This means that you have an 81.5% chance of being hit. In the 2-3 attack sample size, probability states you won't evade. In fact, it would take a sample size of 6 attacks or more for probability to state that you SHOULD evade at least one.
Just to show you that you should go back and check your stats. You define the probability of getting hit as 81.5%, and claim it takes 6 attacks to expect an evade. Actually, you should look at it like this: What is the probability of getting hit 6 times in a row with no evasion? 0.815^6 = 0.29. Only 29%. In fact, your odds of getting hit 4 times in a row are only 0.815^4 = 44%. In other words, more often than not you will get an evade within 4 hits if you have an 18.5% evade rate.
You completely contradicted your previous post in this. Keep in mind that Evasion attempts are completely independent of one another.
If you have a 5% chance at something and have a sample size of 1000 samples, the result will average at 50 for the 5% chance. If you have a sample size of 100, the result will average at 5 for the 5% chance. If you have a sample size of 10 samples, you will average 0~1. If you have a sample size of 1, you will average 0 or 1. The likely result, however, for a sample size of 9 or less 0. Regardless of how many you have evaded before hand. Taking consecutive mobs into consideration is a moot point simply because varying levels and stats will change results.
The whole conversation is a moot point, however, simply because... you're still barely breaking the evasion floor if at all.