I was playing chess against an AI, and found myself in this position (Black/me to move):

r1b2rk1/4n3/p1p1p1p1/1p1pPp1p/1PnN1P2/2NB4/2PP2PP/1RB2RK1 b - - 0 1

The only moves I could think of were either obviously bad (e.g. trying to move the knight on c4), obviously pointless (e.g. shuffling the king around), or would mess up my pawn structure.

I tried plugging this into an analysis engine, and it suggested 1. ... a5 (which, now that I look at it properly, makes a great deal of sense to me). It also said that I'm down about four pawns in evaluation, which is unsurprising as I'm down a bishop, the other bishop is bad, and neither of my knights has much freedom of movement. Meanwhile, White's bishop and knights are active and greatly limiting my options. So, on the whole, this game does not seem to be going well for me.

But my broader question is not "what specific move should I be making?" Instead, what should I be trying to accomplish, in general terms? Because on this board, in this position, I really have no idea what my next goal should be. I feel like most of my pieces are either already developed, or impractical to (further) develop; I cannot see any reasonable lines of attack that might actually yield an advantage; and it's also rather difficult for me to infer what White should be doing, so I can't even effectively defend myself.

The AI doesn't allow draw offers. Should I just shuffle the king around and count to 50, or do I actually have a chance of pulling this one back?

(Lest anyone accuse me of cheating against an AI, the game is already over, but I still find this position utterly baffling.)

  • Something worth a thought: Placing a dark squared bishop anywhere (except f6) behind the pawns for black yields an advantage for black. Perhaps the problem is not what to do in this situation but what should have been done to preserve material equality.
    – user25182
    Commented Feb 20, 2021 at 5:39
  • 1
    @AndréArmatowski: Yes, I can see that the bishop pair would be nice to have, but I'm a beginner. I tend to blunder from time to time, and sometimes it's nice to at least try to continue playing from that point.
    – Kevin
    Commented Feb 20, 2021 at 5:52
  • 1
    Interestingly, my choice was a5 without even thinking. Then someone pointed out Black is a piece down. Ups. It still makes sense - exchange any pawn you can. Commented Feb 20, 2021 at 17:42

2 Answers 2


Let's look at the position and try to evaluate the situation on the board. This is always a good idea, especially if you're struggling to come up with a concrete plan.

r1b2rk1/4n3/p1p1p1p1/1p1pPp1p/1PnN1P2/2NB4/2PP2PP/1RB2RK1 b - - 0 1


In this position black is a piece down for a pawn, so this in and of itself should be cause for concern. But this is actually not the only issue with black's position, although the material situation is of course the easiest thing to consider in a position since it doesn't require anything besides basic arithmetics to calculate.

Factors that are just as important, but often much harder to appreciate for beginners, include king safety, piece activity and coordination, pawn structure, space, etc. Let's begin by looking at black's pawn structure, as it will make clear just how problematic black's position really is.

The "sawtooth" kind of structure that black has set up is something I often see beginners play for, as they feel like it's easy to make sure that everything's protected. But the structure has some serious flaws. First off, the pawns in the back of the structure (g6, e6, c6, a6) are at serious risk of becoming weak since there is no way to defend them with other pawns and it's very hard to advance them without losing them; for instance, moving the g- or c-pawn forward here just lets white capture the pawns for free. But this is not where the problems end. The pawns that are further advanced (h5, f5, d5, b5) are also virtually immobilized; if they were to move forward they would immediately lose support and be vulnerable to attack (imagine if black were to play ...h4, then white would win that pawn if they just manage to threaten it somehow). So in the end, black's pawn structure is rigid and inflexible, and much weaker than it seems at first glance. Another problem with this structure is that it leaves your dark squares very weak. If you want to defend a dark square it has to be done with a piece, and since you're lacking a dark-squared bishop it will be difficult to do this efficiently.

The inflexible pawn structure causes a lot of other headaches for black as well. For instance, black already controls less space than white (in this case it can be clearly seen by the fact that black is relegated to play on the 6th to 8th rank with almost all of their pieces, while white's pieces have a lot more opportunity for movement). This means that black will have less space available to move around their pieces. This, in turn, makes it harder for black's pieces to coordinate without stepping on each others' toes which makes it harder for black to react to white's threats in the long term.

