I've just played a few very annoying games where I would've had better results but for failures on actions which should've succeeded, especially after re-rolls. This is the fault of the RNG... but only because it generated the failing numbers. On the boardgame, you'd say it was the fault of the dice. Anyone who has played the board game will have had dice fail on them in similar circumstances providing they've used normal, evenly-weighted dice. Failed to pick up a ball on two attempts with an Ag5 character and no TZs on the ball? It has happened in the real world and on here, and will happen again. The RNG is meant to replicate dice and sometimes dice go bad... how many players have people seen "warming up" their dice or doing some other strange rituals to stop this happening?

About the only thing I'd like to see changed is the analysis of it - at the end of the replay, you get a breakdown of how many times each number was rolled. If that was displayed for both coaches so they could see how many times they rolled each number individually they would show (hopefully) that the dice for each coach were about even. If it consistently showed a bias toward one coach (i.e. the server coach or the client coach) then it would need to be changed, but I suspect it's just that it is how it appears if you have failures at critical moments. Just get on with it and accept sometimes you'll beat someone because they've had a bad dice run and it'll probably even out in the end.