opinions please

[Bayes]

If you could absolutely 100% guarantee a long term 2% edge over roulette by flat betting would that be enough for you to take up semi or full time play?

It certainly would. Even with a paltry 0.5% edge you could make millions over time. You would need to work out your risk of ruin but there are plenty of resources on the net for this. You bankroll would depend very much on the distribution you're playing (how many numbers). Playing only 1 or 2 numbers would require a bigger bankroll than if you were playing 30 numbers (due to the increased variance).

I think it would be easy enough to stay under the radar by playing at multiple casinos. E.g if your target is $600/day, open 6 accounts and take $100 from each.

Here's something from the Wizard of Odds' site (an article on gambling systems) which might be of interest.

System salesmen usually promise ridiculous advantages.  For example, even with just a 1% advantage on an even money bet, it would not be difficult to parlay $100 into $1,000,000 by betting in proportion to bankroll.  I was asked to prove this claim so I wrote a computer simulation based on the toss of a biased coin, with a 50.5% chance of winning.  At all times the player bet 1% of his bankroll, rounded down to the nearest dollar.  However, if a winning bet would put the player over $1,000,000 then he only bet as much as he needed to get to exactly $1,000,000.  In addition, I ran simulations with a 2% advantage and for a starting bankroll of $1,000.  Following are the results of all four tests.
$100 Bankroll, 1% Advantage

   
• Bets won = 7,182,811,698 (50.4999%)
• Bets lost = 7,040,599,544 (49.5001%)
• Player achieved $1,000,000 first = 79,438 (83.019%)
• Player went bust first = 16,249 (16.981%)
• Average number of bets to reach $1,000,000 = 174,972 (364.5 days at 8 hours per day, 60 bets per hour)

$100 Bankroll, 2% Advantage

   
• Bets won = 7,027,117,205 (51.0000%)
• Bets lost = 6,751,539,769 (49.0000%)
• Player achieved $1,000,000 first = 215,702 (98.099%)
• Player went bust first = 4,180 (1.901%)
• Average number of bets to reach $1,000,000 = 63,775 (132.9 days at 8 hours per day, 60 bets per hour)

$1,000 Bankroll, 1% Advantage

   
• Bets won = 5,213,026,190 (50.4999%)
• Bets lost = 5,109,817,544 (49.5001%)
• Player achieved $1,000,000 first = 74,818 (99.0285%)
• Player went bust first = 734 (0.9715%)
• Average number of bets to reach $1,000,000 = 137,208 (285.8 days at 8 hours per day, 60 bets per hour)

$1,000 Bankroll, 2% Advantage

   
• Bets won = 6,332,837,070 (50.9996%)
• Bets lost = 6,084,596,671 (49.0004%)
• Player achieved $1,000,000 first = 267,445 (99.9996%)
• Player went bust first = 1 (0.0004%)
• Average number of bets to reach $1,000,000 = 46,428 (96.7 days at 8 hours per day, 60 bets per hour)

These simulations prove that with just a small advantage of as little as 1% and a bankroll of as little as $100 you can grind your way to a million dollars through the gambling equivalent of compound interest.

So even with only a $1000 bankroll, your chance of going bust with a 2% edge is negligible.

Is this just a hypothetical question? 

If not I could write a simulation similar to the above to find your risk of ruin. Just let me know how many numbers you're betting on. 

[Skakus]

Thanks Bayes.

>>Is this just a hypothetical question?

Of course it is! 

The problem with these % of bankroll simulations is that in order for them to work out you can never touch the money >> I like to touch the money! But I get the point.

Thanks for the offer but I’ll have to get back to you on the risk of ruin simulation because I’m not sure how many numbers would be bet. Is an average alright, or is an accurate breakdown needed?

[Bayes]

yeah best if you say it's hypothetical, otherwise you'll get the proof police on your case. 

Compounding is the best way to build up your profits, but you could still withdraw a % when you say double or triple the bank. Kelly betting is the optimum MM if you have a real edge, but I expect you already know that.

