Should you always play like a grandmaster no matter how (un)skilled your opponent is?

I thought of a situation where intentionally not playing to the best of your abilities is a good strategy. Like, let's say you're playing against a chess noob. You don't need to waste moves defending your army because your opponent probably won't even notice the hanging pieces. You can checkmate your opponent faster by doing things that're usually considered bad strategies, such as moving your queen out too early in the game.

Additionally, I discovered in the Chess.com mobile app that its "best move" hints change relative to the computer opponent's level. Seems like you should play against newbies differently than how you play against grandmasters.

Is purposely "not trying your best" sometimes advantageous?

  • 5
    Does a chess grandmaster always play chess at a grandmaster level? Probably not, unless the person is a jerk and beats kids all day with his/her grandmaster skills – Huangism Aug 16 at 14:19
  • Is your goal to have fun with your kids/friends or to win in the quickest way possible, while still ensuring that you do win? For your kids/noob friends to have fun you might consider playing quickly (challenge yourself as to how fast you can make a move, hopefully still good, rather than how extremely thorough you can be with your calculations). – Cullub Aug 16 at 16:38
  • 4
    Your question is ambiguous. What is your goal? If it is to maximize your odds of winning, you should always play your best. If it is to minimize the total game time while maintaining a 99% chance to win, this likely requires a different strategy. If it is to maximize your opponent's enjoyment while maintaining a 99% chance to win, this would be yet another strategy. – MooseBoys Aug 16 at 19:06
  • This question is more interesting in Go than in chess. In Go, there's a handicap system: a strong player who plays handicap Go against a weaker player starts from a losing position and has to claw their way back over the course of the game. In that context, it can make strategic sense to play moves you know would be bad moves against someone who was your equal - overly greedy, aggressive moves you know can be devastatingly countered - and hope your opponent is too weak to spot the correct reply. (These kind of plays even have a name in Go culture - they're called "dishonest" moves.) – Mark Amery Aug 17 at 10:40
  • The hint generator and the AI move generator are basically the same thing. The app is probably just conserving battery power by not searching for the best moves for both the AI and you. – chepner Aug 17 at 15:12

15 Answers 15

up vote 16 down vote accepted

It seems to me the real question here is:

Should you always play as if your opponent is a grand master?

You seem to have defined "playing your best" as playing the moves you would make against a grand master, and any other moves, even when played against other opponents, as being lesser. But the rules of chess do not prohibit using your knowledge about your opponent to your advantage, and so I would argue that the 'best' way to play is the moves that are most likely to win against your current opponent. So, when playing someone with less experience, changing your strategy accordingly is still "trying your best".

As for whether you should always play as if against a grand master, that depends on your goals. If you want to be sure to win, the grandmaster strategy seems best. A novice would have to be extremely lucky to find a vulnerability in it and exploit it effectively. If your goal is to 'smoke' the novice and checkmate as quickly as possible, then your best strategy is probably as you said, to take risks, banking on your opponent not knowing how to punish you for them, and to be aggressive, banking on being able to slip away if you get in trouble. If you play as if against a grand master while actually being against a novice, you'll waste time protecting against moves your opponent won't make.

Is it a good strategy to always play like a grandmaster?

Obviously yes, but unfortunately 99% of chess players are unable to do this even if they wanted to. ;)

Is it a good strategy to knowingly play moves that you know (or strongly suspect) to be mistakes?

No.

"Mistake" in the sense that there is an answer that screws you if your opponent finds it. Because either they find it (short term downside) or it works out and you just reinforced a flawed pattern in your brain (long term downside). Blundering your pieces left and right is not a sound strategy, not even against "noobs" (after all, you still need to have some material left to actually mate).

Is it a good strategy to knowingly play moves that you know (or strongly suspect) to not be best (but still okay)?

Sometimes.

In many situations, there are multiple candidate moves and it may have some benefit to play the move you judge second or third instead of the "best" one.

Sometimes it's better to settle for a small but safe strategical advantage (that you are confident to transform into a win against a "noob" anyways) instead of a potentially big advantage that needs a lot of precise tactical play to hold (with the additional risk that you miscalculated somewhere).

Keeping the game less complicated when ahead gives the better player some meta advantage (he's less likely to blunder, and as long as he doesn't, he will eventually win), which might make an otherwise suboptimal move the most advantageous.

