Would you play well known Opening theory, or avoid it? What other strategies would you employ to beat a GM in a simultaneous? I have my own strategies which I've used. See link in my profile. But what is best?
I've answered a very similar question here
My take is that GM try to avoid calculating, and try to rely on their superior positional understanding. So striving for a tactical position, seems like a good bet. Your game seems to support this hypothesis.
Here is interesting story, that involved somebody I used to know over the internet and Kasparov:
[Kasparov] rather nervously asked me if I had a rating. I replied 2084 and he, not looking very happy, asked why I didn't write it down in front of the board next to my name. I replied that no one asked me too and that the Belzberg people knew my rating and only today OK my playing since I am a client. He then stated that if he had known I was rated that highly he would have played differently rather than getting into a very theoretical line. Being aware of the 2000 cutoff I again appologized for the misunderstanding and we quickly realized that the Belzberg staff had erred in not telling people to write down their ratings (these were different people than from the NYC event). I simply stated "I'm sorry for the confusion, I won't play" and started to get up and Garry, reiterating that he'd have played differently vs a rated expert, started resetting the pieces to the starting position and said we can restart. Being rather shocked by this and what was over a minute of commotion and confusion I knew the gentlemanly thing for me to do was to step away and let someone else under 2000 play.
The position was a very sharp double-edged Scandinavian. Something even Kasparov wouldn't want to play against a high-rated player.
A GM that had many simultaneous games experience is the best man to answer your question. But as a chess fan, this is my idea. The best strategy is playing your best game. Play the opening which you know it thoroughly.
Usually GMs avoid positional variants and they try to play tactical and open games in simultaneous matches. I heard from one GM, they force the weaker opponent make a blunder to lose one or two pawns then they simplify the game to reach to an endgame.
So, you can make the game closed and play positional strategies. Avoid simplifying the game and also don't forget a GM always can find a tactical combination just in a few seconds.
Its hard to say how to prepare because you can only play to your strengths and what you really know.Just because Kasparov will find difficulty against the Berlin doesn't mean you should play it against him if you don't understand the subtle nuances yourself.However having said that I've always felt that certain grandmasters are more at home in certain positions than others and it pays to prepare for them as if it was a 1 v 1 game.For e.g When Nigel Short came over to Sydney to give a lecture and play in a simul I did some quick analysis of his games and he doesn't seem to do really well against strong aggressive tactical players like topalov, kasparov, polgar, shirov etc(Relatively speaking of course). He is however a midlegame positional maestro and his feel for shall we say taking care of the pieces (esp the king)in the middlegame are insanely good. So if he had played e4 a sveshnikov would have been played and if he played any other opening other than e4 I would have played a KID set up which in the end was successful but only for a draw.To be honest players at my level only have a good forehand shall we say and any deviation will be inferior for e.g I have played KID and Svesh for the better part of 10years and only 2 years ago have started incorporating e5 berlin/QGD Ragozin setups.It also depends on your own rating (mine at the time was low 2000s)for a 1500 I would recommend just treat it as 1v1 and play to your strengths and incidentally Short did draw to 2! 1500-1600 players in a 30 game simul.He won quite convincingly btw 28.5-1.5(3 draws) and completely routed North Sydney.The point I am trying to make is you have to encapsulate all your experience and knowledge into that one game so really try to play to your strengths as opposed to your opponents weaknesses.I was just lucky that my style matched well to Shorts'. (BTW he is quite the charismatic character)
Some psychologists hypothesized that grandmasters were better at recognizing chess positions that other people. When they ran the experiments, they were proven "half" right.
That is, grandmasters were better at recognizing standard chess positions, because of greater practice, and playing them all the time. On the other hand, they were no better than amateurs at recreating random "chess" positions with pieces dumped randomly on the chess board in defiance of chess logic: e.g. White King at e4, Black king at a1, White pawn at h6, Black pawn "behind" it at h4, etc.
So a trick for an amateur is to steer the game into "unconventional" lines that the grandmaster may not be familiar with, and then use the extra time to "think through" these unusual positions, hopefully better than the grandmaster.
To take one example, playing as Black, Queen's gambit accepted is not as well understood as Queen's gambit declined. This is particularly true if the grandmaster is not Russian. The Russians have produced a whole set of literature on Slav defenses and other variations of QGA, meaning that if you master the literature better than the grandmaster, you may have a chance.