Everyone seems to have completely misread what I said, even after I explained myself. Uncommon lines are not always inferior. Uncommon != unsound. I would also make the argument that playing main line openings is going to play directly into a GM's strengths because a GM is far more likely to have studied and played through every main line position in greater detail than you ever would. I absolutely agree that you should play something that you're comfortable with, but if you're comfortable with lines that your opponent has had a less chance of being prepared for, I'd say that's a plus.
What Defense Would Slow A Grandmaster?
Everyone seems to have completely misread what I said, even after I explained myself. Uncommon lines are not always inferior. Uncommon != unsound. I would also make the argument that playing main line openings is going to play directly into a GM's strengths because a GM is far more likely to have studied and played through every main line position in greater detail than you ever would. I absolutely agree that you should play something that you're comfortable with, but if you're comfortable with lines that your opponent has had a less chance of being prepared for, I'd say that's a plus.
True, but there aren't all that many uncommon lines that perform well for you, particularly with black, and with either colour, these openings are more common than other lines. I could see that if you were very comfortable with, say, the Polish, that you might have some pretty good chances of surviving for quite a while by playing this.

Everyone seems to have completely misread what I said, even after I explained myself. Uncommon lines are not always inferior. Uncommon != unsound. I would also make the argument that playing main line openings is going to play directly into a GM's strengths because a GM is far more likely to have studied and played through every main line position in greater detail than you ever would. I absolutely agree that you should play something that you're comfortable with, but if you're comfortable with lines that your opponent has had a less chance of being prepared for, I'd say that's a plus.
On the contrary, you have a complete misconception of how a GM or any high-ranked player thinks! A GM/IM/NM has a far more deeper understanding of a position than an ordinary player. When an ordinary player follows accepted lines, he is mimicking a person with deep positional understanding. A GM does not need to "study" except against players of equal strength! When an ordinary player tries an uncommon line, he/she has no clue whether it is sound or not! If per chance the line is sound, no problem! If not, he/she can kiss the game goodbye!
The question is, have the majority of the "ordinary" players enough positional understanding to branch into uncommon variation which are not unsound?

An uncommon line is ok against a GM, provided you're reasonably familiar with it, but an unsound line is not...

I've always thought that the best way to defeat a stronger opponent is to play the sharpest opening you're familiar with and hope you get lucky. Even Super GM's regularly lose their way in sharp positions. I'd roll the dice and aim for razor sharp openings like the Perenyi Attack, Hamppe-Allgaier or Pierce Gambit.
actually an extremely complicated position would be a better chance
Yes, provided you have previously analysed those positions beforehand - is it a joke to assume you will do better analysing an extremely complicated position OTB than a master-level player? That's what GMs do in tournaments - keep some surprise lines up their sleeves to floor their opponents...
I've always thought that the best way to defeat a stronger opponent is to play the sharpest opening you're familiar with and hope you get lucky. Even Super GM's regularly lose their way in sharp positions. I'd roll the dice and aim for razor sharp openings like the Perenyi Attack, Hamppe-Allgaier or Pierce Gambit.
actually an extremely complicated position would be a better chance
Yes, provided you have previously analysed those positions beforehand - is it a joke to assume you will do better analysing an extremely complicated position OTB than a master-level player? That's what GMs do in tournaments - keep some surprise lines up their sleeves to floor their opponents...
If you want a win and are much lower-rated, I'd say that it's your best shot even if you haven't analysed them extensively; if you just guess, you have a small chance of coming out ahead. A really really really really small chance. But I think that the idea was that these type of positions offer better chances rather than positions which you know comparably well

Actually, if you're looking for strategy against a much higher rated player, it would be better to play off lines of openings to catch them off guard, force them to think and give them a chance to make a mistake. If you just play the main lines, you're playing into positions that they have most likely seen before and could play without effort.
The main difference between GMs and lower-rated players is their ability to assess a position quickly and focus on the most important factors in creating a plan. Choosing off-lines plays into their strengths.
Against GMs you should be playing for positions with which you are familiar and comfortable, to maximize the chances of finding your way through the position. In this way you will also learn more about your chosen opening, which will only help you in future games.
The reason I would play the hyperaccelerated dragon against 1.e4
A nimzo, benko or kid against 1.d4
a kings english against 1.c4 and
as white
1. e4 with way to many openings to name that I will be able to play.