Or vice versa, sometimes it may be worth it to go for a move that gives your opponent some advantage in one aspect of the game, while you only get a smaller one in some other aspect. Most gambits come to mind, for example. What matters is that you still should not play clear mistakes, only slightly "worse" moves where you do get some compensation. Sacrificing a pawn for initiative that's probably worth less than a pawn, or even sacrificing a piece for a mate attack is fine. Sacrificing initiative by moving out the queen too early for some hit-or-miss chance to apply the Scholar's mate is not a good strategy, because if they find a good answer, you end up with no compensation at all; and if they don't, they are so inexperienced that any healthy opening would probably have won easily (and quickly!) as well without using a strategy that is useless against better players (again: either short or long term disadvantage).

Making the game more complicated when things are even gives the better player some meta advantage (he's more likely to find a better way through the jungle), which might make an otherwise suboptimal move the most advantageous.

However, these kinds of strategies are used by grandmasters as well, so ironically you are still playing like them.

Is it a good strategy to knowingly play moves that you are not sure about (they may be good, they may be bad)?

No.

Often, this cannot be avoided, of course (your skill sets limits to calculation depth and positional judgment). Thus everyone has to do it to some degree. It's not a "good" strategy though. Ideally you should always try your best to evaluate a given position.

  • 1
    "Is it a good strategy to knowingly play moves that you are not sure about (they may be good, they may be bad)?" There is some value in that in that seeing your opponents' response gives you more information about whether they're good or bad. If you're confident that you can recover from the disadvantage a suboptimal move gives you, then it can be useful to try out lines you're not familiar with. – Acccumulation Aug 17 at 16:47

No, you should not.

There are serious games, these are less serious games, there are chess lessons with kids. If you play all games at your maximum strength:

  1. kids are going to hate you, because they don't like losing.
  2. people is going to hate you, because spending 5-10 minutes thinking about every move in a casual game makes everyone bored to the death.
  3. in a very short period of time you'll "burn out", you'll hate the game and you'll never play it again.

So, unless you're playing for the world title or in a very serious tournament, try to make it fun. Don't intentionally blunder or give away pieces (though this might be required sometimes when dealing with kids =), but play risky, adventurous games, so even if your opponent loses, (s)he does it with a smile and tells you: "Thank you for a wonderful game!" afterwards.

After all, chess is just a game...

  • 4
    Regarding point 2, which propagates to the two other points as well, there is a difference between playing bad moves (what some, including me, might like to call "cheezy" because the strategy or tactic is full of holes), and just not spending too much time and energy to do the necessary calculations. Playing intentionally cheezy is never a good idea, but when you play casually, for fun, against kids, you might not want to spend minutes calculating whether a given move is solid or cheezy. Personally, I think you are right, but that you answered a different question than the one which was asked. – Arthur Aug 16 at 13:14
  • 2
    Regarding point 2, any opponent who is also trying to play at their absolute best will not at all be bored to death by a game that challenges them. I often like my opponent taking time to move because then I can be thinking about my possible follow ups. – Todd Wilcox Aug 16 at 17:30
  • 1
    After too serious play the kids are likely to start hating the game too. Which means that you do not treat them like one should treat kids: by inspiring them to learn more. – Roland Pihlakas Aug 17 at 20:41

If your goal is simply to win chess games, then yes; you should always play like a grandmaster (to the best of your abilities) even against less skilled opponents (again, this is if your goal is simply to win. Really putting the squeeze on a novice is a great way to make that person lose interest in chess).

You said against novices that :

You don't need to waste moves defending your army because your opponent probably won't even notice the hanging pieces.

To this I say a few things. First of all, a chess strategy that assumes "your opponent probably won't notice" is pretty flawed. While you may have a high chance of success against novices, people tend to develop the strategies they use the most often. If you repeatedly leave hanging pieces, you'll slowly find yourself checking for hanging pieces less often.

In addition, if you've misjudged the quality of your opponent and blunder away a free knight early, you've begun that game down on material. Blundering a piece away isn't so bad when you're up substantial material, but that's a decision made on the quality of your opponent's position, not the quality of your opponent.

Lastly, again, "people tend to develop the strategies they use the most often." The more games you play sub-optimally, the fewer games you play optimally. I know that's obvious, but you have to consider that in the long term, your skills won't be developed if you avoid using them.

You can checkmate your opponent faster by doing things that're usually considered bad strategies, such as moving your queen out too early in the game.

That's potentially true, but the goal of the game isn't to checkmate your opponent as quickly as possible. With the exception of a few notable games, it's not important if you beat your opponent in 4 moves or 100, so why risk the game? For example, you could try Scholar's mate against a novice and probably win most of the time. However, if your opponent is familiar with this, now you've played a sub-optimal opening for pretty much no reason.

