shredder, fire, stockfish are all machines... they're unpredictable.... there's no "real deal"
Is shredder the real deal???


AI is the real seal of the deal, I feal.
?
Then again, maybe not.

Held stockfish solidly as black
Sure he might have been worse but he was black vs a winner of many championships
The nightmare of Alpha Beta search engines(stockfish ) exposed!
Obviously stockfish could not detect the fortress of black pawn structure. There is only one weakness in black position which is f7, however there is no way for white to force capture that pawn.
1. From 45 to 75, Stockfish was stacked with +2.00 evaluation for 30 moves without progress and suddenly dropped the evaluation to 0.00 when 50 moves rule kick in.
2. However shredder has correct evaluation +0.60 somehow, ??? probably due to pawn structure evaluation function!!
If stockfish would have seen these fortress ahead(know the evaluation function in fortress ), Stockfish wouln't probably played ???23. Rd4 which allow 23... Rxc3!! that forced lead to that endgame. Strangely, after 23....... Rxc3!! Streddeder evaluation dropped and stockfish buggy evaluation up(not knowing fortress).

So that makes Shredder smarter than Stockfish!
Of course, in that particular evaluation function, Shredder> Stockfish. There are tons of evaluation function codes and Stockfish may have better in other positions.
However overall elo is the result of search + evaluation function.
If you have more complicated( more knowledgable) evaluation function code, your program will run slower, and likely to perform worse in general.
Laury Kaufman said, Komodo evaluation is much ,much better (complicated/ knowledgable) than stockfish. But in rating/ game play, stockfish search deeper and faster ,and get better rating.
Better knowledge doesn't mean better player, but the overall result is search+ evaluation, how fast you can search and how fast you can evaluate those positions make a better player/ better rating.
It beat Fire while the great Stock fish couldn't