Spider Solitaire History: Origins, Early Rules and Evidence

Short answer: Spider was already a developed two-deck patience game by the time Morehead and Mott-Smith published detailed rules in 1949. Specialist research points to an earlier, uncredited description in Games Digest in 1937, so 1949 is not a defensible invention date. A frequently repeated 1917 attribution has not been supported by a convincing primary source in the material reviewed. The popular explanation that eight completed runs represent a spider’s eight legs is plausible as wordplay, but its exact documentary origin is unknown. Microsoft later gave the game a much larger computer audience by including it with Plus! 98 in 1998.

Spider has an unusually tidy modern identity: two decks, ten columns, descending builds, eight suited King-to-Ace runs. Its history is less tidy. Search results and modern summaries often compress several different questions into one date: when the game was invented, when it was first written down, when a major rulebook described it, and when a computer edition made it familiar to millions of Windows users. Those events are not the same.

This history separates directly inspected evidence from specialist leads and unsupported repetition. The strongest primary document used here is a 1949 rulebook. A respected specialist source reports a 1937 magazine item, but the underlying pages were not directly reproduced in the sources reviewed for this article. The often-copied 1917 claim lacks adequate documentation. That hierarchy matters: an earlier lead can be credible without being treated as fully verified, while a later printed rule can be certain without proving the date of invention.

For the rules played on PlaySoli now, see the complete Spider rules. For the practical effect of changing the number of suits, see One Suit versus Two Suits versus Four Suits.

Contents

• What counts as evidence for an early game?
• The reported 1937 Games Digest description
• What the 1949 book actually documents
• Why 1949 is not an invention date
• The unsupported 1917 attribution
• Where the name Spider may have come from
• How the rules changed in computer editions
• Microsoft Plus! 98 and mass distribution
• Evidence-based timeline
• Common historical mistakes
• In brief
• Frequently asked questions

What counts as evidence for an early game?

A card game usually exists before a historian can prove that it exists. Players teach rules orally, alter them at home, and use names that later change. Printed descriptions arrive only when an editor thinks the game is worth recording. Consequently, the “first known description” is not automatically the first game, and the first widely circulated book is not automatically the work of the inventor.

For Spider, four kinds of evidence should be kept separate:

1. A directly inspectable primary description. A dated book or magazine page that actually gives the rules is the strongest evidence for what was known by that date.
2. A specialist bibliographic report. A researcher may identify an earlier item that is difficult to access. This is useful, but detailed claims should remain qualified until the item itself is checked.
3. A later recollection or unsourced summary. It may point toward a real lead, but repetition does not make its date reliable.
4. A distribution milestone. A computer release can explain later popularity without explaining the game’s origin.

This article therefore uses “the earliest description located in specialist research” where appropriate, not “the game was invented in.” Exact origin remains unknown.

The reported 1937 Games Digest description

Michael Keller’s specialist research identifies a Spider item in the November 1937 issue of Games Digest, volume 1, number 3, on pages 42–43 SRC-023 SRC-032. He reports it as the earliest reference he had found. The item was apparently uncredited, which means it does not identify an inventor even if its date and rules are confirmed.

That report matters because it moves documented discussion of Spider at least twelve years earlier than the familiar 1949 book. It also changes the correct wording of a history article. A statement such as “Spider was created in 1949” is too strong. At minimum, a game called Spider was described before that date according to the specialist bibliography.

Keller’s comparison reports another potentially important difference: the 1937 description began with 50 tableau cards, while the directly inspected 1949 text deals 54 SRC-022 SRC-023. This detail remains qualified secondary reporting in the present package because the underlying 1937 pages were not directly inspected. It supports the conclusion that Spider’s printed layout was still not completely standardized, but it should not be presented as a fresh transcription of the magazine.

The qualification is equally important. The sources reviewed for this article did not directly inspect a stable scan of the two 1937 pages. It therefore does not reproduce the magazine’s exact wording or strategic advice. The defensible statement is narrower: specialist research reports a 1937 description and a 50-card opening, and the primary pages should be checked before those details are treated as settled.

A useful example of the evidence distinction is this:

• “A 1937 item has been reported by a specialist researcher” is supported.
• “The 1937 rules were identical to PlaySoli Spider” is not supported here.
• “The anonymous 1937 writer invented Spider” is not supported at all.

The magazine lead also leaves open the possibility that players knew the game earlier under the same name or a related one. A publication supplies a latest-possible date for existence, not an earliest-possible date for oral play.

What the 1949 book actually documents

Albert H. Morehead and Geoffrey Mott-Smith included a full Spider section in The Complete Book of Solitaire and Patience Games, published in 1949 SRC-022. Unlike the disputed origin stories, this is a directly inspectable, dated rule text. It establishes that a recognizable form of Spider was sufficiently established for inclusion in a substantial American patience collection.

