@Scep: The chart is for winning debt back, so the statistics are subsequent to a 1 2 4 8 miss. I am a bit tired today, weak in math and easily confused, so maybe you can help me disentangle the apparent divergence in statistics. The chart is meant to show the devolving percentages after receiving 4 misses and the goal is to recover all debt in X number of coup attempts, where I took 96.43 as the base and devolved it by itself for each spin (and it really should be by each coup >.<). This approach is obviously terrible?

Of course one could not examine cumulative probabilty, staying with static probability, and we can always say that we have .4865 chances, but we would be kidding ourselves if we don't understand that those chances devolve after every subsequent consecutive hit and evolve after every subsequent consecutive miss.

Here is another game that was more interesting, all on EC's, using a flexible divisor:

1 02 1 02 04 1 02 04 08 16 01 2 1 02 04 08 01 1 1 2 1 1 2 4 8 1

8 16 8 12 18 **9 14 14 19 25** 13 18 9 14 14 19 18 9 5 7 4 3 6 6 9 5

which recovered 15 units in 26 spins, with an all time high debt of 98 units. This session was somewhat bad with the worst sequence running 2 hits in 252 numbers bet, which is 1:126 [is this -3.4 SD?] (vs. 1:37). The divisor fluctuated between 1 and 4.

The red bolded represents what would have been a full 4-step progression loss (.9643 event, occurring after only 2 hits), which is what we are trying to avoid, in using a divisor.

However, it is interesting to note that after 2 hits, we only would have went down to 75 units instead of 98. Things are worse if its after only 1 hit (90 units) and back to back is 135 units.

So, very interesting that only a back to back loss (.9952 event) would do worse than our divisor here!