Your engine is very great opponent for human playing and training.
Please, add a parameter for risky playing versus human opponent, for training tactical skills.
My idea for this parameter.
Humans can search about 3-8 moves down and narrow.
Engine can search more deep and wide.
Engine can detect situation for wide and deep tree of variants, and multiply self positive score progressively.
In search algorithm engine just multiply score by K * deep * wide for positive score, but not multiply negative score (!).
In that manner of score, tight variants for human wold be scored very higher, and engine will be play very risky and gracefully.
Your engine is very great opponent for human playing and training.
Please, add a parameter for risky playing versus human opponent, for training tactical skills.
My idea for this parameter.
Humans can search about 3-8 moves down and narrow.
Engine can search more deep and wide.
Engine can detect situation for wide and deep tree of variants, and multiply self positive score progressively.
In search algorithm engine just multiply score by K * deep * wide for positive score, but not multiply negative score (!).
In that manner of score, tight variants for human wold be scored very higher, and engine will be play very risky and gracefully.