I suppose the huge majority of all chess engine research is focused at bumping the silicon monsters yet another ELO point. Why is it more interesting to buy a new version that is five points stronger than the six month old version you already own than to buy a software that plays at a very human like 1500-level?
As a DIY chess improver working full time and eager to spend time with the kids, it is hard to find time for a slow game. Playing against software when a gap suddenly appears in the overloaded time schedule is often the only alternative and it is a nice kid surprise insurance to have the pause button available.
The problem is that most engines are totally useless to mimic patzer chess. The mastermind programming to produce an Average-Joe Chess style seems to be to play at top strength and to mix it up by moving Ra1-Rb1 for no obvious reasons at random time points.
Which engines would you consider to be most humanlike when playing at lower levels?
2nd Place tie in the Mid-America Open U1900
9 hours ago