The best two-player games when you only have one phone
Three genuinely different answers to the problem of fitting two players onto one handset, what each one costs, and the games on this site that get it right.
By Aayush Kumar · Local multiplayer · 8 min read
Two people are waiting for a train. One of them has a phone. Nobody wants to install anything, the station wi-fi is asking for an OTP, and there are nine minutes before the train arrives. This is the single most common two-player situation there is, and almost nothing is designed for it.
The usual answer — two devices, two accounts, a lobby, a matchmaking screen — solves a different problem. What is actually needed is one screen that two people can both reach, with two sets of controls that do not collide, in a game short enough that the loser gets a rematch before the train comes. Every two-player game on this site works that way, because there is no networking code anywhere on it. That constraint turned out to be productive. It forces a choice, and there are only three sensible answers.
Approach one: cut the screen in half
The cleanest solution is to give each player physical ownership of a region of glass. Your thumb works in your half, mine works in mine, and the game reads two simultaneous touches as two separate players. Modern browsers handle multi-touch properly, so this costs nothing technically. What it costs is design: the game has to be legible from both ends of the phone at once.
Potato Pong is the neatest example because it solves the orientation problem outright. Held upright, the court runs top to bottom and you are at the bottom. Turned sideways, the whole court rotates so you are on the left. Two people can therefore sit opposite each other across a table with the phone flat between them, or side by side over a laptop, and neither of them ends up reading the game upside down. Each player slides a finger anywhere on their own half. There is no control to learn at all, which matters when the person you handed the phone to has never seen the game.
Air Hockey does the same with a wrinkle worth knowing about: the mallet does not teleport to your finger. It chases your finger through a heavy smoothing filter, so it carries momentum, and how hard the puck leaves depends on how fast the mallet was already travelling when it arrived. A stationary mallet absorbs a shot. That means flicking beats shoving, and it means the split-screen version rewards the player who understood that first. The puck keeps about 98.5 per cent of its speed per frame, which is close to frictionless, so a rally that starts fast stays fast.
Tennis 3D and Table Tennis both put one player on the lower half and one on the upper. Tennis is the more interesting of the two to share, because there is no shot-select button anywhere in it — which shot you hit is decided entirely by when you meet the ball and how you were moving at contact. You steer by running through the ball. That is a lot to explain, but nothing to teach: the person opposite you works it out in about four points.
Math Duel makes the sharpest use of the split by mirroring it. Fifteen questions, four answer tiles each, and in two-player the upper half is flipped so it reads the right way up from across the table. Both players see the same question at the same instant and the first correct tap takes the point. It is the only game here where the split screen is the entire game rather than a delivery mechanism.
Foosball and Smile Ball round the category out. Foosball gives each side a drag zone and a Kick button; the twist is that all three of your rods are welded to one slider, so covering your near post drags your forward across the table too. Smile Ball gives each player their own Jump and Grab buttons on their own half, and punishes goal-hanging with an actual rule — stand in the band in front of your own net for a second and a half and the other side is awarded the goal.
Approach two: opposite corners
Splitting the screen only works when the game world can be divided. A shared arena seen from one camera cannot be. So instead of dividing the picture, you divide the edges: player one's controls live in the bottom-left and bottom-right of their side, player two's in the opposite corners, and both of them look at the same middle.
Car Football is built around this. One steering stick and one action button per player, placed at opposite corners, both watching the same walled arena. It works because the game is legible from any angle — a ball stuck to the nose of a car, two goals, a clock — and because possession rather than aim decides the scoreline. Drive over the ball and it sticks to you; the other car takes it by arriving, with no button involved. A pair of timers stops that becoming a scramble: the thief is protected for about four tenths of a second and the victim cannot steal straight back for a full 1.4 seconds.
Boat Tagger duplicates its thumbstick and dash button on the opposite side of the screen when two people share it. Four paper boats on a pond, one of them It, ninety seconds. The boats turn at a fixed rate however hard you swing, so a chase is geometry rather than reflexes, and the no-tag-backs rule stops two people trading a tag forever in one corner.
