the agents increase the average to around 1/2 and by game 45 or so they seem all seem to agree on jumping 13 times for every 15 possible jumps.
This is much higher than we would have expected, indicating that jumping over small obstacles, rather than moving around them, may be a better idea than we thought.
Such a result is a good illustration of the benefits of evolutionary algorithms: the agents had evolved a behavior that might be
better than what we would have hard-coded.