I have occasions where I disagree with the analysis engine in reviewing games. it seems it has a real dislike for sacrifices. Ie temporarily giving up material to get advantage several moves ahead. I can see that a perfect defense....might make it less than best option.
recently, it said best move for an opponent was taking a rook off the 8th rank, to capture a piece of mine, which was protecting his king from checkmate. That's exactly what he did....and immediate checkmate. there was another series of moves that could have extended the game , still ultimately ending in checkmate, if he made a different move.
do others feel that's common with the analysis?
A *lot* of engines place far more value on material gain than on positional advantage. Always take the engine-suggested move with a grain of salt unless there is an immediate demonstrable tactical advantage, especially if you aren't running the engine at full power.