An average is ok for the numbers. More accurate would be to take a weighted average. Example: 30% of your bets are on 3 numbers, 20% are on 5 numbers and 50% are on 10 numbers, so the average quantity of numbers bet will be 0.3 x 3 + 0.2 x 5 + 0.5 x 10 = 6.9 (7 numbers rounded up).

[GARNabby]

... yeah best if you say it's hypothetical, otherwise you'll get the proof police on your case. 

I think you're confusing (historical) fact with (logical) proof.  The hypothetical is no less subject to proof.

Pretty soon we'll all be opening one of those "plain 5% bank-accts"?

[GARNabby]

Here's something from the Wizard of Odds' site (an article on gambling systems) which might be of interest.

People like the Wizard forget, if ever realize, that the field work reigns in the simulations, as much as the other way around.  Ie, we're not a bunch of computers living in lalaland.

Besides, the Wizard isn't an expert at anything.  He started out as a simple actuarial science graduate, and ended up a "gambling writer"... neither a mathematician nor a writer.  That's what happens when someone "bounces around" aimlessly from one gimmicky "system" to another.  But you can still "sell yourself out", nay the ones who believe in you, and have someone like Bodog come along to "pick you back up".  LoL.

[Skakus]

Ok Bayes, what's the risk of ruin?

Edge = +2%

Weighted average = 5.62 rounded up = 6 numbers per hypothetical bet.


Kelly betting is one of the best if you have an edge. What do you think of The Turnaround?

My favourite MM method is the KISS method. I don't like holding calculations in my head when I'm playing. It's hard enough picking the numbers without trying to calculate how much to put on. 1 unit for each selected number is as complex as I get these days.  

[Bayes]

Ok, I thought it would be more useful to know the biggest drawdown over many sessions, rather than the RoR, that way you'll know what bankroll you need. 

To simulate a +2% edge, and assuming you lose 6 units per loss and gain 30 units per win, we have:

30p - 6(1 - p) = +0.02   where p is the probability of a win.

Solving for p, we get p = 6.02/36 = 301/1800. So this represents a "wheel" with 1800 pockets, and if the ball lands in say pockets 1 - 301 it corresponds to a win.

I simulated 10,000 sessions at 200 spins per session. The largest drawdown from the 10,000 sessions was approximately 700 units (I ran the test a few times and it was consistently around this number).

Since you're actually betting somewhat less than 6 numbers, this is an underestimate of what the actual drawdown would be, but I reckon a 1000 unit bank should be enough to cover the worst case scenario.

Kelly or % of bank betting will probably be safer, and more profitable over time. 

What do you think of The Turnaround?

Not sure what you mean, are you talking about the Turnaround system for EC bets?

[Bayes]

I think you're confusing (historical) fact with (logical) proof.  The hypothetical is no less subject to proof.

A hypothesis may be subject to proof, but this isn't a hypothesis in the sense of trying to explain anything, so it doesn't need proof. It's just a "what-if" question.

[TwoCatSam]

I have the time and the money.  Let's do it.

Sam

[GARNabby]

It's just a "what-if" question.

Well, the "proof police" have one up on the BS cops.

[TwoCatSam]

Skakus

"If you could absolutely 100% guarantee a long term 2% edge over roulette by flat betting would that be enough for you to take up semi or full time play?"

Are you saying this can be done or is this just a "what if" question?

Because if I could fly, Homeland Security wouldn't get to feel my junk!

Samster

[Skakus]

>>Homeland Security wouldn't get to feel my junk!

Samster

EEEUW!

[Skakus]

Thanks a lot for the drawdown calculation Bayes.

I reckon 1000 units is probably close but  doesn't account for much as long as things turn around. It's when you lose the whole 1000 then lose it again, and agian that gets you into trouble.

This is only a hypothetical question after all so it doesn't really matter.

Even If you could guarantee a 2% edge you still need to be able to afford the investment of both time and money. And that is a sizable investment. $80,000 locked up and 5-6 hours a day work for $400 a day pay cheque.

I don't think the casinos have much to worry about.

[Skakus]

>>
Not sure what you mean, are you talking about the Turnaround system for EC bets?

