Imagine you're playing vs someone who would win 100 out of 100 on a level playing field and he's a competent blindfold player.

What is your best strategy to beat him when he's blindfold?

Things I'm assuming:

-Maybe play a more obscure opening but keep it fairly mainline since he has a skill advantage
-Keep a lot of pieces on the board
-Maybe fianchetto to set up some overlooked attack once the centre of the board clears up

Thoughts?

Keep pieces on the board and keep it messy for sure. You probably aren't good enough to judge what's an "obscure opening" to someone 600 points higher, so you are at risk of FPS if you try that. But you want to prioritise messiness and tactical opportunities. Poisoned pawn Najdorf, good. Main line Benko, bad. You'll most likely be worse out of the opening either way, but you have a good chance to out calculate him after that.

Keep pieces on the board, but I'm not sure with the advice to make it "messy".

Avoid simplicity, sure, but proactively trying to make a position more complex without making weakening moves is difficult. If you're playing someone that much better than you are, your idea of "messy" moves are probably just bad moves.

Steal his rook.

Im 2050ish, completely incapable of beating even a 500 blindfolded. I have no idea how anyone does this, I also never memorized where like "c4" is when using the Black peices

Having done some blindfold simuls in the past I'd say just play the best moves and just rely that his moves will be weaker than normal. You can aim for a position that isn't "messy" (whatever that means) but one that maintains tension; having pawn faceoffs where you have to consider every move the impacts of a trade.

However, if you want blindfold tricks, if you see a bunch of moves being relatively similar, consider making a series of the same bishop moving on the same diagonal, or shuffle two rooks back and forth; the more the same piece moves, the more ghost memories the blindfold player has of it and the more likely he is to lose track of it.

Play openings like 1.e4 d5 2.exd5 Qxd5 3. Nc3 Qd8 or 1.e4 Nf6 2. e5 Ng8. Carlsen beat Caruana and Fressinet with these lines. The idea is to get out of theory immediately and leave white without any obvious plan. Other than that, just try to make good moves.

Wouldn't you go Qh5 rather than Qd8?

4.Qxh5 1-0

Qd8!

The point is not to make the best move, but the one that allows the least amount of forced variations. It's like a mini-gambit. You sacrifice something like 0.20 in computer pawns in order to get your opponent out of book and also out of his stereotypes.

Sorry I meant a5 not h5

this is a nice point. So it doesn't matter at all that I've nothing developed at this point and the guy I'm playing has the total control of the centre and a very easy board to remember?

Qd8 is actually a line, probably as good as Qa5 or Qd6 IMO

Also I doubt that fianchettoing the bishop just for the sake of it wont really help, when it's in a fianchetto it is obvious where the bishop is and hard to keep track of. Also if you're familiar with the concept of Chunking as a memory tool then you'll realise that the fianchetto confuguration is one that is easy to remember. I'm not saying don't fianchetto, just do it for a reason other than long diagonal tricks

Sorry, but you've got to ask Carlsen on that one. He played it despite the fact that Caruana could clearly see the entire board, so there must be something going for this approach. There is no clear-cut plan for white and that counts too. If the plan doesn't exist, it doesn't matter if you can look at the board or not.

http://www.chessgames.com/perl/chessgame?gid=1768345

I am not claiming that this is a great opening or even a good one. Kramnik is playing 1.d4, 2. Bf4 to get people out of book lately. In the age of forced draws, this is simply a new philosophy.

Taking advantage of chunking is a good example of what I referred to. For example, let's say you fianchettod with black, then played Nh5-Bf6-Ng7. That's a very unusual piece configuration within a known chunk, and more likely to be messed up or forgotten.

However I stand by my original assertion - just play your best.

and a little chuckle at any strong player being confused or out of book after Qd8 in the scandanavian.

what are the exact ratings?

A 1400 vs. 2000 and a 1800 vs. 2400 would require different strategies imo. I also think the higher the ratings, the less of a shot you have assuming villain can beat you 100/100 easily. Even blindfolded, I don't like your chances that much.

Maybe do research on villain's games and styles. Play some unorthodox opening and aim for a style that puts villain in unfamiliar or uncomfortable territory (something villain sees and can't intuit).

But yeah, generally go out of theory (eg. not some main line opening) OR play an opening villain is unfamiliar with (but also will not hinder your play significantly).

Another thought is keep pieces spread out and active on both flanks of the board if possible. I'm not saying get out of your way to do this, but if you can have Bc4, Bg5, and Qf3 for instance, it might be more confusing to play against than when you have Bd3, Bd2, and Qe2.

As to open or closed positions, I'm not sure. I think a position that is dynamic and has a lot of stuff going on would be confusing for the blindfold. BUT if villain can easily spot tactics blindfolded then my guess would be to keep the position closed. A closed position would result in a longer game since I think mistakes are less costly. A mistake in an open position usually results in loss of material whereas a mistake in a closed position is usually positional and harder for villain to completely realize.

So yeah... if villain is a tactical attacker, trading is def in your favor. If villain is a positional player, I'd keep as many pieces on the board as I can. If you dunno, my default would be to keep pieces on both flanks (assuming I'm comfortable with that and wouldn't blunder easily in those type of positions).

If you dunno, maybe checkout villain's opening, middle, and endgames online. Strive to gain an edge (or in this case, lose the least edge) by figuring out which part of the game villain struggles the most in.

If villain draws a lot in the endgame, trade! If villain is a middle game monster, trade! If villain grinds out wins very technically, DON'T Trade. Think of it like poker I guess.

I think the BEST HOPE of winning (however small it may be) is if villain blunders. It's almost impossible to outplay someone who can beat you 100/100. I'd heavily rely on cheap trickery.

I'm around an 1100 blitz player on chess24 and he's around 1900.

He's also got a standard rating of 1657 and rapid rating of 1694 (live chess)

I also have this information about his play:

When he was younger he used to lose games to a blunder and as a result he has trained himself to be very cautious. In the times that I've played him I've expected to be blown out of the water, but he usually just has a couple of pawns advantage at the endgame (which is obviously enough)

so this means DON'T Trade?

A player with a 1650 regular chess rating will blunder. Just play normally.

OBV

Okay, so 1900 blitz really isn't that strong. If we were talking about 1600 vs. 2400, my advice would be to just play your normal game. But 1100 vs. 1900, your best chance might be to randomize the position with a weird set-up, hoping that he will forget the position of some piece or pawn. Something like the https://en.wikipedia.org/wiki/Hippopotamus_Defence (if you're White just play 1.d3, 2.g3 etc.).

Given your difference in skill levels, if you start an open fight in the opening, it's likely that you will make some gross strategic blunder that your opponent will be able to exploit and get a dominant position very quickly. Your best bet is to play a closed set up and just try to stay in the game as long as possible hoping for a mistake.

hmm interesting

I will definitely let everyone know what happened when this takes place

i don't know if it's cheating but i would try to improve from 1100 to 1400 before the game. 1100 is really bad, you are almost guaranteed to blunder early.

How should I most effectively improve?

I mean since I'm going to improve then why not improve to 2000 so that I'm better than him...

1100->1400 is something that can be achieved by most people by doing a trivial amount of tactical puzzles/studying/learning opening strategies - it's really not much for a thinking adult to get to

2000 is unrealistic