You're kidding, right? Deaths are, on average, 1/6th of all cas. We know this from the casualty table requiring 61-68 to be a death. Nothing modifies the casualty table (as opposed to the injury table) so it's not exactly rocket surgery to work out the translation from cas to deaths.
You're welcome, btw.
Deaths are not covered here. I asked for average number of succesful blocks needed for causing a death.
I'm not convinced you're clear on what you're asking. Since we're talking about probability what you're asking is how many iterations will it take before the probability reaches a certain threshold.. often 50%, meaning the number of iterations at which point your chances of having achieved success (or whatever) is over 50%. Then, given your past threads, you're going to compare that number to your own number and cry foul, as though eyeballed comparisons have meaning.
I know you're aiming for this because this is the second thread like this - in the last one you asked what the probability of getting only 1 success on 32 trials was, and what the average number expected in 32 trials was... then proceeded to say that you only got 20% of that... which itself has no real meaning.
As for getting people to do all this math for you... its time you learned to do it yourself.
First, calculate the probability for something to happen on each repetition - in your case, we want the probability of causing a death on a single successful knockdown for a given AV. That means the chance of rolling over the AV value on two dice, then rolling a casualty on the injury table, then rolling a death on the casualty table.
To break armor you need to roll more than the value on two dice, so consult a probability table for dice if needed and find the probability of rolling that number, and add it to all the probabilities of rolling each of the higher numbers. That, out of 36, is your chance of breaking armor.
Next, look at the game's injury table and see what you need to roll to cause a casualty. You cause one on a roll of 10-12 on two dice. So go check that same table and add together the probabilities of rolling 10, 11, or 12.
Finally, look at the casualty table and see that there's a 1 in 6 chance for death, which is obviously 1/6.
Now multiply all those chances together. That's your per instance probability. Invert it (subtract that value from 1) and you get the probability that you will NOT cause a death on each successful block... now keep multiplying that value by that value over and over again until the number you see is below 0.5, and the number of times it took to do that is the number of blocks you have to make, on average, before you'll cause a death.
For MB you are doing the same process but you calculate it with the injury roll being 9 to 12 instead of 10 to 12, and then you calculate the probability of rolling exactly one number below the needed armor break value (not combining it with higher rolls), with the normal 10 to 12... and you combine the two results... which is the probability of causing a death on one successful block with MB.
Once you know the per-block probability you can use a binomial probability calculator of your choice to see how likely your outcomes were based on that probability.