Can you tell if someone is cheating in Wordfeud?

We can say that with some confidence, because we make one. TileVision does exactly what a suspicious player imagines. You take a screenshot of your Wordfeud board, switch apps, and your phone lists every word you could play, highest score first. The app never connects to Wordfeud and doesn’t read the game’s data or anyone’s account. All it sees is a picture, on the phone that took it.

To Wordfeud, a move is a move. Whether someone found it by staring at their rack for ten minutes or by checking an app, the word that lands on the board is the same.

Can Wordfeud detect cheating?

We can’t speak for Wordfeud’s side of things, and we won’t guess. The problem itself is an old one, though. In 2012 the Norwegian tech site Tek.no ran a piece on cheating in Wordfeud. The Norwegian Scrabble federation said a game should be played on equal terms, and that using helpers your opponent hadn’t agreed to was unacceptable. Wordfeud’s founder said cheating was hard to guard against because it happened outside the game.

Helpers still work that way. A website on a laptop, an app on a phone: the thinking happens away from the board, and what reaches the game is a move. So you’re left with signs. Some are worth noticing, but none of them is proof, and every one is also something a strong player does on a good day.

Signs your opponent might be using a helper

Words they can’t explain

The classic. Your uncle, who answers every message with a thumbs-up, plays a seven-letter word you’ve never seen and lands it on a triple word. You ask what it means and get a shrug.

The catch is that serious players learn words without their meanings all the time. Memorising long lists of valid words is a big part of getting good at this kind of game, and many strong players couldn’t define half of what they play. What’s worth noticing is someone who never studies and can’t say where they came across the word at all.

The top-scoring play, turn after turn

A solver puts the biggest score at the top of its list, and the easy thing is to play it every time. People don’t, even very good ones. They miss things, and they pass up points on purpose, to keep the board closed or to hang on to good letters for the next turn. Someone who always takes the maximum, even when it hands you a triple word, is playing like the top line of a list.

But you never see their rack. You know the word they played, not the seven letters they were choosing from, so you can’t be sure there wasn’t something better. “They always score big” stays an impression, however strong.

A sudden jump in level

Someone who used to lose to you most weeks starts winning comfortably, with a seven-tile play and its 40-point bonus in nearly every game. That kind of change can happen overnight with a helper, and rarely does without one.

People do get better, and some improve fast once they start learning words or playing several games at once. A climb over a couple of months is what learning looks like. Harder to explain is a change from one game to the next with no story behind it. So is play that’s wildly uneven: a long word through two premium squares, then a turn where they miss an easy spot for the Z.

Moves that don’t fit the time they take

A lot of Wordfeud gets played in the gaps of a day, so the time between moves usually tells you about someone’s schedule, not their vocabulary. It says more when you’re both playing at the same moment. A hard board normally takes longer than an easy one. If every reply comes back after roughly the same couple of minutes, whether it’s a two-letter word or a 60-point one, that steady rhythm fits someone running each turn through an app.

Then again, some people just play fast, and some work out their next move while they wait for yours.

What to do if you think your opponent is cheating

Be honest with yourself about what you’ve got. One odd word means very little. Several of these signs across a run of games, from someone whose play used to look different, are worth a conversation. They still aren’t grounds for an accusation.

• Talk to them. Ask what a word means or where they found it, as a question rather than a charge. If you’re wrong, nothing is lost. If you’re right, you’ve given them an easy way to say so.
• Agree on house rules. Settle it before the next game: no helpers at all, helpers for both of you, or helpers only in practice games and for looking back at finished ones. The Norwegian Scrabble federation drew the line in the same place back in 2012. Their objection was to help the other player hadn’t agreed to.
• Play people you trust. With a stranger you’ll never get a straight answer, so there’s little point losing sleep over one game. Keep the games you care about for people whose word you’d take.

Why people use a helper at all

Not everyone with a solver on their phone is out to beat their in-laws. Lots of people use one to learn which words exist and, more usefully, where they fit: the short words that squeeze into a corner, or the hooks that let one tile make two words.

Others use one to look back at a game once it’s over. Screenshot the positions where you got stuck, then go through them when the result no longer matters and see what was there. It’s humbling at first. After a while you start spotting the same openings during the game, without the app.

And some practise alone. Screenshot a board, find your best move, then compare it with the list. It’s a quick way to train your eye for premium squares.

Where we stand

We make one of these apps, so weigh our view accordingly. A helper in a game where both players know about it is fine. As a teacher, in practice or once a game is over, it does something a word list can’t, because it works on your own board. Against a friend who doesn’t know, that’s between you and them, and if they found out they’d have every right to be annoyed.

If you think someone is using one against you, ask them. And if you want to see what a helper actually shows a player, here’s how TileVision reads a Wordfeud board.