So, to summarize, black is material down and has a weak and inflexible pawn structure which makes their position cramped and gives their pieces a harder time to move effectively (the bishop on c8 is especially terrible here). The only "good" thing about black's position is that their king is pretty safe at the moment. Therefore your aim in this position as black is to draw, since winning is out of the question as long as white doesn't blunder something serious.


Depending on the opposition a draw for black can be easy or difficult or impossible to achieve. A grandmaster would easily win this position as white no matter what black does, the average player could potentially struggle against optimal play and a beginner is probably not going to win this as long as black hunkers down and plays passive defense. Let's discuss a bit how white actually wins this position so that we know what we must try to prevent.

At the moment, the position is pretty locked, so white should ultimately aim to open it up somehow. An open file is enough here to work with, so a pawn breakthrough with h3-g4 is going to be enough to open the position eventually. But the fact of the matter is that with a locked central structure there is no real need to play this immediately, and white can instead focus on improving their position by developing the dark-squared bishop on c1; if this bishop were to awaken it could easily become very strong, since black's dark squares are desperately weak as I described earlier. In fact, a pawn break with h3-g4 may not even be necessary to win here if the c1 bishop manages to worm its way into black's position.

In general, white should try to use their extra space to their advantage in order to "optimize" the position of their pieces and threaten as many of black's weak points as possible at once while at the same time take away squares for black's pieces to occupy safely. Therefore, the following would be an excellent sequence of moves for white to be able to play: Nb3 (this opens the potential scope of the bishop on c1 later on, and it blocks black from making any pawn breaks; also, this knight could be very annoying for black if it were to reach c5, as this would practically immobilize the bishop on c8 completely), Be2 (opens up the d-pawn, and it also starts to prepare for a future g4 pawn break), d3 (pushes black's knight away from its active post on c4), Be3 (this bishop may go c5-d6 or f2-h4, depending on what black decides to defend against). If white so pleases, they could begin the entire operation by playing Be2, d3, Be3 first.


This is by far the hardest question to answer. Black is objectively lost here, and to be honest, I'm struggling to see a decent way for black to play this in order to make life hard for white. If I were playing a weaker opponent I would probably just hunker down here and play passively, hoping that white doesn't find a winning plan (which they probably wouldn't find, at least immediately).

However, the problem with this is that white just wins then by playing along the plan I outlined above, and this is something I kind of expect the average player (~1600 Elo) to find eventually. So I guess that ...a5 is really the best move to try here in practical terms. This move is good to try because it eliminates the pawn on b4, which gives black some chances to take back some terrain on the queenside by means of just rolling their pawns forward if white is careless enough. This would solve the space issue nicely, and give black's pieces a fighting chance to activate themselves. Note, however, that black is dead lost with decent play from white here and even if ...a5 is relatively best it's still far from nearly enough to ensure a draw. For instance, the line 1...a5 2.bxa5 Rxa5 (if 2...Nxa5 white just plays Ba3, activating their dark-squared bishop immediately and threatening the knight on e7) 3.Nb3, and white dominates completely. But since the pawn on b4 has been removed this means that the pawns on the queenside are a bit more mobile, which means that white has to be a bit more tactically alert.

  • (+1) Explaining the positional evaluation in accessible language. Commented Feb 20, 2021 at 15:39
  • This is extremely helpful and detailed. Thank you.
    – Kevin
    Commented Feb 21, 2021 at 6:04

Once clear material down against an AI, with no compensation, I would resign immediately without shame. No point limping on. There is nothing to learn by going forwards since the imbalance dominates all positional thinking. Have a look to figure out what was the mistake, and then use the released time to play another game.

After development, if there are no tactical hammer-blows for either side, positional assessment is key. My favourite book for this is Silman’s “The Amateur’s Mind”. It identifies the key things to check and then gives many, many worked examples of the alchemical magic of “improving one’s position” over the course of a few moves.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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