Streaking D'Alembert

[Colbster]

Just about everyone on this board is aware of the D'Alembert, which starts at 1 unit and increases by 1 after a loss and decreases by 1 after a win.  The theory is that if you lose at 1, lose at 2, win at 3 and then win at 2, you will have won 1 unit for each win that you have had (the 3 won offsets the 2 lost and the 2 won offsets the 1 lost).  Anyone who has taken it for a spin knows that the streaky nature of roulette can turn against us quickly, with a series of losses taking our drawdown to steep levels quickly.  However, we know that a balanced number of wins and losses gives us a bunch of +1 wins equal to the total number of our wins for the session, with the wins at high points to offset the lower losses.

It is great that the +9 offsets the -8 and the +8 offsets the -7.  What is even better is when a +9 offsets the -8 and +9 also offsets the -7.  If a win on a D'Alembert is always at a recent high-water mark, so is a streak of wins back to back.

What I am doing with this progression is moving up on a loss like the typical D'Alembert.  If I get a win, rather than dropping a unit down, I repeat my bet.  If I lose, it is a wash and I repeat again at the same level.  However, if I get a string of wins, I let them continue winning at this high level for as long as the wins come.  I stop betting only after my first loss.  For instance, if I am at level 9 and I hit a string of 5 wins and then a loss, I have effectively won 9 units a net of 4 times.  This is +9 against -8, -7, -6, and -5, meaning that I have netted +1, +2, +3, and +4, or a total of +10 instead of the usual +4.  At this point, I will reduce my betting level by the net wins (4), meaning in the example I would begin betting again at level 5.  If I lose again, I move up to 6.  If I win at 5, I repeat again as before and look for another profitable streak.

Variation:

What you will notice is that a series of chops is nothing more than a wash in this method.  In the typical D'Alembert, each chop is a +1 unit for your bankroll, such as +9 -8 +9 -8 +9 -8.  The same can be accomplished with this method by decreasing after the first win and then holding at the next level down for the remainder of the streak.

[Colbster]

I'm getting good results using the variation that I mentioned in the last post.  Keeping the bets higher after wins instead of dropping allows us to take advantage of the streaks.  Thus, the recovery/profits come quick when we get a burst at the higher amounts.  To illustrate, if we were at level 9 on the usual D'Alembert, we would drop to level 8 after a win.  The win at 9 covers the previous loss at 8 and we now hope to win at 8 to cover the 7.  Staying at 8 instead of dropping to 7 after another win doesn't just cover the next loss amount of 6, it also gives us the loss back at level 1.  If we get another win, we cover both the 5 and the 2.  Finally, another win covers the 4 and the 3.  With just a couple wins, we are back even rather than needing to march all the way back.  Any additional wins at any level would results in profits now, including some really attractive returns if we happen to get a nice streak at a higher bet unit.

For EC, this is playing safe and giving good stability to what I already think is my favorite progression.

[ati]

Looks interesting, I'm trying to test it at the moment, but my luck fails big time. I'm betting on black and it hit only 4 times against 17 red.

[Colbster]

I had my first 200 unit bankroll lost today with similar. I'm testing black only and this session went straight to crap. 7 reds to begin and never came back, although there were some times I could have walked away about even. Still pleased, needs rules like anything else.

[ati]

Just finished a 50 spin test. I hit 21 times and ended up with -19 unit. I also compared to standard D'Alembert and that would have ended with -15 units. The longest series was five wins in a row at the very end, before that I was down over 100 unit. I'm gonna do more tests tomorrow.

[ausguy]

ati - How would it compare to a Guetting progression ?

[vladir]

This could only work if you have a bet selection with a good hit rate...  I tested this betting always HIGH  first and then Follow the Last for sets of 10.000 spins in excel ... end result is, most of the time, negative... once you get in a hole, it gets harder and harder to get out of it...

[superman]

This could only work if you have a bet selection with a good hit rate

There's no such thing as a trigger, any trigger you "imagine" will eventually settle at the expected percentage for the event you bet on, any EC trigger will = 50% return, on average guaranteed.

If you want to find a mechanical progression that works EVERY time you deploy it, you need to know the lengths and grouping of the losing runs, by that I mean 40 reds in 200 spins, how long where the losing runs between reds, studying groups of 200 spins that way, no 2 groups would be the same when fully disected, therefore no 1 mechanical progression will do. No matter how many times we revisit them, random is dynamic, we must be too.