Divisors and debt banks are the same thing. Statistically it is easier to achieve a smaller win amount than a larger win amount. So my theory is, chain together smaller win amounts for a better statistical chance of elcipsing the target debt without encountering a counter-swing.
So, like this:
100 units of debt accrued at 1 bet unit.
We can try to regain this with 50 2 unit bets in one shot, OR we can divide the debt into 4 sections of 25 units, achieving each target goal of 25 units with only 13, 2 unit bets.
The second method offers the following advantages:
1) A smaller exposure in number of spins per session
2) A smaller exposure in chips at risk at one time
This is not the entire picture because our bet selection should also include a statistical component that will indicate to us when we have larger chances than normal to obtain a hit and raising during those moments will also provide a positive incline to our recovery.
So far this has worked very well for me and (I have heard from) others as well. The theory seems a bit sketchy but it would revolve around an exponential risk factor over a larger amount of spins. If we can prove that there is an exponential increase in risk over a number of spins (as opposed to a static increase) then we can prove this theory.
A common example of this might be understood through computer simulations. I have noticed, time and time again, that the very worst loss streaks don't appear until after 1M or even more spins/coup events; it just seems to happen every time that a larger number of spins/events is required for them to even show up.
Does this mean they can't or won't show up sooner? No, but it would tend to indicate that the chances of encountering rare events in a session increase the longer the session runs until we reach a "100% point" and the rare event occurs.
Both Dobble and the IDG have referenced this where IDG says, "At a certain point the statistics will "cross" your session and you will start losing".
He is actually saying two things here:
1) Set up your base progression system and profit goal to win in a short term span that will be consistently advantageous statistically
2) Set up your recovery progression system to win faster than your base progression system
I have attempted to emulate this and have met with suprising success.
This doesn't take into account another important factor which is bankroll.