This one:


Roulette System - The Turnaround System
First, a glossary of Turnaround terms.-
This one comes from a PROFFESIONAL, and i haven't tested it myself. (lack of time)
A "Series" is a sequence of bets, usually one in which the player is in the hole and working towards a profit. You'll hear the term "recovery series" from time to time, and it tells you that every losing series is assumed to be reversible. There's also an animal called a "winning series": more about that later. Don't worry about losing series: Turnaround will run out of money once in a while if you impose a sensible stop loss limit, but it won't happen often.
The "LTD" is a pair of figures you will need to keep in your head at all times. If you can't handle simple mental arithmetic, this method is not for you. If you are a blackjack player who counts cards, relief is at hand -- Turnaround is a whole lot easier and more effective than counting and is virtually undetectable by pit cops. "LTD" is short for loss-to-date, and the first figure in the pair tells you how many chips you're currently down according to the special tracking method the system uses. The second figure indicates the correct next bet (NB); basically, your NB stays the same after a loss and increases according to a fixed formula after a win.
The "Steal" is a process by which the final profit when a recovery series is resolved is deliberately inflated. When you win, say, 3 units you steal one unit before deducting the win from the LTD and calculating the value of the next bet (NB). Example: -1, -1, -1, -1, -1, -1, -1, -1, -1, +1, +3, +4, -5, -5, -5, +5, +6, +7, +5. The LTD after the win that ended a streak of 9 losses was "9,3" because the single unit won was not deducted from the loss-to-date. The NB jumped from 1 to 3 units for reasons I will explain in a moment. The LTD for the series changed as follows: -1, -1, -1, -1, -1, -1, -1, -1, -1, +1 (9,3), +3 (7,4), +4 (4,5), -5 (9,5), -5 (14,5), -5 (19,5), +5 (15,6), +6 (10,7), +7 (4,5), +5 (EOS = end of series).

JUMPING AND PUMPING

The "Jump" uses the first figure in the LTD to set the value of the NB.  My recommendation is that you do not divide the LTD by less than 6 to get the revised NB, meaning that a seven-bet win streak will end the recovery series.
Don't misunderstand the thinking here: I am not saying that seven-bet hot streaks are frequent or even necessary to Turnaround's success. But I am saying that your next bet must always be guided by full knowledge of how much you must win in order to turn a recovery series around and come out with a profit.
The "Pump" is another device to maintain the value of the NB at a level that makes recovery attainable. Oscar's "plus one after a win" rule is pathetically inadequate: it won't get you where you need to go unless the dealer is having a very bad day. When the game starts going your way, you must press as hard as you can, keeping in mind that if you overdo it, a reversal will do you very serious harm.
The Pump percentage I use is 20 percent, meaning that after the first win in a potential hot streak, you multiply the previous bet by 1.2 to get the NB. The first win after a losing streak (the previous bet when you're trying to figure the Pump) was set by the Jump, and now it's time to press a little harder.
Here's a sample sequence with the Steal, the Jump and the Pump all in play: (49,8) -8, -8, -8, +8 (66,11), +11, +13, -16 (60,16), -16, +16 (61,19), +19, +23 (21,22). In this sequence there were 5 losses and 6 wins, but to get you to the opening LTD of 49,8 the house advantage must have been running high beforehand. Generally, you should expect to lose more bets than you win (the casino depends on that). The trick is to make sure that you "win more when you win than you lose when you lose," and that's what Turnaround is all about. In this example, the 5 losses averaged -11 units, and the six wins averaged +15. That's a good thing! Note that when the first figure in the LTD dropped to 21, the NB became 22 and not 23 x 1.2 (28), applying a rule that the NB after a win can never be less than the LTD divided by 6 or more than the LTD+1 unit. More about that in a moment.
