don't find my posts useful, just ignore. It doesn't bother me. It hurts if I wrote the reason why.

 

I unit a spin.

I wrote this detail to tell about the possibility when most people assume to be impossible.
Not to brag.

Remember, I wrote there are many "ways" to exploit. But the general principle is the same. TwoUp wrote the principle.

 

This thread is incomplete without some form of evidence.

I posted my Rsim sessions on TurboGenius thread.
https://www.gamblingforums.com/threads/when-does-it-lose.22770/#post-127108

 

Those two calculations are incorrect. I've explained why on the other thread:
https://www.gamblingforums.com/threads/when-does-it-lose.22770/page-2#post-127147

 

You could not be more right. But what if you don't care about the times that you get it wrong? It's possible to win anyway without knowing what is going to happen. Reading Randomness is waiting for it to be playing your game. It always has been since 2006. It's published here in the Roulette section. It tells you how it works. The thread serves as a self taught instruction tutorial and allows you to prove to yourself that it works. It does this so that you won't have to spend any money on learning to gamble before becoming an expert at it.