Can someone play without openings?

Here's what I've found:

"I go through games on a database, or more often read a book about them, and I’m really not looking for anything. I just enjoy the games, I enjoy the chess and then hopefully I can learn something from them."— Carlsen, 2016

"At a time when all players prepare themselves with software, my goal is not to see if my computer is better than my opponent's. In the openings, I just need to reach a position that gives me play. The idea is to be smart rather than trying to crush the other.

I try to figure out where he wants to take me and I do my best to not put myself in positions where I could fall into his preparation.

I try to play 40 or 50 good moves, and I challenge my opponent to do as much. Even if the position is simple and seems simple, I try to stay focused and creative, to find opportunities that lie within. Not to play it safe. It is important to know how to adapt to all situations.

In this sense, I have that in common with Karpov in his heyday: he believed deeply in his abilities, he was very combative and won a lot of games in tournaments because even when he was not in a good position, he felt he could still win, and played all the way. I'm somewhat similar in spirit: during a competition, I always believe in myself." Carlsen, 2014

To me, Carlsen sounds like he doesn't worry too much about opening preparation. He glances at games to see what he can pick up from them, but doesn't sound like he sits down to study or memorize variations.

Rather, he trusts in his ability to figure things out over the board.

That's what I gather from his comments, anyway. Your mileage may vary.

Those are really helpful quotes from King Magnus.  Thanks for posting them, Slow Move.

I really like his approach to Opening Theory.

Sure thing. I like his approach, too. Especially, "In the openings, I just need to reach a position that gives me play."

The problem is that chess has three phases. The opening, middlegame and the endgame. 

No remove the opening phase. 

Tactics that one can realistically expect to be able to work out over the board?

Is that the same as not studying openings at all?

"... In the openings, I just need to reach a position that gives me play. ..."

@SlowMove Carlsen prepares more than you think. Carlsen's second Peter Heine Nielsen helps him a lot with opening preparation. Just look at the game Aravindh-Carlsen where Carlsen followed the exact same line which Heine Nielsen analyzed in an opening article he wrote in a book. Even Jobava prepares for his games. Also I think that studying theory can make you creative because you see how other people have played and you can ask yourself: Why not this move? Why is this considered to be bad? I have a better idea and so on and so on.

If you just memorize an opening setup like the London System and then repeat the same lame moves in everygame you will never find new or interesting moves. Your games will always be the same old stuff.

Just look at how Carlsen's games. Even if he plays sidelines he still knows the theory but is able to  findsnew ideas like Na7 in the modern defense.

He does prepare even in sidelines. In one interview he mentions the book of Tiger Hillarp Perrsson about the modern defense. He followed Tiger's analysis and equalized easily.

It's funny how so many people think that using sidelines means you don't have to put any effort into your opening play.

I've answered this question of yours already.

Also: Do you provide opinions of your own, or do you only question the opinions of others?

Perhaps he does prepare more than I'm aware of. I'm not familiar with his training regimen—I'm only going off what's been said in interviews.

Though I found this bit about him interesting, from The New Yorker a few years ago:

"In the lead-up to tournaments, when other players are testing out strategies on their computers, Carlsen is often staying up late playing video games or online poker. Before tournament days, he likes to get plenty of sleep—optimally, ten or eleven hours—waking up an hour or two before the start. 'It’s no secret that the best players’ opening preparation is much deeper than mine,' Carlsen told me. In London, he went into some games with only the first move chosen . . ."— The Prince's Gambit, D.T. Max

I do like your point about looking at theory and asking yourself questions about it. I can certainly see the learning potential there.

I mean Carlsen is a GM so of course he forgets more theory than we'll ever know lol. But still there are plenty of shortcuts and easy ways to just get out of the opening without something terrible happening.

If someone again declares that tactics are the reason for past ideas no longer working, it seems apppropriate to me to again raise the question of whether or not it is realistic to expect such tactics to be worked out over the board. (That sentence, by the way, is an opinion of my own.)

Portisch advocated that sort of thing in a contribution to a 1974 book about how to play the opening. The Portisch chapter is about 40 pages and concludes with, "... I urge the reader to do his own analysis in the development of an opening repertoire."

@SlowMove The book i was reffering to was Experts on anti-sicilian where his seconde Heine Nielsen suggests a line to fight for an advantage against 1.e4 c5 2. b3 from black's point of view. Carlsen followed the analysis in this book. He does prepare but maybe not so intensively like Caruana but there is still some effort going into his preparation. By playing Colle or London System he of course avoids heave theoretical discussions in the nimzo or KID.

