GameVui123 logo GameVui123Casual games, honestly reviewed
Casual and Mobile Games

How Mobile Card Games Shuffle and Why Hands Feel Rigged

My aunt has played the same solitaire app every evening for six years and she is certain it cheats. Her evidence is a streak of eleven deals in one week where the last king sat buried under a stack she could never open.

A phone showing a solitaire deal in progress beside a shuffled deck of real playing cards on a table

My aunt has played the same solitaire app every evening for six years and she is certain it cheats. Her evidence is a streak of eleven deals in one week where the last king sat buried under a stack she could never open. I sat with her phone for an hour, wrote down the opening layout of nine deals, and worked through them on a table with real cards. Seven were winnable. She had missed the same kind of move in five of them, which I know because I missed two of them myself on the first pass.

The feeling that a card app is rigged is almost universal and almost always wrong, but the reason it is wrong has nothing to do with trust. It has to do with how shuffling is actually implemented, what share of deals in each game can be won at all, and a setting that a surprising number of apps turn on without telling you. Once you know those three things, a bad run stops looking like a conspiracy and starts looking like arithmetic.

What a shuffle is inside the app

A physical shuffle is a messy mechanical act. Seven decent riffles get a deck close to random, and most people at a kitchen table do three. A digital shuffle is a list of 52 numbers rearranged by an algorithm, and the standard one is Fisher and Yates: walk the list from the end, and for each position swap the card there with a card drawn from the positions not yet visited.

Done properly, every one of the roughly 8 times 10 to the 67th possible orders is equally likely. Done slightly wrong, it is not. The classic error is to loop through all 52 positions and swap each with any random position in the whole deck, which looks identical in code and produces a measurable bias toward certain arrangements. That bug has shipped in real card software more than once, including a famous online poker client in the late 1990s whose card order could be predicted from the clock.

The second ingredient is the random number generator. Phone games use a pseudorandom generator seeded with something like the system clock in milliseconds. Same seed, same deal, every time. That is not a flaw. It is exactly how daily challenge modes give every player in the world the same hand, and how a replay this deal button can reproduce your layout after you have ruined it.

Most deals you lose were genuinely losable

This is the part that settles most arguments, and it is not a matter of opinion. Researchers have worked out, for several of these games, what share of random deals can be won by a player who can see every face down card. Those figures are ceilings, and they are lower than people assume.

GameShare of deals winnable with perfect informationWhat a good human actually wins
FreeCellVery close to every deal; of the 32,000 numbered Microsoft deals, only number 11982 is unsolvable75 to 95 percent with practice
Klondike, draw oneAround 82 percent35 to 50 percent
Klondike, draw threeRoughly 82 to 92 percent in theory, far harder in play10 to 20 percent
Spider, one suitAlmost all80 percent and up
Spider, four suitsEstimated near 95 percent, with brutal lines5 to 20 percent
PyramidAround half10 to 30 percent

Look at the Klondike draw three row. A ceiling in the eighties against a real win rate near fifteen percent means the game is not withholding wins from you. It is offering wins you cannot see, because the information is hidden and one early choice closes a door you did not know was there. In Pyramid the ceiling itself is the story: lose six in a row and you may simply have been handed three unwinnable boards.

The setting nobody mentions

Here is where apps genuinely interfere, and it is the opposite of cheating against you. Many modern solitaire apps ship with a winnable deals only mode, sometimes called solvable deals, and in several of the biggest apps it defaults to on. Instead of shuffling freely, the app draws from a stored library of deals already verified as solvable, or it shuffles and tests until it finds one that passes.

That changes the game more than people realise. If every deal is winnable, a loss is always your fault, and the statistics screen stops meaning what you think it means. A 60 percent win rate on guaranteed solvable Klondike is worse play than a 45 percent rate on truly random deals. Dig through the settings for it before you compare your numbers with anyone.

It also changes strategy at the margins. On a guaranteed solvable board, a line that requires an unlikely card to appear in the right place is worth gambling on, because the deal promises a path exists somewhere. On a random board that same gamble is often throwing away the one safe sequence you had.

Against the usual advice: the app is not adapting to you

The theory I hear most is that these games watch your performance and hand out harder deals when you win too much, to keep you installed. It sounds plausible, since that kind of difficulty tuning really does exist in match three and merge games, where the board generator and the booster offers respond to how you are doing.

