It's relatively simple.
Stockfish on lower levels are programmed to make mistakes. The higher the level, the less bad moves it plays in a game.
If you play Stockfish on level 6, it statistically makes 1-3 inaccuracys in the first 10-15 moves and by some relatively low percentage 1-2 blunders.
If you play for example the same 5 opening moves 100 times, chances are high, that in 1 or 2 games Stockfish blunders something very early.
Its just simple statistics. I showed that some time ago with a scolars mate on a lower level of Stockfish...
lichess.org/forum/game-analysis/i-won-against-the-computer-level-2-twice?page=3#24
Notice, that I played it anonymously in a correspondence-mode, so I could easily abort the game, if the comp didn't fell for the trap... I guess I tried it about 30 times, before it worked.
Edit: I just watched the game history of that person with the "impressive" easy mate in 4: I stopped counting after Stockfish was in lead of 30 wins by 2 losses...
Have fun!
Stockfish on lower levels are programmed to make mistakes. The higher the level, the less bad moves it plays in a game.
If you play Stockfish on level 6, it statistically makes 1-3 inaccuracys in the first 10-15 moves and by some relatively low percentage 1-2 blunders.
If you play for example the same 5 opening moves 100 times, chances are high, that in 1 or 2 games Stockfish blunders something very early.
Its just simple statistics. I showed that some time ago with a scolars mate on a lower level of Stockfish...
lichess.org/forum/game-analysis/i-won-against-the-computer-level-2-twice?page=3#24
Notice, that I played it anonymously in a correspondence-mode, so I could easily abort the game, if the comp didn't fell for the trap... I guess I tried it about 30 times, before it worked.
Edit: I just watched the game history of that person with the "impressive" easy mate in 4: I stopped counting after Stockfish was in lead of 30 wins by 2 losses...
Have fun!