Editorial Standards

We write what we believe is best for solitaire players. That is the whole job. We do not accept pay-to-rank placements in our product reviews, we do not let advertisers choose the angle of an article, and we do not publish anything we would not be comfortable defending on the record. If an advertiser ever objects to an editorial decision, the editorial decision wins. The SolitaireStack.com Editorial Team has final say over every published page.

We treat product coverage the same way a trade publication treats its review section. Apps, games, and tools we recommend are ones our editors actually use, not ones that have paid for placement. When we compare two competing products, the comparison is organized around the criteria we think matter to a real player — not around which brand has the larger affiliate rate.

Every factual claim in a SolitaireStack.com article traces back to a source we can name. For game history, that means period rulebooks, Lady Adelaide Cadogan’s patience collections, Dick’s Games of Patience, early Hoyles, and later standards such as The Penguin Book of Patience. For rules and variants, we cross-check Hoyle’s, Pagat.com, and the digital implementation that actually runs in the browser. For numerical claims, we cite the simulation or academic source that produced the figure.

Some claims are legendary rather than documented — for example, the story that Napoleon played patience on St Helena. We hedge those claims in public, explain what is known and what is disputed, and do not present folklore as history.

Different claim types require different evidence, and we hold each to its own bar.

• Win-rate and probability claims are backed by simulation. When we say FreeCell is solvable roughly 99.999 percent of the time, we can point to the run: sample size, solver configuration, treatment of auto-moves, and the confidence interval around the number.
• History claims require a citation. If we cannot find one, we say the claim is disputed or legendary and move on.
• Rules claims are tested against the game engine that actually ships on the network. If our written rule and our implemented rule disagree, one of them is wrong and we fix it before the page ships.

We use AI tools to accelerate research, summarize sources, generate first drafts of rules primers, and spot inconsistencies in long articles. We do not use AI as the final author. Every page on SolitaireStack.com is reviewed, rewritten, and signed off by a human editor on the owning desk before it is published. AI never publishes unreviewed.

In practice that means AI output is treated the same way we treat a first-pass draft from a new contributor: useful as scaffolding, not trustworthy as a final product. Numbers it produces get re-derived. Rules it summarizes get tested. History it recounts gets traced to citation. Where AI assistance materially shaped the angle of an article, we note it.

Articles are updated on a rolling calendar and whenever something material changes in the game, the rules, or our data. Updated pages carry a visible updated date. When we get something wrong, we fix the article, note the correction, and explain what changed. For the full procedure, see our correction policy.

Any affiliate relationship or sponsored arrangement is disclosed plainly on the page where it applies. We do not run paid placements dressed up as editorial reviews. If a link in an article earns a referral commission, we say so at the point of the link. If a partner ever pays to place content on the network, that content is labelled as sponsored, segregated from editorial flow, and held to our other standards for accuracy.

We write for players, not for search engines. We prefer clear language over clever language, short sentences over long ones, and concrete examples over vague maxims. Each desk writes in first-person plural voice (“we”, “our”) because it speaks for the desk, not an individual writer. We try to name things plainly and to show our work when we reach a judgment.

We avoid the two failure modes that plague most writing about card games. The first is overclaiming — telling readers that a strategy “always” works, or that a rule is universal, when neither is true. The second is underclaiming — hiding behind “it depends” when a defensible answer exists. When we know the answer, we give it. When we do not, we say we do not, and we explain the shape of the uncertainty.