How to Play Against the Four Knights Game
By move four, the Four Knights Game has developed the same two pieces for both sides in the same order, and if you keep mirroring White's choices, the position stays symmetrical enough that some lines lead toward forced draws by repetition rather than a real fight. That symmetry is the whole problem to solve here — not a specific threat, but a structural tendency toward a game where neither side ever really commits to anything.
The way to actually fight for a win: the Rubinstein Variation
After 4.Bb5, instead of mirroring with 4...Bb4, play 4...Nd4 — the Rubinstein Variation. This immediately breaks the symmetry, offering to trade a knight for a knight while creating genuine imbalance in the position. White's most testing response, 5.Nxe5, wins a pawn temporarily, but Black gets real compensation with 5...Qe7, hitting both the knight on e5 and, indirectly, the bishop on b5, forcing White to navigate real complications rather than continuing the comfortable, symmetrical development both sides have followed so far.
Why the mirrored line is worth avoiding on purpose
4...Bb4, mirroring White's pin exactly, is completely sound and far from a mistake in the sense of losing material or position — but it leads toward some of the most heavily drawn structures in all of open-game chess. Several main lines from this symmetrical setup are known to allow forced repetition if both sides play the most natural moves, which is exactly the outcome a player trying to win with Black should actively want to avoid rather than walk into by following the most obvious-looking sequence.
If White avoids the pin: the Scotch Four Knights
Not every Four Knights game reaches 4.Bb5. If White instead plays 4.d4, striking the center directly, the position transposes toward Scotch Game ideas but with an extra pair of knights already developed. After 4...exd4 5.Nxd4, you can play 5...Bb4, pinning the newly developed knight and applying pressure similar to the Rubinstein idea above, or continue with straightforward development. This structure carries more inherent imbalance than the symmetrical Bb5 lines simply because the central pawn tension has already been resolved by a real trade rather than a series of mirrored developing moves.
What real compensation looks like after 5...Qe7
In the main Rubinstein line, Black isn't playing for a quick tactical shot — the compensation for the temporarily lost pawn comes from active piece play and the awkward coordination White's pieces (particularly the knight on e5 and bishop on b5) can fall into if White tries to hold onto every material gain. Play actively toward completing development and creating threats rather than immediately trying to win the pawn back by force; the position rewards patience with the initiative more than it rewards an immediate tactical strike.
The mistake that turns a fighting choice into a passive one
Some Black players choose the Rubinstein Variation specifically to avoid a draw, then play the resulting complications too cautiously, offering trades and simplifications the moment any tension appears. This defeats the entire purpose of choosing 4...Nd4 in the first place — if you're not willing to navigate genuine complications once you've created them, the symmetrical mirrored line would have been the more consistent choice from the start.
A quieter middle ground: the Spanish Four Knights without Nd4
If offering a pawn outright doesn't appeal to you but the fully symmetrical line still feels too passive, there's a middle path: play 4...Bb4 as normal, but rather than mirroring every subsequent White move, look for the first reasonable opportunity to break symmetry yourself with a timely ...d5 or ...Bxc3 followed by active piece play. This doesn't create the same immediate imbalance as the Rubinstein Variation, but it avoids drifting all the way into the most heavily analyzed drawish sub-lines simply by refusing to mirror indefinitely once a genuine alternative appears.
Decide whether you actually want a fight before move four
The Rubinstein Variation only pays off if you follow through on the complications it creates. If a calm, likely-drawish game genuinely suits your goals for this particular game, the mirrored 4...Bb4 is a completely reasonable, sound choice — just make the decision on purpose.
The Ruy Lopez page covers the closely related Bb5 pin reached one move earlier, before Black's second knight has developed.
Frequently Asked Questions
Is the symmetrical 4...Bb4 line actually forced to end in a draw?
No — 'known to allow' forced repetition in some specific sub-lines isn't the same as every game reaching one. Plenty of symmetrical Four Knights games are won by accurate middlegame play from either side; the point is that the drawish lines exist and a player specifically trying to win should be aware of and avoid them.
Does the Rubinstein Variation risk losing material for good if White plays accurately?
White can hold onto an extra pawn with precise play in some lines, but Black's compensation in piece activity and development is well established at every level — it's a real strategic trade-off, not a miscalculation, and it's been played successfully by strong players for exactly that reason.
What if White plays 3.Bb5 instead of 3.Nc3, skipping the Four Knights structure entirely?
Then you're in Ruy Lopez territory instead — a different opening with its own theory, not a delayed version of the Four Knights. Watch White's actual third move rather than assuming the Four Knights automatically follows from 1.e4 e5 2.Nf3 Nc6.
Is it worth learning the Four Knights specifically, or should I just treat it as an occasional transposition?
If you already play 1...e5 as your main defense to 1.e4, it's worth a small amount of dedicated study — it appears often enough at club level that treating it purely as an incidental transposition means facing real decisions unprepared.