Next: "Halving." This is exactly what it sounds like: at a certain point in a losing streak, Turnaround requires you to reduce further potential losses by cutting your bet down the middle. Again, balance is important here -- do it too soon or too often, and your bet will bounce up and down like a yo-yo gone wild, punching you further into the hole with each downturn.
The Halving rule limits the ploy to bets in double figures and requires you to hold off until you have lost two "whole" bets in a row or, in blackjack, have seen a double or split raked into the dealer's tray. What you're trying to avoid is the damage from a "volley" -- (60,10), -10 (70,5), +5 (66,11), -11 (77,6), +6 (72,12), -12 (84,6), +6 (79,13), and so on -- and since paired losses are about half as frequent as isolated ones -- the "two lost bets" rule works well.
If you can get away with it, keep halving until the NB gets down to single digits, remembering that the lower you go, the bigger will be the Jump when you finally win a bet and have to recalculate the NB as LTD/6. Here's another example: (169,58), -58, -58, -29, -15, -8, +8 (330,55). It's quite a jump from 8 units to 55 -- more than a parlay, that's for sure -- and in blackjack particularly, you should limit yourself to no more than two bet reductions if the pit cops are "on heat."
THE DUMP
The final Turnaround oddity is the "Dump." Translation: run for the hills. This is a concession to the tendency we all have to beat a safe retreat when a dealer gets hot or our "luck" runs cold. Computer simulations presume that a player will stick it out through thick and thin, betting bravely onward regardless of repeated punishment and waiting patiently for a negative trend to reverse itself. Human beings just don't play that way.
Take a walk if the house gets, say, 10 bets ahead.
I'd rather not apply the Dump at all, but human nature -- mine, at least -- doesn't permit me to bet on and on like a senseless robot when I'm getting my butt kicked. So, it makes sense to regulate the bail-out process. A guiding principle of Turnaround is to provide a rule to cover every situation that could possibly arise in a game of chance, relieving the player of the agonies of indecision. OK, that makes me a robot after all, but not a senseless one, since all of the system rules are based upon logic and common sense.
When Turnaround made its debut on the Internet in April of 1997, several math-mad computer mavens leaped forward to offer to write source code for simulations which would prove that this particular method of progressive betting is no better than any other. After months of to-ing and fro-ing, caused by the inability of all these experts to apply the Turnaround rules correctly, the most persistent among them pronounced that my method had an expected win rate of over 99.5 percent or about the same as a Small Martingale.
Secondly, the Martingale or double-up method is doomed as soon as it hits a losing streak of a dozen bets or more, whereas Turnaround freezes the bet in response to a loss and requires it to be repeatedly halved after a certain point. So, a Martingale bettor trying to win at a casino with a 3,000-unit house limit will run out of options once a dozen consecutive losses require that his next bet be 4,096 units. He's 4,095 units "in the hole" and the math guys say that if he reduces his next bet to 3,000 units and wins, then he's still in trouble.
 