For solitaire style card games it does not hold up, for a practical reason. The business model is advertising, and advertising revenue goes up when you play more hands, not when you suffer. A deal you abandon in disgust is fewer ad impressions than a deal you finish. Worse, a tuned difficulty system would be a liability: if it were discovered, a free app with no stake in the outcome would have burned its only asset for nothing.

What actually happens is recency. Eleven frustrating deals in a week feel like a pattern because you remember them, and the forty forgettable deals in between vanish. Run twenty deals writing each result on paper and the clustering usually looks ordinary.

Real money card apps, rummy and poker in particular, are a different world. There the shuffle is a certified component, audited by an outside lab, with the generator seeded from hardware noise rather than the clock. The incentive to interfere is real there, and so is the regulation. If an app handles money and will not say who certified its generator, that silence is the answer.

How to settle it for yourself in twenty minutes

Pick one game mode and stay in it, since mixing draw one and draw three makes your numbers meaningless. Turn off any winnable deals setting you find. Play twenty deals and write down the result of each, plus a one word note on why you lost. My aunt's notes came out as buried king four times, no move at all twice, and ran out of stock five times, which is a reading of her play rather than of the app.

Then use the replay feature. Almost every one of these apps lets you restart the same deal, and that is the only honest test available to you. Lose a deal, replay it, and see whether a different opening changes the outcome. If the same layout turns winnable on the second attempt, the first loss was yours. In nine deals from her phone, five of them fell on the replay.

If the answer is that your reads are fine and the game is simply hard, pick a mode with a kinder ceiling. FreeCell is the honest choice, because almost every deal is winnable and the whole board is face up from the start. For the hardest end of the scale our guide to spider solitaire strategy that clears more deals goes through the ordering decisions that matter, and if you want a card game that works with no signal at all, the method in testing mobile games offline before a flight applies to card apps too. The rest of our phone game coverage lives in casual and mobile.

FAQ

Does the daily challenge give everyone the same cards?

In nearly every app, yes. The date supplies the seed, so your deal and mine match exactly, which is why leaderboards on those modes mean something while leaderboards on regular play do not.

Can an app tell if I am about to quit and give me an easy deal?

It could, technically, and in match three games something close to that exists. In solitaire apps I have never seen credible evidence of it, and the money points the other way. I would not bet heavily either side, but the burden of proof sits with the claim.

Is a shuffled deck ever repeated?

Practically never for a full 52 card order, since the number of possible arrangements is larger than the number of atoms in anything you can picture. Repeats you notice come from a seeded deal, a solvable deal library, or your memory filling gaps.

Should I leave winnable deals mode on?

If you play to relax, yes, and enjoy it. If you want to improve or compare yourself with anybody, turn it off, because a guaranteed solvable board quietly flatters your win rate and hides the one skill that matters most, which is reading a board that may not have an answer.

DV
Dimitri Vance

Dimitri learned rummy before he learned to read, and now writes the rules nobody explains clearly.

More posts by Dimitri

More in Casual and Mobile Games

01doorganico 084goad 3hot8 4465456 after5internprogram ahandfulofblessings ahmadpc ahomcybersecurity allodialresources annodex antiquesandperiodcollectibles aotc asianworldkashmir beedlectrics bloggertemplatefree bluefeatheragency bodyweightsolution campaign08blog carfixadvisor catstreetcollective cbstvsports cnfanspurchasing committeeof500yearsofdignityandresistance covertopsdjs cs-kaiyuangame cuestacomunicaciontotal cxoutlet dancebetgame deskgearhub destinyvillastay detectmagazine di33 diamondsteelbuildings donmo dyanadropship emarsysfinance emimattl escapechan eskitarih evautoplay everythingweddings-orlando farmzone feedviewerapp foxstudiosscratch gleasongrade greetingsday hacklabasu haskell iw-kaiyun jagoanproperti jilidskripsi joltbusinessfinance juegodepokergratis juegospucca kaeru-festival latestmehndidesigns lestanzediottaviano litemagazines localcommunityhero lustrsazan mariashebets menswheels newwebstechs1 nikkit ootheday oxigenvpn plurfitness prefetial probudgetcarpetcleaning psychicmrsmartin pyaten qvantor rejoignezlogis rexzbot riselayouts ruchin rwxqsgd5 sakaltime shibiaowang shop-bot skrtips spotlesshaven strcoding tetapatrium topselller uneplanteartificielle vipertechcr webtech1223411 wo-kygaming yamaha-mekar yeknava yoybuysheets zenunwa yamaha-mekar jilidskripsi ankertoto ankertoto ankertoto ankertoto