Turn 3 has a reputation for being unfair, and the reputation is half-earned. Plenty of Turn 3 deals genuinely cannot be won by anyone. But most of the games people lose were winnable, and they were lost to the same misunderstanding: treating the stock as a random pile you flip through, when it is actually a fixed sequence you are allowed to re-index exactly once per pass.
The difference between Turn 1 and Turn 3 is not difficulty in the vague sense. It is one specific mechanic, and it is worth understanding properly, because it changes which move you should make first.
Why does Turn 3 feel impossible?
Klondike deals 28 cards to the tableau. The other 24 go to the stock. In Turn 1 you meet all 24 of them, one at a time, in order, as many times as you like.
Turn 3 deals them in eight packets of three, and only the top of each packet is playable. That is the whole rule, and it has a consequence people rarely state plainly: on any given pass you can reach only eight of the 24 stock cards. The other sixteen are sitting there, visible for a moment as the packet flips, and completely untouchable.
Cycle the stock again and you get the same eight. Again, the same eight. The sequence does not shuffle between passes — it is the same fixed order, split at the same points. This is the part that makes Turn 3 feel rigged: you can see the card you need, you can count that it is second in its packet, and no amount of cycling will ever put it on top.
Until you change the offset.
How do you break the offset?
The moment you play a card off the waste pile — onto a tableau column or up to a foundation — the stock has one fewer card in it. The next time you cycle, the packets of three fall in different places. Cards that were buried second are now on top.
That is the single most important idea in Turn 3, and it reframes what a "good move" is. In Turn 1, taking a card off the waste is worth roughly what the card is worth. In Turn 3 it is worth the card plus a complete re-index of everything you could not previously reach.
Which leads to the counter-intuitive part: sometimes the right move is to take a card you do not particularly need, purely to shift the offset by one and unlock a card you do. Players who lose Turn 3 consistently are usually playing every move on its face value. Players who win are spending moves to buy access.
What does "winnable" actually mean here?
Two very different questions get asked with the same word, and mixing them up is why the figures people quote for solitaire vary so wildly.
- Solvable with full information. Could a perfect player who can see every face-down card win this deal? This is the number research papers measure, and it is the honest upper bound.
- Winnable as you actually play it. Could a good player win it without knowing what is underneath? Always lower, because some deals require a first move you had no way of justifying.
Both are lower for Turn 3 than Turn 1, and for the reason above: restricting stock access removes options, and removing options can only ever reduce the set of solvable deals, never grow it. Any specific percentage you see quoted should tell you which of these two it measured and under which redeal rules, because a figure without those two qualifiers is not really a figure.
Does every deal here come out?
On this site, yes — and that is a deliberate choice rather than luck. Every Turn 3 deal is run through a solver before it is dealt to anyone. If the solver cannot find a win, the deal is discarded and never appears.
So losing a game here means a winning line existed and was missed, which is a much more useful thing to know than the usual ambiguity. When a deal will not come out, the question is never "was this one of the impossible ones" — it is which move closed the door, and that is a question you can actually learn from.
How do you win more Turn 3 games?
Five habits, in the order they tend to matter. The first two are the ones that separate a 20% win rate from a 60% one.
- Count the packet position of the card you want. If it is second or third in its packet, no amount of cycling will reach it. Stop cycling and go find a waste play instead.
- Spend a move to shift the offset. Playing any waste card re-indexes the entire stock. Do it deliberately, not incidentally.
- Empty a column before you fill one. An empty column is the only place a King can go, and a King is the only thing that unblocks a long buried run.
- Do not send Aces and Twos up automatically. Low cards on the tableau are landing spots. Once they are on a foundation they cannot catch anything.
- Turn face-down cards over first. When two moves look equal, take the one that reveals information. Klondike is a game about reducing what you cannot see.
Should you play Turn 1 or Turn 3?
They are different games with the same cards, and neither is the beginner version. Turn 1 rewards planning several moves ahead. Turn 3 rewards understanding one mechanic deeply.
| Turn 1 | Turn 3 | |
|---|---|---|
| Stock cards reachable per pass | All 24 | 8 of 24 |
| Main skill | Sequencing and planning | Managing stock offset |
| Share of deals solvable | Higher | Lower |
| Typical game length | Shorter | Longer |
| Where beginners lose | Emptying columns too early | Cycling the stock on a loop |
If you are learning, Turn 1 teaches you the shape of the game faster. If Turn 1 has started to feel automatic, Turn 3 is where the interesting decisions are.
Frequently asked questions
Yes, and for a specific reason rather than a vague one. Turn 3 restricts you to eight of the 24 stock cards on any given pass, so fewer deals are solvable and fewer still are winnable without knowing what is face down. Restricting access can only remove options.
Eight out of 24 per pass. The stock is dealt in eight packets of three and only the top of each packet is playable. Cycling gives you the same eight every time until you play a card off the waste and shift where the packets fall.
Because the stock order never changes between passes. It is one fixed sequence sliced at the same points, so you meet the same eight cards on a loop. The only way to change which cards surface is to remove a card from the waste, which shifts every boundary by one.
Not in general — plenty of random Turn 3 deals cannot be won by anyone. On this site they are, because every deal is checked by a solver before it is dealt. A deal without a winning line is discarded rather than shown to you.
Usually the one that turns over a face-down tableau card, because information is worth more than position this early. If nothing does that, prefer the move that gets a card off the waste, since that re-indexes the stock and changes everything you can reach.
No. An Ace on the tableau is a landing spot for a Two, and a Two is a landing spot for a Three. Sending low cards up immediately removes the places your other cards were going to go. Send them when you need the space, not on reflex.
It helps, but less than people expect. Extra passes through an unchanged stock show you the same eight cards again. Redeals only become valuable once you have played waste cards between them, because that is what changes the offset.
It is if you are losing to the offset rather than to the deal. Count the packet position of the card you are waiting for. If it is second or third, cycling will never reach it, and recognising that one pattern tends to move people's win rate more than any other single habit.
No. There is no account, no download and no advertising anywhere on the site. Open the page and the deal is already waiting.
Solvable means a perfect player who could see every face-down card would win. Winnable means a good player can win it playing normally, without that information. Solvable is always the higher number, and quoted solitaire percentages usually mean the first one.
Final thoughts
The reason Turn 3 gets called unfair is that its central mechanic is invisible. Nothing on screen tells you the stock is a fixed sequence being sliced at moving boundaries, so a card you need looks like it is being withheld at random when it is really sitting in a position arithmetic will not give you.
Once that clicks, the game stops being about luck and starts being about access — which cards can you reach, and what will it cost to reach the ones you cannot. That is a better game than the one most people think they are losing at.
Next time a Turn 3 deal stalls, before you cycle the stock a fourth time, look for a waste card you can play anywhere at all. That one move is usually the whole game.