But whenever he plays he alsways gives the london system a different touch. Even in sidelines he finds unconventional but good moves and doesn't play autopliot moves like most london system players.

Here is a nice game against Tomashevsky:

Notice the following move: 13: Bxg6:! followed by Ne5! What a move! While every typical london system player would have played h3, Bh2 and Ne5 followed by f4 (because he blindly follows "theory") by now Carlsen decides to give up his lightsquared!!! bishop to cement his knight on e5. 

Since fxg6 isn't advisable because of the weak pawn on e6, Tomashevsky played hxg6. After Ne5 black has a problem because the white knight will dominate the position because it will stay their for a long time (Notice how many important squares it controls).

Black can't play f6 to drive the knight away because the pawn on g6 is hanging. Without the doubled pawn black would have the possibility to free his position with f6. The knight on e5 ended up being one of the deciding factors of the game. Tomashevsky gave up immediately after Nc4.

What I'm trying to say is that playing the london system or other "automatic" setups (like grand-prix attack) can be dangerous because weaker players just bang out the moves without thinking. The position might be playable afterwards but will offer nothing new and fresh. I think playing like that kills your creativity. Looking at theory moves at a later stage of your chess learning process (not at master level) can benefit you because you get new ideas. By looking at theory moves we learn from experiences of other people. We can either follow them if we undestand their moves or find other and possibly better ways. Just playing one-setup to avoid theory will limit your mind to only what you know. 

I guarantee you absolutely nobody on this world would even consider giving up his lightsquared bishop like Carlsen did on move 13. I sometimes have the impression that all those london system books say: Play d4, Bf4, Nf3, e3, c3, Bd3, Nbd2, h3, Bh2, Ne5 and know the game begins.

That's just wrong because the game starts from move 1 even in the london system. It's funny how the people who criticize other people for studying theory are the ones who actually don't understand their london system (grand-prix, alapin whatever) moves either. They do the same moves again and again without understanding why these moves are played. The london system can offer false safety in the opening and is a threat to their chess understanding. 

Note: I'm not talking about strong players who have already a good chess understanding and will be able to handle the london system in a non automatic way. I'm talking about lower levels where even trainers recommend the london system as a way to avoid theory and get a playable position.

Learning the london system alone isn't enough and I think it is not as tactical like e4-openings but it is a strategically complex opening which mgiht be difficult for beginners to understand.

There are some subtleties one only undrestand if you study strategic topics.

Whether or not such tactics will be worked out over the board depends entirely on the players. Either way, it's still a matter of tactics.

Opening play doesn't exist in a vacuum; it's not one-shot deal. A player's opening ability is constantly evolving, due to practice and experience. If one misplays an opening, it's not reason enough to point a finger and say, "Ah-hah! See? This proves that this player needs to sit down and study openings! He wasn't able to work out the tactics over the board!"

Except, losses can be learned from. The player can review their loss. They can identify the mistakes they made (likely tactical in nature). They can recognize the faulty logic that guided their mistakes, discover improvements, and remember the lessons learned.

In this way, the player best internalizes the ideas, as they've been honed through personal experience, rather than absorbed from second-hand exposition.

I agree that Carlsen likely does prepare, though I also agree that it's probably not very intensive at all.

He's at such a high level of chess comprehension that merely glancing a line once is likely all he needs to grasp the inherent plans and tactical ideas.

The reason he seems so fluid with his handling of the London is (in my opinion) due to a lack of intensive study—he seems more interested in experimenting with new ideas on the board, rather than relying on home prep.

And I agree with you that robotically parroting an opening system isn't the best way to approach it—by doing so, the player might miss positional or tactical opportunities that may arise, simply because they're too busy trying to "stick with" the same system or structure, regardless of what the opponent is doing.

my chess teacher states that shakriyar mamedyarov reached to 2700 with almost no  special opening knowledge, but also emphasizes that man has solved so much tactics that when it comes to calculation  nearly a little bit weaker version of stockfish.

he puts him above kramnik in calculation but believes that kramnik is much better on overall strategy and positional nuances.

If the "tactics" are not something that a player can be realistically expected to work out over the board, then, is it necessarily a mistake to try an opening book?

"Likely", but not definitely, and apparently not necessarily something that the player can be realistically expected to work out over the board.

Can this be realistically expected for all players and all opening difficulties?