The opening layout

The book uses two standard decks. It deals 54 cards into ten tableau piles: six cards to each of the first four piles and five to each of the remaining six. Only the top card of each pile is face up. The undealt 50 cards remain for five later rows of ten SRC-022. This is the familiar structural skeleton of modern Spider.

The arithmetic is informative. Two decks contain 104 cards. After 54 enter the tableau, 50 remain. A row places one card on each of ten columns, so the stock supplies exactly five complete rows. That ten-column rhythm is not a later invention of the Windows interface; it is present in the 1949 printed rules.

Building and moving

The 1949 rules allow cards to build downward by rank without requiring matching suits. They also distinguish an ordinary descending build from a same-suit descending sequence. A properly ordered same-suit sequence may move together as one unit, whereas a mixed-suit build cannot be lifted as a whole SRC-022. This is the central tension still associated with Spider: cross-suit placement creates temporary access, but suit continuity creates mobility.

Spaces may receive an available card or an available build. Before another stock row is dealt, however, every empty pile must be filled SRC-022. The modern rule that a new row is unavailable while a column is empty therefore has clear documentary continuity.

Completing a run

The book treats a complete suited King-to-Ace sequence as a removable thirteen-card packet. Once eight such packets have been taken away, the full two-deck game is accounted for. Yet the printed wording differs from many computer implementations: removal is permitted but described as not compulsory SRC-022. A digital game such as PlaySoli can remove a completed run automatically. That interface behavior is a rule implementation choice, not proof that every historical edition handled the moment identically.

Historical strategy language

Morehead and Mott-Smith call same-suit builds “naturals” and prefer them because they remain reversible: the sequence can later be moved intact. They recommend creating a space early and using spaces to rearrange mixed structures into suited ones SRC-022. These observations anticipate modern strategic language about mobility and workspace.

They also discuss winning chances, but the surrounding introductory text explicitly warns that the book’s odds are estimates based on insufficient data rather than scientific measurements SRC-022. Therefore, a number printed there should not be converted into a universal Spider solvability percentage. It refers to the authors’ experience with their rules and their own uncertain estimate, not a solver result for every modern implementation.

Why 1949 is not an invention date

A book can establish publication by a date; it rarely establishes invention on that date. The 1949 Spider chapter does not present a named inventor, a patent-like claim, or a dated account of creation. It presents rules to readers. The reported 1937 magazine item already shows why the distinction is necessary.

Even without the 1937 lead, historians should avoid turning a convenient book date into a birthday. Editors normally collect games that already circulate. Rules may have been taught privately, published in a source not yet digitized, or known under another title. The responsible conclusion is:

• 1949 is a verified influential publication date;
• 1937 is the strongest earlier lead in the reviewed specialist research;
• the exact date and person of invention are unknown.

This is not excessive caution. It is the normal method for histories of traditional games, where documentary survival is uneven.

The unsupported 1917 attribution

A number of modern pages repeat a claim that Spider appeared in a 1917 book associated with Ely Culbertson. In the sources reviewed for this article, that assertion was not accompanied by a convincing scan, stable bibliographic record, or directly inspectable rules page SRC-023 SRC-032. The attribution therefore remains disputed.

The problem is not that 1917 is impossible in principle. Spider could have existed then. The problem is that an exact date and author require exact evidence. A bare statement copied across modern websites cannot substitute for the alleged book. Until a primary item is located and checked, the correct editorial forms are “one repeated version,” “documentary confirmation is insufficient,” or “the exact origin is unknown.”

A future discovery could change the evidence hierarchy. If a genuine 1917 text is found, it should be examined for title, authorship, rule identity, publication date, and whether “Spider” refers to this game rather than another patience. Historical claims should be revisable; they should not be upgraded merely because they are old or popular.

Where the name Spider may have come from

The familiar explanation connects the spider’s eight legs with the game’s eight completed suited runs. It is memorable and structurally appropriate. The 1949 rules already use the name Spider and require the two decks to be removed as eight thirteen-card sequences SRC-022. What the source does not provide is a clear statement that a particular person chose the name for that reason.

That gap changes the claim. “The game has eight runs, like a spider has eight legs” is an observable analogy. “The game was named because of those eight runs” is an etymological assertion. The second needs an early statement from a rule writer, editor, or creator. None was located in the reviewed evidence SRC-023.

Other arachnid names appear in the wider patience literature, which may have encouraged a family resemblance in naming. That still does not identify the first use of Spider or the intention behind it. The best current wording is: the eight-legs explanation is common and plausible, but the exact naming origin is unknown.

How the rules changed in computer editions

Printed Spider and computer Spider share a stable core, yet interfaces can standardize choices that paper rules leave to the player. Three examples show the difference.

First, a printed player must turn exposed cards, keep the stock count, enforce group-movement restrictions, and physically remove completed runs. Software performs those operations automatically. This reduces bookkeeping but can make an interface convention look like an ancient rule.

Second, the 1949 text says a complete suited run may be removed and that removal is not compulsory SRC-022. Modern computer editions often remove it immediately. Automatic removal prevents a finished run from being used as temporary tableau material and makes the game state easier to read.

