Bot results - Interpretation thread

When I study the InforRNA bot-solving of my puzzles, I get confused (all the results for my puzzles can be found at my in-game profile page).

It has previously been said that the bots only solve shape, without concerning themselves with constraints (max GC, min GU etc).

How then, can Crossroads be almost twice as hard to solve as Crossroads II, for the InfoRNA bot?
The only difference between those two are a free base at the end of Crossroads II (apart from the constraints, which the bot won’t consider).

Crossroads ViennaRNA bot fail, InfoRNA bot pass - 00:09
Crossroads II ViennaRNA bot fail, InfoRNA bot pass - 00:05

Furthermore I wonder whether the long solving time for “Up, up and away” - 09:01, could be caused by the size of the big loop?

InfoRNA bot is programmed to send requests to InfoRNA and receive results through network. Thus, the duration to solve a puzzle can be inaccurate since the time to receive results might delayed not because InfoRNA was taking too long, but because some network issues were delaying it (i.e. busy traffic). Generally, differences in some minutes are not that important since it might not be accurate for InfoRNA bot. We will launch another bot soon with accurate timer, so stay tuned!