Additionally, I discovered in the Chess.com mobile app that its "best move" hints change relative to the computer opponent's level.

This is likely because chess engines view time as a precious resource. When you ask for the "best" move and the computer knows it's playing a novice, it's not going to delve as deep into lines, because it doesn't have to. This does undercut my answer a little, but keep in mind computer play and human play is very different. A computer has a harder time allocating time for future moves, and so you'll find it trying to shave precious seconds off where possible in case it needs them later.

A more interesting question though, is Should I make the best move available every time?

Interestingly, the answer to that question is no. That's because chess is a strategy game, and a large part of strategy is evaluating your opponent. When outmatched, you'll sometimes see grandmasters start with a traditionally weaker opening. This is because they've prepared it; they've analyzed most of the opening lines, and they know their opponent probably hasn't. As such, they might be able to leverage a small advantage that's less obvious to their opponent right away.

In addition, there's some interesting psychology behind making sub-par moves. In particular, Mikhail Tal was notorious for deliberately ignoring the best move available. However, he had the skill required to back up these moves, so players quickly became very cautious when he gave them an "obvious" response. That often led to them making their own sub-par moves.

Tal's style of play was so intimidating that James Eade listed Tal as one of the three players contemporaries were most afraid of playing against (the others being Capablanca and Fischer). However, while Capablanca and Fischer were feared because of their extreme technical skill, Tal was feared because of the possibility of being on the wrong side of a soon-to-be-famous brilliancy.

There are plenty of stories of Tal making "bad" moves which confused or concerned his opponents enough to give him an advantage, but that's similar to the "weaker opening" point as well; Tal knew his advantages weren't strictly based in material.

When evaluating their own play, your opponents will likely assume you'll make optimal moves. If you do, they'll have a line prepared and it's truly a battle of who has better chess strategy. But when you throw in a few moves they didn't expect, they have to reevaluate their own line and they have to try and understand what you're thinking. Tal was an expert at this; his habit of making unexpected moves made it harder for his opponents to evaluate deep lines. If done correctly, this can give you a larger advantage than the material or positioning you've offered.

As a lower level player, I go a little farther away from this then most others seem to.

In general, I assume that my opponent will find the best move I can think of (that isn't anywhere near grandmaster strength, honestly, but that's the best I can do).

At times, I am willing to do the following:

  • Make a slightly inferior move (according to a grandmaster) that puts me in a position I play better (for me, that means endgame, for others, it may mean tactics or positional play)
  • Make an opening move that the book says is inferior, but it would take a higher level opponent to exploit it (that way, my opponents stop playing grandmaster moves they learned in a book)
  • If my opponent is in time pressure, play a sacrifice I don't think they can find a way out of in their remaining time
  • If I believe my opponent is obviously weaker, play a move that will end the game quicker so I get some more rest time before the next round (I'm taking a chance here, but the reward of being fresh for the next game is more than enough for the risk of my opponent actually finding the best play this game)
  • If losing, set a trap for my opponent and see if I can get back in the game (losing quicker is better than losing slower anyway, and maybe I can get a draw out of it)

Depending on how "best" means. Say you play Sicilian Dragon the best against e4 but also have decent knowledge in French. Your opponent always play as white 1.c4 but never 1.e4.

Now you two meet at the last round of a major tournament while you two are tied for first. Whoever wins this game gets the trophy. He now opens 1.e4.

Should you go into Sicilian Dragon where he's obviously heavily prepared or do you go French?

While it's a different game, this is somewhat done in at least one game where two opponents play tactically with pieces on a board with no hidden knowledge, which makes me think it would apply to chess as well.

A few years ago, a computer (AlphaGo) beat a world-class Go player (Lee Sedol) for the first time.

One thing many Go experts picked up in, was that AlphaGo's approach favoured a more certain win (even if only by a much smaller margin or less 'dramatically'), over a bigger win. So it might well prefer a play that was less "optimal" than most human masters or grandmasters would choose, if it concluded that it would give a more sure 50.1% chance of an eventual win (even if by just one counter), for example.

(A chess equivalent/analogy might be a calculation (in machine play terms) that following 'ordinary' grandmasters quality play would give a player an estimated 63.1% chance of an 'ordinary' and 'good' win, but allowing the play to degenerate and drag out into much slower and apparently even 'pointless' or 'incomprehensible' endplay actually gives a 63.2% chance eventually of a very narrow win, even though this is a course that (almost) no human player would pursue. Note that this is my own understanding and analogy, so I could be a bit wrong, but hopefully it gives the idea.)

