First game was played at TC 2' + 2'' for the engine, and 5 times as much for me, with Komodo using contempt(hence the strange evaluations, Larry has always claimed Komodo plays better against humans with contempt, so we tried that out, and I alloted myself more time to avoid any unpleasant tactics).
The second game was played at fully equal TC, 5' + 3' for both opponents, with Stockfish pondering, so you see it is possible to beat the tops even with fully equal conditions.
I don't believe you beat stockfish in a classical game, and anyone with a brain will agree. Engines are stronger than players and can see a tree of 30 moves ahead and can calculate millions of moves/second. If you had that calculating power and speed in your brain you were in the top 10.
I have never played classical, LTC games against the top engines.
Most of my wins are blitz and rapid, with myself having more time, and the engine pondering.
You should not forget that I am computer chess expert, and know almost by heart all top engine specificities and behaviour on the board.
When one has followed every single game of TCEC, and analysed for long hours daily all possible types of positions with the top engines, one certainly gets the knack of it. It is a matter of experience.
Below one game from the second part of 'Human versus Machine':
As you see, winning is possible even in not fully closed positions, Komodo simply made some inaccuracies.
And one more, this time against an earlier version of Stockfish, just to show that my style of play is unmatched by anyone or anything, and all fake stuff allegations are fully unfounded.