Well app wouldn't be the place to start as randomness generators for software are not really random, but code faking it (some more sofistacted than others as far as i understand)
And I wouldn't not have mentioned it if it was a rare occurance it actually was a majority of time, especially obvious when rolling not with d20 but d100. I much more often had streaks of 4 fails or hits on 5 attempts for 50% than other distributions combined.
Naturally my perception isn't more accurate than other people's so it can easily be just biased observation. But a appropriate analyses would require more static skills than I want to deal with since simply counting the total results isn't going to show that something was off in this case