This also makes it difficult to answer the question, because it raises the question of what the "best" move should actually be defined to be. Perhaps not necessarily the one that grandmasters think it is?

  • I think you have some good points that are applicable also to chess, but could you clarify what you mean by a "more sure 50.1% chance of an eventual win"? – Dag Oskar Madsen Aug 16 at 12:05
  • 2
    In chess, there's no such notion as a “win by bigger margin”. Evaluating a position really means you estimate the odds of it ending in win, loss or draw. If anything, a “win with small margin” would be if you sacrifice a significant amount of material but then manage to put your opponent in checkmate, but that's usually considered a more dramatic win than boring crushing through material advantage. – leftaroundabout Aug 16 at 12:54
  • 1
    Your description of AlphaGo seems to contradict itself. You say that "AlphaGo's approach favoured a more certain win (even if only by a much smaller margin or less 'dramatically'), over a bigger win" but then you say that it was happy with moves that would give it only a "50.1% chance of an eventual win (even if by just one counter)." That can't be right. Surely you mean that it would prefer a high chance of winning by one counter to a lower chance of winning by multiple counters. 50.1% isn't high at all, especially for a narrow victory. – David Richerby Aug 16 at 12:54
  • But, in any case, the only reason that this would be surprising is that Go players aren't used to computers. Anyone who's programmed computers to play games knows that you program it to maximise the chance of winning, not the margin of victory. – David Richerby Aug 16 at 12:55
  • 2
    @DavidRicherby To put my own spin on it: nobody was surprised that the computer played to maximize its chance of winning. What surprised people was just how visible this was in its strategic choices; a priori it seemed to us mere humans that increasing your score difference was a pretty natural way of maximizing your chance of winning. It took superhuman calculation abilities to teach us that we severely underestimated how common it was to be in a situation where increasing your apparent score doesn't coincide with maximizing your chance of winning. – Daniel Wagner Aug 17 at 17:24

If you play a strategy that can be countered more easy but can also lead to a faster victory against someone with lower expected skill, that's just normal risk and reward for me. You increase your chance of defeat for a faster victory.

Playing worse, or even losing is useful, if you want someone to play more games with you. Like when you teach someone or if you for any other reason want others to play with you instead of not doing it any more because of frustration. In that case, chosing a nonoptimal strategy in a single game is part of a larger strategy over multiple games.

Should you always try your best?

No.

Simply put some things are not worth it. Sometimes you run into a position that would take too much time to play for no reward. Sometimes you need your opponent to teach you want to do in certain positions because you have no idea. In those cases it's usually best to resign.

Really, you need to figure out how hard you should go and how hard you're willing to go for every game.

A word of warning, every single action you make starts to build a habit; if you play too carelessly against a noob it will effect your chess game against stronger opponents. It might not be worth the extra effort to play "nice" when you have to unlearn some of the habits you've already developed and then relearn them later when you have to play against Magnus Carsen - that's 100% who you're playing against I already know lol.

You gotta know when to fold 'em.

No.

My friend and I would often play chess, just as something to do while we're hanging out. One day, I paused and started analyzing the board. He said, forget it. Let's do something else. Where's the Frisbee?

He wasn't afraid of losing. He was a much, much better player than me. It's just not the type of game he wanted to play that day.

Should you always play your best: Yes
Does that mean playing like your opponent is a grandmaster: Sometimes.

Chess is a game. You get to decide what the purpose of playing that game is. It's entirely up to you. I can't tell you why you should play chess, Kasparov can't tell you why you should play chess, your wife can't tell you why you should play chess (okay, maybe you should listen to your wife). It's all up to you.

Now the "always play your best" argument is typically given to lower ranking players to encourage them to not play sloppy. It's easy to get dependent on cheap strategies which don't work at higher levels of play. If you don't "play your best," you risk such dependency. If you always play as though the other player is a grandmaster, you avoid this issue.

Now I will note that one does not just play a grandmaster on the board. Grandmasters study the personality of their opponent at every single chance they get. They aren't trying to beat the king on the board. They're trying to beat the person. Thus, if you see your opponent has a personality glitch that may cause them to make a bad move, abuse it! That's the best way to win.

But what if your goal is not to win? Chess is just one part of life. There's a lot more out there. One of the other answers points out that if you do nothing but slaughter novices left and right, they stop wanting to play with you. Did you care about them spending time with you? If you did, then you probably shouldn't slaughter them. You should't play like a grandmaster was across the table.

But you also don't want to be sloppy. You can't just go make stupid moves because then you aren't challenging yourself. Then, when the grandmaster does come along, you're hosed because you're out of practice playing carefully.

