A Money Management Challenge

[Kav]

Hello,

We can argue forever if roulette can be beaten or not.
However here is a challenge for you to test your creativity, knowledge and intelligence: Beat roulette given specific (hypothetical) certainties.

The challenge

You have 20.000 units bankroll. Your bet can be from 1 unit to 500 units (500 is the table limit).
You are only allowed to bet on RED, as often as you like.
In the next 100 spins the RED will appear exactly 30 times, but we do not know in which spins.
Make a profit of at least 1 unit after the 100th spin.

How would you do it?

[Proofreaders2000]

Two questions Kav:

1.) If we are only allowed to bet on Red, why
not play Baccarat (which has a better payout.)

2.) How do we know if Red has appeared 30 times unless it
has already happened and the game in question is not playable?)

[Kav]

Hello Proofreaders2000,

Thank you for the interesting questions.

1) Baccarat may have better payout, but don't know much about it. I work with roulette only.

2) The exactly 30 REDS in 100 spins, is the hypothesis on which the problem/challenge is based. Indeed, in reality you do not know for sure that in the next 100 spins, 30 of them will be RED. However this is a theoretical mental challenge, so we assume that we know that 30 Reds will appear, but we do not know when. 

Believe me this is an extremely difficult mathematical problem I pose here. If one can see the solution to this, that would be of great help even for sessions when we do not know the future out comes at all.

PS: At least you didn't ask me "why don't we bet on BLACK?" ;-)

[Chrisbis]

When you say we are only allowed to bet on RED as often as we like, do your soley mean the outside bet of RED,
or can we bet on the INSIDE numbers that are RED?

(Reminds me of that old joke........
What's Black and White and Read ALL over?)
[reveal=answer]
A Newspaper
[/reveal]

[Kimo Li]

Wait for 29 RED numbers to show up. Count how many spins did it take to get to the goal of 29 reds. Take the remaining spins and calculate the progression that you will use to make a profit of one unit when the 30th red does hit, using the 500 units as fight fund.

Kimo Li

[GLC]

Kav,  the modified labby was tested in this topic:



link:://:.rouletteforum.cc/index.php?topic=3280.msg29753#msg29753


The main difference is that the worst distribution in 200 spins was 75 or maybe 74 of the lowest hit even chance.  74 our of 200 is 34%.  Bayes couldn't find a series as bad as 60 out of 200.  I'm assuming that 30 out of 100 could be encountered because we know that the smaller the sample, the more out of balance the distribution can be.  15 out of 50 is easier to believe and I've seen 8 out of 25 and all of us have seen 3 or 4 out of 12.

I do think this is a very good challenge because if we can learn to beat 30 out of 100 or even 15 out of 50 it would give us something solid to work with.

Another parameter to consider, is the maximum number of even chances that can hit in a row.  So if we don't expect to hit more than 23 of an even chance in a row, we can use that in our computations.

Another thought is, okay, maybe we can get 20+ e.c.'s in a row, but what about some of our bet selection methods that use triggers to set up a bet opportunity.  Can they have 20+ losses in a row?  I personally have never had more than 12 losses in a row using my favorite bet selection trot.

It is worth noting that Fripper and Al did defeat these horror sessions.

[GLC]

It sounds like your challenge is for exactly 100 spins in sequence.  This eliminates the idea of 100 placed bets after triggers because this would take us into more than 100 consecutive spins.  Right?

[GLC]

One more thought is that if we can learn to defeat even 15 out of 50 or maybe 7 out of 25 would be acceptable, then we can use that as an attack procedure which can be incremented in case of the rare loss.

Let's say we use bet sizes from 1 to 60 to defeat a 7 out of 25 run.  And we only lose 1 time out of 100 attempts.  That would mean that if we had a recovery bankroll of 2-120 units, we have an excellent chance of staying ahead of the game.  And, heaven forbid, if we should lose 2 attacks in a row, and we were kin to Tomla021 and could have him back us with a tiny portion of his winnings to the tune of 4-240 units, we might have something to take a shot at.

Of course the other option is to just flat bet and have a good back up bankroll of say 500 units and it should give us an "almost" guaranteed shot at beating the game.

[Kav]

Wait for 29 RED numbers to show up. Count how many spins did it take to get to the goal of 29 reds. Take the remaining spins and calculate the progression that you will use to make a profit of one unit when the 30th red does hit, using the 500 units as fight fund.

Kimo Li

Smart thinking, but wont work if the first 29 REDs come in the first 60 spins. Then you'd be left with 40 spins and a single RED, which you still don't know when it will come.

Kav,  the modified labby was tested in this topic:

link:://:.rouletteforum.cc/index.php?topic=3280.msg29753#msg29753

The main difference is that the worst distribution in 200 spins was 75 or maybe 74 of the lowest hit even chance.  74 our of 200 is 34%.  Bayes couldn't find a series as bad as 60 out of 200.  I'm assuming that 30 out of 100 could be encountered because we know that the smaller the sample, the more out of balance the distribution can be.  15 out of 50 is easier to believe and I've seen 8 out of 25 and all of us have seen 3 or 4 out of 12.