THE Turnaround RULES
Here are the system rules, step by step:-

1. Bet the minimum (your own or the table minimum) on the first decision in a series. I recommend $5 units to start with, meaning that if the table minimum is $25, your first bet is 5 units and will go up or down in increments of $5. Don't be tempted to use a table minimum above $5 as your increment (raising your bet from $25 to $50 to $75 and so on at a $25 table).
2. WIN, bet MIN+1, and be ready to follow the Turnaround win progression: (+1), 2, 2, 4, 7, 11, 16, 22, 29, 37 etc. LOSE, bet the same amount. At a $25 table, the progression will be (+5), 6, 6, 8, 11, 15, 20, 26, 33, 41 etc. When the win progression ends, set the next bet (NB) at 2 units, except when a split/double increases the loss to more than the initial bet, in which case freeze the NB at that point.
EXAMPLE: +1, +2, +2, +4, +7, +11, -16, +2.
EXAMPLE: +1, +2, +2, +4, +7, +11, -32, +16.
3. Any win after a succession of losses invokes the Steal principle, whereby a single chip is deducted from the win before the loss to date (LTD) is revised.

4. The LTD consists of two figures, best expressed as "n1,n2." The first figure, n1, is the amount DOWN at any point, and is inflated by the Steal; n2 is the amount OUT -- meaning the value of the next bet (NB) "out" on the table. Think of the LTD as "Down comma Out" and the mental chore of revising it after each outcome will quickly become second nature.
EXAMPLE: -1, -1, -1, -1, -1, -1, -1, +1 (7 DOWN, 2 OUT), +2 (6,3), -3, -3, +3 (10,4), +4 (7,5), +5 (3,4), -4, -4, +4 (8,5), +5 (4,5), +5 (EOS = end of series). HA 16% but TA wins 8 units = 16% of Action of 50u.
5. A win after one or more losses also invokes the Jump, whereby NB is calculated by dividing n1 (the inflated loss to date) by a predetermined amount. I recommend that if n1 is 20 or less, NB = n1/4; if n1 is more than (>) 20 and less than (<) 51, NB = n1/5; if n1 > 50, NB = n1/6.
6. After a win, NB must never be LESS than the first figure in the LTD (n1) divided by 6 or MORE than n1+1. If EOS is achieved, NB always falls back to the minimum.
7. After a win to which the Jump has been applied, use the Pump to press the NB further. This increase should be 20 percent, rounded, meaning that it has no effect until the previous bet (PB) hits 8 units (8 x 1.2 = 9.6 = 10).
EXAMPLE: (99,17), -17, -17, -17, +17 (134,23), +23, +28, -34 (119,34), +34 (86,41), +41 (46,47), +47 = EOS. NOTE that 41 x 1.2 = 49.2 = 49 but the final Turnaround bet was reduced to 47 that your next bet (NB) must never exceed DOWN+1.
8. Use Halving to limit potential losses as soon as you lose two successive bets of 10 units or more, and keep Halving until NB reaches single digits. In blackjack, start Halving if a double/split (two or more hands, in effect) costs you 20u or more. Blackjack players conscious of dealer/pit scrutiny should not use Halving more than twice in any sequence.
EXAMPLE: (99,17), -17, -17, -9, +9 (134,23), -23, -23, -12, -6, +6 (192,32), -32, -32, -16, -8, -8, +8 (271,46), +46, +55, +66, +79 (28,29), +29 = EOS.

9. Halving does NOT apply if a win progression ends with a lost double/split but takes effect if another double-digit loss is suffered.
EXAMPLE: +1, +2, +2, +4, +7, +11, -32 (17,16), -16, -8, +8 (34,10), +10, +12 (14,14), +15.

10. If, as above, raising the NB to match DOWN+1 will increase it by not more than 10 percent, make the adjustment in the hope that EOS will be achieved. Otherwise, a series will be extended for an extra one or two bets, which is contrary to our objective: wrapping up a recovery series in as few bets as possible before reverting to a minimum bet and starting a new series. In our example above, a win at 14,14 would have given us an LTD of 1,2 because of the Steal.

11. Use the Dump to stop play in a damaging location, transferring the LTD in full to a new deck, shoe or table whenever it's convenient. Experienced gamblers know that trends exist, and that a persistent downturn will sometimes seem unbeatable. The basic rule is to quit playing and move as soon as the second of two successive losses takes the LTD above 100 units, then 200, then 500, 1000, 2000 and so on. Don't Dump if Halving has already begun, or within five bets of the last move. An alternative is to track wins and losses and Dump as soon as the house gets, say, 10 bets ahead of you in any series (for instance, if the house has taken 16 bets vs. your 6). The advantage of this method is that it's more consistent than the original variable trigger approach.

12. In blackjack, play strictly according to basic strategy. There are about as many variations on the BPS (basic playing strategy) as there are "experts" on the game, but if you stick strictly to whichever book you choose, you will reduce the house edge to substantially less than 1 percent. Here's the BPS I recommend...

STRATEGY TABLE #1 Turnaround FINAL
6 5 1 1 1 1 1 1 1 A A A A A A A A A A 2 3 4 5 6 7 8 9 X
2 3 9 0 1 2 3 4 5 6 2 3 4 5 6 7 8 9 X A 2 3 4 5 6 7 8 9 X
------------------------------------------------------------
A: H H H H D H H H H H H H H H H H S S - T H H H H H H T S S
2: H H D D D H S S S S H H H H D S S S - T T T H D H T T T S
3: H H D D D H S S S S H H H H D D S S - T T T H D T T T T S
4: H H D D D S S S S S D D D D D D S S - T T T H D T T T T S
5: D D D D D S S S S S D D D D D D S S - T T T T D T T T T S
6: D D D D D S S S S S D D D D D D S S - T T T T D T T T T S
7: H H D D D H H H H H H H H H H S S S - T T T H D T T T S S
8: H H H D D H H H H H H H H H H S S S - T H H H D H T T T S
9: H H H D D H H H H H H H H H H H S S - T H H H D H H T T S
X: H H H H D H H H H H H H H H H H S S - T H H H H H H T S S