A hybrid solution, which I highly recommend is, "Play to maximize your learning." Don't play risky moves just because you can get away with them. Play risky moves to practice playing in difficult situations. Handicap yourself to challenging yourself. Play just bad enough to get into a position where their skill level plus the handicap you give them is just enough to force you to play your best -- to play like there was a Grandmaster on the other side of the table. Hang a rook, not because you think you can get away with it, but because you believe that, with a 1 rook advantage, your particular opponent will actually challenge you and force you to play your best.

Or, perhaps less drastically, if you're used to playing an open aggressive style of chess, consider putting yourself into closed detailed analytical positions just to see how you handle them.

Then every game you play teaches you about yourself.

And, as an added bonus, you are intentionally putting your opponent into a position where they stand a chance to win if they play their best. And that makes for better opponents. Soon you may find they learn enough to surprise you!

No when you want to teach something specific

A complete novice (in any game) does not know some of the strategies that lead to a win (in chess to a checkmate) and you need them to become aware of them. It doesn't even have to be an immediate win, but a specific moves combination or gaining specific position in order to get a strong advantage. If you want to teach a person to be aware of this position, you need to expose them more often than you would in a normal course of a game. So you tend to build your strategy upon reaching this specific position rather than winning itself. It doesn't mean though that you should give up defending your pieces or just leaving easy picks (unless its an intentional sacrifice to show a trap one might fall into) - it teaches bad habits rather than anything else. But you still use moves that could be considered bad if you play towards win only.

Yes if you play a general game

No mater how skilled your opponent is, if you don't intend to teach something specific, but rather give a general practice, play as if your opponent is a grand master. Use handicap to even your chances. Explain that you do this to even your chances as the opponent might feel down for loosing despite you have removed one or two of your own pieces before the game even started.

You may give a different versions of a handicap. Recently I played with a nephew who is still very novice. He "invented" a "great" strategy with just one flaw - it was ignoring I can move pieces as well ;-) I agreed for him to use this strategy and not use any of the pieces he was going to beat with his strategy nor defend them in any way under a condition the only moves he does with his knight (that was supposed to beat 3 of my pieces including the queen) were in this specific order. He could use any other piece in the meantime but knight had to take pieces in the exactly agreed order whenever he moved it. As you can imagine I did all my efforts to, despite being bound to this agreement, take as many other pieces as I could and get as close to winning as possible. I really had to challenge myself and the kid played one of his best parties (in terms of how much he analysed each move) - he did less casual moves than ever before. He eventually managed to complete the "3-moves-3-picks" strategy in something like 20 or more moves and only when he realised he is going to loose anyway and at that moment can no longer do anything about it.

The benefit was he had to focus on slightly less pieces (3 of mine were actually out of the game) so the game was somewhat easier for him to analyse but he knew I will put all my best efforts to make it a hard time for him ;-)

Other option is to limit the time you have for the game or burden yourself playing with more than one unskilled opponent at the same time (simultaneous exhibition)

On the other hand - I still remember the first game I won against my father without any handicap from his side. It was a real pride for me. Not even closely comparable to that when I was wining a handicapped game.

If I am in an inferior position against a player I deem no better than I am, instead of playing what Fritz will call the best move later, I'll probably look to set traps. In a similar situation, the Grandmaster (playing another Grandmaster) might just resign, if the position is bad enough, or reconcile himself to playing for a draw. Setting the opponent up for a trap is probably not a GM's best plan.

Always try your best no matter who you're playing. You never know if someone is underrated and better than they seem. Also, even if someone isn't that good of a player, you playing recklessly gives them more chances than they should have.

The only reason for playing differently against lower rated players is to finish the game slightly faster, but this is a poor reason. You can play safely without wasting too much time.

In general, I find that when I (subconsciously) play carelessly against low-rated players, I end up playing worse. This sometimes leads to the game being closer than it should be, which makes the game longer than it would have been.

As an interesting aside, I also find that, on the odd occasion I have played a grandmaster, I tend to play extra safe, again subconsciously. This also makes my play worse, since I'm not playing like I would have. So the strategy that yields best results is to play against everyone the same, if possible ignoring their playing strength. If I could consistently do this, I'd probably go up 50-100 rating points (and I'm around 2200).

Timur Gareyev, a 28-year-old grandmaster set a world record for playing 48 games at once.

In the article about him it was noted he would play sharp gambits to bring the game to a quick conclusion.

Strategy varies.

protected by Phonon Aug 19 at 20:21

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.