I do think this is a very good challenge because if we can learn to beat 30 out of 100 or even 15 out of 50 it would give us something solid to work with.

Another parameter to consider, is the maximum number of even chances that can hit in a row.  So if we don't expect to hit more than 23 of an even chance in a row, we can use that in our computations.

Another thought is, okay, maybe we can get 20+ e.c.'s in a row, but what about some of our bet selection methods that use triggers to set up a bet opportunity.  Can they have 20+ losses in a row?  I personally have never had more than 12 losses in a row using my favorite bet selection trot.

Thanks for the reply GLC,

The problem with that thread - which is extremely valuable food for thought - is that the progression soon became too complicated (play a martingale mini-progression, change EC after 4 losses etc.). So, technically there is no clear answer in that thread.
What is interesting is that when faced with such a problem, we instinctively think of a negative progression (up as you lose).

The thing is, in case the hits come late, you try not to increase the bets too fast. on the other hand, in case some hits come relatively early you don't want to let them pass by without taking advantage, so you must increase the bets...


Another interesting twist to the problem would be the following:
Even if you don't manage to make a profit how would you minimize your losses. How can you minimize your losses in such a 100 session.

It sounds like your challenge is for exactly 100 spins in sequence.  This eliminates the idea of 100 placed bets after triggers because this would take us into more than 100 consecutive spins.  Right?

Since we take as granted, that whatever triggers or bet selection we have only 30 wins in 100 spins, bet selection doesn't count. Anyway, yes  "(it) is for exactly 100 spins in sequence" - yet you are not forced to bet on every spin if you don't want to.

One more thought is that if we can learn to defeat even 15 out of 50 or maybe 7 out of 25 would be acceptable, then we can use that as an attack procedure which can be incremented in case of the rare loss.

Let's say we use bet sizes from 1 to 60 to defeat a 7 out of 25 run.  And we only lose 1 time out of 100 attempts.  That would mean that if we had a recovery bankroll of 2-120 units, we have an excellent chance of staying ahead of the game.  And, heaven forbid, if we should lose 2 attacks in a row, and we were kin to Tomla021 and could have him back us with a tiny portion of his winnings to the tune of 4-240 units, we might have something to take a shot at.

Of course the other option is to just flat bet and have a good back up bankroll of say 500 units and it should give us an "almost" guaranteed shot at beating the game.

Now that's a VERY interesting thought! (flat bet would provide no advantage - it is as good as not playing roulette at all ;-)

[GLC]

I just looked at the worst horror session posted by Bayes and it starts out with only 4 wins out of the 1st 25 spins??

What if we bet 1 unit after each loss and then try to recover the losses after a certain trigger to increase our bets?

I have posted recently and attempted to do this by betting 1 unit until 3 losses in a row and then bet to recover those 3 losses with 1 bet.  My thinking is that losing 4 times in a row is not that common.  If the logic worked, it would improve by waiting 4, 5, 6 or more losses in a row and then attempting a recovery.  It's all trying to cheat the odds because we know that there's still a 50/50 chance for a win or a loss (not counting zero) no matter how many losses you've had in a row.

To be honest, unless someone can come up with something unique and never thought of before, I have nothing left in the gas tank.

[RouletteMaster]

think either win-loss +1unit shd work fine/place bets on red always/stop when you are +1 unit at any time.


assume you have first 70 spins as black/You are  -2485.
next 30 spins as red. You are +2565/Net +80/maximum placed bet is 100 - much below 500 table limit.

did i miss something?

[GLC]

Oops.  There was a drop of gas still in the gas line.

I like Flat's new BP system idea of using a 9 bet series to determine the next bet amounts.  Since we have up to a 500 unit bet limit, we could do something like bet 1 unit for 9 spins.  If we're up, repeat.  If we are in the hole, our next 9 bets will be whatever amount we're in the hole, but if we get back to even, we stop and reset back to 1 unit starting a new 9 spin sequence.

If we're in the hole at the end of the 1st attack and also in the hole more at the end of the 2nd attack, our 3rd 9 spin attack will be based on the total units we're in the hole.  This is just a little more aggressive than Flat's system.  It gives us the possibility of recovering anytime we start a 9 spin attack with a Win or a Loss-Win-Win or L-W-L-W-W-W etc...

If 9 spin attacks are too small, we can expand it to 11, 13, or more.

Just thinking out loud.

[GLC]

Another change to his idea is that any time we get ahead by at least +1 in the 1st 9 bet attack, we reset back to 1 on a new 9 bet attack.  This means that we don't increase our bet amount unless we end a 9 bet attack in the minus.  With luck, we could end our 100 spins and never have had to bet more than 1 unit.  Not likely, but possible.

[Kav]

Another change to his idea is that any time we get ahead by at least +1 in the 1st 9 bet attack, we reset back to 1 on a new 9 bet attack.  This means that we don't increase our bet amount unless we end a 9 bet attack in the minus.  With luck, we could end our 100 spins and never have had to bet more than 1 unit.  Not likely, but possible.

If we are ahead 1 unit, we just stop and do not bet on the next spins. That's not the problem. You don't have to bet each spin. The problem is to reach profit at some point.

[D1]

Hey Kav

Is there a way of doing this ?