D = DOUBLE S = STAND T = SPLIT H = HIT

Notes: With 4,4 against 5 or 6, split if DAS, otherwise D; some
"experts" (usually card counters) recommend standing with 7,7
against 10 and standing with 16 against A but it's a bad idea;
never, ever, take insurance or even money (which is the same
thing); there's a case to be made for doubling A,8 or A,9
against 4-6, but it's a very aggressive play.

"CUSTOMIZING" Turnaround
Many of the rules of Turnaround can be adapted to suit the style of the individual player. For example, since the Steal drives the required bet rapidly upward in a downturn, you might feel that once it has been applied often enough to guarantee a reasonable win when EOS is eventually achieved, it should be abandoned after a certain point. Let's say you decide to ditch the Steal once the LTD hits 50. Here's how a tough sequence of bets might look: -1, -1, -1, +1 (3,2), -2, -2, -2, -2, -2, -2, -2, +2 (16,4), -4, -4, -4, -4, -4, -4, -4, -4, -4, +4 (49,9), -9, -9, -9, -9, -9, +9 (86,15), +15 (72,18), +18 (55,22), +22 (34,26), -26, -26, -13, +7 (93,16), +16 (78,19), +19 (60,23), +23 (38,28), +28 (11,12), +12 (EOS). Here, the house won 27 bets out of 40, but Turnaround bested an HA of 14/40 = 35% by exploiting upturns and backing off during downturns, overcoming the fact that losing streaks were longer and more frequent than winning ones. The 27 losses cost the player a total of 163 units for an ALB (average losing bet) of 6 units. The 13 winning bets recouped 178 units for an AWB of 14 units, kicking the HA all the way back to breakfast-time and making a profit of 15 units against horrendous odds.
If we "replay" those same hands with the Steal abandoned about half way through, the sequence looks like this: -1, -1, -1, +1 (3,2), -2, -2, -2, -2, -2, -2, -2, +2 (16,4), -4, -4, -4, -4, -4, -4, -4, -4, -4, +4 (49,9), -9, -9, -9, -9, -9, +9 (85,15), +15 (70,18), +18 (52,22), +22 (30,26), -26, -26, -13, +7 (88,16), +16 (72,19), +19 (53,23), +23 (30,28), +28 (2,3), +3 (EOS). The effect in this case was simply to reduce the "Turnaround bet" -- the last bet needed to end a recovery series with a profit -- from 12 units to 3 units. Since the critical bet was a winner, that was a bad thing, but if it had lost, the next bet (NB) would also have been much smaller, thus confronting the player with a greatly reduced exposure.
The recommended levels for the Steal (-1), the Jump (LTD/4,5,6) and the Pump (PB x 1.2) could all be said to be "aggressive," and a cautious player could limit his exposure while still following the Turnaround rules and principles. Keep in mind that the more timid the play, the longer it will sometimes take to exploit an upward trend and end a series with a profit before falling back to a minimum bet. When we reduced the Steal above, we cut the EOS win from 15 units to 6 -- a high price which a conservative player might consider worth every penny. Note that in the first version of the sequence (the recommended approach) Turnaround won 15 units on action of 341 units, achieving a player "hold" of 4.4 percent. A 1-chip flat bettor would have lost 14 units on total action of 40 units, giving the house a "hold" of 35 percent. If you play Turnaround for long enough, even if all you're willing to do is try it out at home, you will learn that making money from sequences of bets in which the house had the clear edge is commonplace.
The concept of "stealing" one unit before figuring out the loss to date and the related next bet is unique to Turnaround in spite of the fact that reward proportionate to risk is as reasonable and sensible a goal in gambling as it is in any other money-making endeavor.
In contrast, Turnaround applies the requirement that the longer a recovery sequence (or series) lasts, the greater should be the final profit. I have experimented with applying the Steal to lost bets as well as winning ones, and naturally the end result is always an even greater profit; the downside is that when wins are few and far between, the strain on the bankroll is greatly increased.
For example, consider a TA sequence which plays out like the one used at the very beginning of this article: -1, -1, -1, -1, -1, -1, -1, -1, -1, +1 (9,3), +3 (7,4), +4 (4,5), -5 (9,5), -5 (14,5), -5 (19,5), +5 (15,6), +6 (10,7), +7 (4,5), +5 (EOS = end of series). With the Steal applied to all bets, win or lose, the same sequence would be as follows: -1, -1, -1, -1, -1, -1, -1, -1, -1, +1 (18,3), +3 (16,4), +4 (13,5), -5 (19,5), -5 (25,5), -5 (31,5), +5 (27,6), +6 (22,7), +7 (16,8), +8 (9,10).
Stealing after wins only earned us a profit of 7 units and delivered a resolution after a series in which the house won 12 bets and we won 7 bets. The more aggressive approach gave us a profit of 10 units but left us in need of another win to achieve Turnaround. The effect of applying an all-around Steal, then, is to win more chips after facing greater risks. Either way, you can be confident that Turnaround will consistently hand you a profit after you have lost more bets than you won.
THE RED/BLACK TEST OF Turnaround

