Ok, good point. You bring up the issue of the PLE (Progresion Loss Event):BSE (Black Swan Event) which varies depending on the Progression Length, which yields a % chance (PLP).
Let's compare (percentages are estimates for ease of use):
The 95% method yields a HR:PLE ratio value of 8 vs. 2.4 of the 99% method.
As I have said several times, I tend to be mentally slow and especially at maths. Your observation is incredible! You are clearly showing that the -495 method is actually LESS risky than the 99% method!! I was INCORRECT in assuming that increasing the win rate automatically decreased the risk!! I failed to take into account the disparate cost.
Thanks so much for this!
Now, this method is about scrapping and so its better to scrap with less risk, so I will switch back to the -495 method at 95%. One good thing that I have noticed is the relation between the short-term expectation of hits:spins and the chances of a BSE. Let's experiment some more and see if this new info can be useful!
+28 +6 +19 +8 +30 +37=+129/+129 6/133 (6) <=== 89 spins ABOVE expectation
This is very close to violating Rule #2, but for now let's institute a 2X expectation delimiter. Continuing...
+33 +3 +14 +29 +52=+131/+260 5/205 (11)