Zombie Fighter is the odd one in this group, and the most interesting. Its two-player mode is not versus at all: one person drives and the other works the turret, and neither half can save itself, because the driver has no weapon and the gunner has no steering. That asymmetry is the whole appeal. It also produces the only argument on this list that is genuinely about whose fault it was.
Approach three: pass the device
The oldest answer, and still the best one for anything turn-based. Nobody has to share glass, nobody has to hold the phone at an awkward angle, and the screen can use its whole area for one player at a time.
Every board game here supports it. Chess, Four in Line, Bigger Tic Tac Toe, 5 vs 10, Tic Tac Push and Carrom Clash all simply hand over. Ludo goes furthest: in pass-and-play all four colours are human, which is the version families actually play. The rule decisions behind those boards are worth reading before you argue about them.
Two games extend pass-and-play to things that are not turn-based at all, by fixing the randomness. Bike Rush 3D gives player two the same 2,400-metre stage with the same traffic, and the faster finish wins. Ridge Rally does the same with the same hills, the same fuel and the same seed, and plants a marker post during the second run at the distance the first player reached. Drive Mayhem has one person run all eight stages before handing over, scored on attempts rather than time. None of these is simultaneous, and all three are more competitive than they sound, because you are racing a number rather than a person and the number does not get flustered.
Then there is the category that needs hidden information, which pass-and-play handles better than a split screen ever could. ONO puts a cover screen between turns so neither player catches a glimpse of the other hand. Hand Cricket locks the first player's pick before the second player chooses, so you turn the screen away and pass it across. Pixel Derby keeps one player's running order hidden while the other commits to theirs. You cannot do any of that on a split screen, because the other person is looking at it.
What each approach costs
Split screens halve the space. On a 6-inch phone that is a genuine problem for anything with fine aiming, which is why the split-screen games here are all about position rather than precision. They also assume both players have a hand free, which rules out playing while standing on a bus holding a rail.
Opposite-corner controls preserve the picture but produce elbow collisions. Two people gripping the same handset from opposite sides is physically awkward for longer than about three minutes, which is exactly why Car Football runs a three-minute clock and Boat Tagger runs ninety seconds. A ten-minute match in this format would be a worse game.
Pass-and-play costs momentum. Every handover is dead time, and the more players you add the worse it gets — four-player Snake & Ladder roughly doubles the wait between your own rolls compared with two. It is the right choice for turn-based games and the wrong one for anything where tension depends on continuity.
The honest exceptions
Three two-player games here need a keyboard and will not work as touch two-player: Stick Fighter (WASD with F and G against the arrows with K and L), Sumo Fighter, where the touch pair belongs to player one only, and Bumper Kart Arena. All three are good games and none of them solves the one-phone problem. They are laptop games that happen to be in the same directory.
It is also worth saying plainly what the site does not have. There is no online play anywhere on it, in any game. The second player has to be in the room. That rules a lot out, and it rules one thing firmly in: you never wait for a lobby, and the opponent never rage-quits without consequence, because they are sitting next to you.
Where to start
If you have never done this before and the other person has eight minutes: Potato Pong first, because the rules explain themselves in one rally. Then Math Duel, which is a fair fight between almost any two people. Then Carrom Clash if you want something that will still be interesting on the fourth game. For a group rather than a pair, the pass-and-play party formats scale further than any of these do. And if the controls feel wrong in any of them, that is a solvable problem with a known set of answers.
More from ARTHIS.space
- Party games Games for a group standing around one screen Six games that put one phone in the middle of a room, what each needs from the group, and an honest note about which are really one-player games with an audience.
- How it works How a whole game fits in a single web page A tour of what a browser hands you for free, why canvas and requestAnimationFrame are enough to build on, and the four places this approach genuinely runs out.
- Game AI How the AI in these games actually works, and how to beat it Four genuinely different approaches to computer opponents, the numbers behind each one, and the specific weakness every approach hands you.