If you're among the doubters, try this little test: take the K and Q of hearts out of a single deck, creating a stack of 50 cards in which there are 26 blacks and 24 reds and black therefore has an advantage over red of 4 percent. Shuffle them rigorously, then turn them over one by one, using the Turnaround method to bet that the next card will be red. The HA will vary before each card is dealt, depending on how many blacks and reds have preceded it, but you can realistically expect a 4 percent house edge before the first card is turned over, and the same house edge before the 50^th card is exposed. Time and again, you will finish the deck with Turnaround substantially in profit in spite of a clear, absolute house advantage. The red/black test will sometimes hit the end of the deck in the middle of the recovery series, but then you do just as you would in a casino and roll over the LTD to the first "hand" of the new "shoe."
 
Lastly, a clarification of the Dump. Walking away from a costly string of losses is not merely an option for a real gambler (as opposed to someone who bases his "knowledge" of casino games upon computer simulations): it's a fact of life. Everyone does it, and everyone should do it. It's a difficult thing to regulate because so many influences push us to the point where we don't want to play any more against a particular dealer or at a particular table. To name a few: dealer hot streaks, prod-nose pit bosses, incompetent and/or inebriated players, fatigue, boredom, hunger and a stressed-out bladder. Proponents of computer simulations argue that their pale imitations of casino games are mathematically accurate, but in truth they can't be. No simulation exists which can imitate more than one or two components of real casino play at a time, and their most critical omission is human nature. I don't recommend changing your bet or varying basic blackjack strategy on a whim because one of the primary principles of Turnaround is that consistency is critical to long-term profit. But I do encourage you to back off from a game the moment you begin to feel uncomfortable or pessimistic: never, ever play if it doesn't "feel right."
There's no way a computer simulation can ape a gut feeling that tells a player when to quit, and it's difficult to devise rules to cover that situation. Turnaround's Dump will sometimes save your bankroll and will sometimes pull you out of the frying pan and into the fire. The one advantage of a computer simulation is that it enables you to test a variety of strategems against the same set of outcomes, but even when analysis of a million plays indicates that you would have won more without the Dump than with it, you can't say, "OK, I'll never walk away from a game again" because uninterrupted betting is an impossibility in real life.
I suggested earlier that one way to "Dump" is to walk away from a shoe (or a series) if the house pulls 10 bets ahead of you, winning 15 bets out of the first 20, for example. At -10, you won't have to move very often, whereas at -5 you'll be abandoning non-profitable situations pretty frequently, taking your LTD with you to another location and resuming play per the system rules. Personally, I favor frequent moves and "natural breaks" -- they alleviate boredom and keep me alert, and reduce the temptation to wobble off the rails and temporarily depart from the rules (always a no-no!). Bottom line: it's up to the player.

