[page2067]

Certainly, with practice matches at competitions considered, we passed 150 - but I am sure others go on further too.

But what can we learn from machines going 120+ matches - is there a wear out plateau - if you run x number matches then you can be fairly confident of reliability?

Or are we really still in prototype mode... still learning embryonic failure modes? even after 120+ matches...

I think in the past couple of years of running a least 100 matches that we uncovered a great proportion of issues, but cannot quantify that very well.

And cannot yet say if we found all issues or still learning.