I've always thought that the best way to defeat a stronger opponent is to play the sharpest opening you're familiar with and hope you get lucky. Even Super GM's regularly lose their way in sharp positions. I'd roll the dice and aim for razor sharp openings like the Perenyi Attack, Hamppe-Allgaier or Pierce Gambit.
actually an extremely complicated position would be a better chance
A complicated opening is a death sentence. You're not going to out analyze a GM and you're never going to understand the position as well.
Lol, then it's pretty much impossible to win or draw against a GM. Because to get lucky you would have to out analyze them. I personally think the better chance (if a draw is ok) is to see if they can outplay you from a solid position. I mean, the chances of holding a draw despite being outplayed seem higher than getting instantly mated from your second best move at any moment. And I remember you saying the game would be horrible. Why?! I think I could appreciate how a GM outplays you positionally in a seemingly equal position! I would learn quite a bit in fact!! I think what you learn is about equal: either a long combination or many good positional moves and ideas. I would appreciate a good combination also since I'm a well rounded player but in any case we should have a full appreciation of chess so now that I think about it no, I don't think it would be so bad to slowly lose to a GM in a solid position. What is there to feel bad about? He's a GM! So you could only get a positive benefit either way. So my decision will be purely based on the odds, and right now caro kann is most reliable since it is 99.99999% likely even if you know the opening just as well as the GM to misstep instead of him, or, if he makes the error, you're likely to make one back anyways. With the caro kann I would give myself a few more percents in chances and hold on for dear life.

I've always thought that the best way to defeat a stronger opponent is to play the sharpest opening you're familiar with and hope you get lucky. Even Super GM's regularly lose their way in sharp positions. I'd roll the dice and aim for razor sharp openings like the Perenyi Attack, Hamppe-Allgaier or Pierce Gambit.
Then again, if a GM plays the Scheveningen as black, he's almost certain to be prepared to play against the Perenyi or the Keres attack. He probably wouldn't be too booked up on the Hampe-Allgaier or Pierce Gambits, but they're less sound than the Perenyi.
Still, this approach would be my best chance. Against 1. d4 I'd play for the Benoni and against 1. e4, play the Najdorf and hope to create some benefical complications. What more can you do?

yes if i were 2 play a gm i wouuld definitely play the modern benoni and launch an attack on the qs. i never played a gm b4 but i was able 2 beat chessmaster 6000 with the benoni. now if he played e4 i wouldnt know what 2 play probably the french since its a solid defence but he would probably break right through it maybe the caro khan but i would have 2 study it and prepare thouroughly

Honestly I don't think there is a clear answer to the question. I think solid and sharp are both fine, about equally unlikely in giving results lol!
If I were to play a GM I would play to win matter of fact I'm not the best chess player but I've taken a lott of losses to become a good deffensive player so I say experience is the best teacher.

I think that the premise of the OP points the club player in the wrong direction. Instead of picking an opening to maximize chances of a win, use an opening that you often use against players of your own level. Then, win or lose, you may learn something relevant to your development as a chess player, instead of having an isolated experience that has no bearing on how you usually play.

Yes, I very much agree with Pegrin's post above. It's not about playing the Dragon or the Caro-Kann. What matters is :
- you play something you're familiar with (so at least you can expect to give your best and learn something, which is very important !)
- you play something sound (ie. main lines don't lead you to a clearly inferior position)
Actually, pegrin's post is based on a wrong presumption. The OP wasn't talking at all about maximising the chances of a win, he's concerned with holding out the longest.
My position on the matter is that you should play your best chess - i.e. to play the best you can, playing whatever opening you think is the best (or that you can play the best). This is what you should do every game. Thus, you will be playing what you're most comfortable with.
Most of the rest of the discussion would be beneficial, in my mind, for someone who is extremely comfortable with many openings.

I've always thought that the best way to defeat a stronger opponent is to play the sharpest opening you're familiar with and hope you get lucky. Even Super GM's regularly lose their way in sharp positions. I'd roll the dice and aim for razor sharp openings like the Perenyi Attack, Hamppe-Allgaier or Pierce Gambit.
Then again, if a GM plays the Scheveningen as black, he's almost certain to be prepared to play against the Perenyi or the Keres attack. He probably wouldn't be too booked up on the Hampe-Allgaier or Pierce Gambits, but they're less sound than the Perenyi.
Still, this approach would be my best chance. Against 1. d4 I'd play for the Benoni and against 1. e4, play the Najdorf and hope to create some benefical complications. What more can you do?
I agree that that a GM is likely to be familiar with the Perenyi if they play the Scheveningen but black's position is so incredibly difficult to play for 3/4's of the game -- even when you're booked up and playing CC -- that I thought it might make a good choice. It would definitely be a fun one to try even if it ended brutally.
Concerning the Hamppe-Allgaier, I came very, very close to beating Becerra with it so I wouldn't hesitate to play it again against another GM. So, so close....ugh.
I checked out your game against Becerra with the Hamppe-Allgaier, very cut-throat! There seemed to be one point towards the end where your initiative stalled slightly, and his queen was unleashed, particularly along the h1-a8 diagonal. Do you have any ideas for improvements in retrospect?

Actually, pegrin's post is based on a wrong presumption. The OP wasn't talking at all about maximising the chances of a win, he's concerned with holding out the longest.
My mistake. Holding out a long time is even less attractive as a goal, in my opinion. At least if you get a win, that's a great memory and story. Going on for 80 moves because it takes a long time for the GM to convert an advantage and you refuse to resign.... Well, that's not for me.
Most uncommon lines are inferior, at GM level.
Respectfully I disagree. Most uncommon lines fall out of favor only until a TN or refutation is found, that's all.
Then everyone's playing them again.
Depends on how uncommon the uncommon line is. You don't see GMs playing the Grob, really ever