WRINKLES

I can't over-emphasize the wisdom of the individual player tailoring Turnaround to meet his/her needs and circumstances, so long as the key elements are not totally disabled. I didn't offer specific suggestions in the original presentation on my web site because I was concerned that the system was already quite complicated enough, and E-mail from all over the world confirmed that assessment. Now that Turnaround is, in effect, "retired" from the Internet, I can cover this topic in greater detail.
Anyone who has bet against a game of chance for just a few hundred outcomes knows that while skilled money management can beat the odds for very long periods of time, eventual damaging streaks in favor of the house are inevitable. Turnaround's damage control is effective but not infallible, and absent a very large bankroll, it is sometimes wise to help it out a little. A technique I call "Breakdown" would take an LTD of, for example, 19,3 x $100 units and break it up into four separate targets of 19,3 x $25 units. It will take much longer to achieve Turnaround -- at least four times longer! -- but the strain on the bankroll will usually be substantially reduced.
"Breakdown" assumes the use of a couple of other Turnaround concepts, "Multi-level Betting" and "Target Betting," both of which also further complicate the system but greatly enhance its effectiveness, and extends the idea of the Dump (a natural break in play in response to a downturn) into something called "Deferred Betting."
The idea of Multi-level Betting is to enhance the system's response when the house edge is biting extra hard. Assuming a start-up at $5 units, you would apply the Dump when the LTD exceeds 100 units and make a quick conversion to $10 or $25 units before resuming play against a new shoe. For example, an LTD of 117,19 x $5 would become 59,10 x $10
or 24,4 x $25, the idea being that the value of incremental increases goes up and the likelihood of achieving Turnaround is often improved because fewer consecutive wins are needed. In essence, use of the multi-level conversion puts a little more power behind the Pump, which increases the next bet (NB) by 20 percent after every win until the LTD+1 ceiling is reached. Say you're at 117,19 x $5 and you log a win, bringing the LTD down to 99,23 and the NB to $115. At 59,10 x $10, a win would take the LTD to 50,12 and the NB to $120, and at 24,4 x $25, the LTD would drop to 21,5 and the NB would therefore be worth $125. Small differences, maybe, but small differences are what Turnaround is all about!
"Target Betting" is a sneaky profit-builder which is, in essence, Deferred Betting without any actual prior losses. It exploits the fact that the mathematicians are right on the money when they remind us that the cards, dice or whatever have no memory and so are unaffected by bets that have been won or lost to date. Usually, Turnaround starts betting in a new series with a clean slate and the opening wager is therefore 1 unit and the LTD is, in effect, 0,1. But what if we open play with a fictitious LTD of, say, 19,3? If the house edge stays reasonably close to the norm, we'll make more money when Turnaroundachieves end of series (EOS) is what.
Here's a blackjack Target Betting example, starting out with a spurious LTD of 19,3: -3 (22,3), +4.5 (BJ), +4, -5 (22,5), -10 (32,5), +5, +6, -7 (30,7), +7, +8, -10 (27,10), -10 (37,10/2), -5 (42,5), -5 (47,5), +5 (43,8), +12 (BJ), -10 (42,10), +10 (31,12), +12 (20,14), +14 (7,8 TA), -8 (15,8), +16 (EOS).
Of course, Target Betting increases the risk that a runaway house edge will push the Turnaround bet into what I call brown trouser territory, and the potential exposure disqualifies it from the repertoire of a conservative player. It's just another wrinkle in the Turnaround technique: an example of how the system can be "customized" to suit the personality of the person who's using it.


The absolute minimum required to beat a pendulum that refuses to swing your way is 100 units. My greatest success has been with a 500-unit bank, but the real objective of Turnaround is to keep building your stash by adding to it after every winning session, splitting the take down the middle and extracting only half of your winnings as true profit.

[Bayes]

If you really did have a guaranteed 2% edge then that means you couldn't lose in the long run. Things won't turn around. Problem is that the only way to guarantee it is to prove it mathematically, and since it's already proved mathematically that you CAN'T get a positive expectation, you're always going to wonder how long your edge will last, and how solid it is based purely on past results (in other words, if it's based on past results, it ain't guaranteed).

I coded the Turnaround system years ago, in fact it was one of the very first systems I played. It can win for VERY long stretches, but it requires a big bankroll and the balls to use it. If you get into a prolonged bad run, it takes forever to recover.