Third, digital editions can offer one-, two-, and four-suit modes while preserving the same two-deck, ten-column framework. These suit levels are practical difficulty settings: fewer active suits make movable same-suit sequences easier to assemble. They do not prove that one mode is the historical original, and they do not justify a universal win rate. The current distinctions are explained in spider suits comparison.

Computerization also made undo, replay, move counters, hints, and automatic legal-move checks possible. Those features alter how people learn and analyze the game, but they do not erase the underlying rank and suit constraints.

Microsoft Plus! 98 and mass distribution

Microsoft announced Spider Solitaire as one of the games in Microsoft Plus! 98, made available on 25 June 1998 SRC-024. This is a well-documented distribution milestone. It placed Spider on a mass-market Windows add-on at a time when many users were already familiar with Microsoft’s Klondike program.

The event should be described accurately. Microsoft did not invent Spider in 1998: printed evidence existed decades earlier. Plus! 98 instead helped standardize a computer presentation and exposed the game to a much broader audience. Later Windows inclusion and countless web editions extended that digital afterlife.

The distinction parallels the history of computer solitaire more generally. A platform can become the route through which a traditional game is remembered without being the place where its rules originated.

Evidence-based timeline

Date	What can be said	Confidence
Before 1937	Spider may have circulated earlier, but exact origin is unknown	Open question
November 1937	Specialist research reports an uncredited Games Digest description	Qualified lead SRC-023 SRC-032
1949	Morehead and Mott-Smith publish a detailed, directly inspectable Spider ruleset	Verified SRC-022
25 June 1998	Microsoft distributes Spider Solitaire with Plus! 98	Verified SRC-024
Later digital period	One-, two-, and four-suit modes become familiar interface choices	Implementation history varies

The timeline deliberately omits a definitive “invented” row. No reviewed source supports one.

Common historical mistakes

Calling 1949 the year Spider was invented

The date belongs to a verified rulebook, not a documented act of invention. The reported 1937 item is already earlier.

Repeating 1917 without a primary citation

An exact early date needs a scan or a reliable bibliographic record. Repetition across derivative pages is not independent confirmation.

Treating the eight-legs explanation as proven etymology

The analogy is plausible, but no early naming statement was located.

Assuming the Windows edition created the game

Microsoft Plus! 98 was a distribution milestone in 1998, decades after printed rules.

Quoting an old winning estimate as a modern solvability rate

The 1949 authors themselves characterize such figures as guesses. Rules, interfaces, information, and player methods also differ.

Assuming every rule stayed unchanged

Optional manual removal in the 1949 text and automatic removal in current software are a concrete counterexample.

In brief

• The exact inventor and invention date of Spider are unknown.
• Specialist research reports an uncredited Games Digest description from November 1937.
• A directly inspectable 1949 book gives a recognizable two-deck, ten-column ruleset.
• The 1949 rules distinguish mixed descending builds from movable same-suit sequences.
• They require spaces to be filled before another row and allow completed suited runs to be removed.
• Their printed win estimates are expressly non-scientific guesses and should not become universal percentages.
• A repeated 1917 attribution remains unsupported in the reviewed corpus.
• The eight-legs/eight-runs explanation is common, but the exact name origin is unknown.
• Microsoft distributed Spider Solitaire with Plus! 98 on 25 June 1998; it did not invent the game.

Frequently asked questions

Who invented Spider Solitaire?

No inventor has been established by the sources reviewed. The game appears in reported 1937 evidence and a verified 1949 book, but neither supplies a documented act of invention.

What is the earliest description of Spider found for this project?

The earliest located lead in specialist research is an uncredited item in the November 1937 issue of Games Digest SRC-023 SRC-032. The earliest primary rules directly inspected for this package are the 1949 Morehead–Mott-Smith rules SRC-022.

Was Spider invented in 1949?

There is no evidence for that conclusion. 1949 is the publication year of an influential rulebook, and specialist research reports an earlier 1937 description.

Is the 1917 origin story false?

It is better described as unverified. The reviewed sources did not provide a convincing primary scan or bibliographic chain for the precise 1917 attribution SRC-023 SRC-032.

Why is it called Spider?

A common explanation links eight completed runs to eight spider legs. The analogy fits the game, but the exact naming origin has not been documented SRC-022 SRC-023.

Were the 1949 rules identical to modern PlaySoli Spider?

They are close in layout and movement, but not identical in every procedure. For example, the 1949 text makes removal of a completed run optional, while modern software may remove it automatically. See the current PlaySoli rules.

Did Microsoft create Spider Solitaire?

No. Microsoft included it with Plus! 98 in 1998, long after printed descriptions SRC-024. Microsoft’s importance lies in computer distribution.

Can the 1949 book’s winning estimate be used as a solvability percentage?

No. The authors explicitly frame their odds as rough guesses based on insufficient data SRC-022. It is not a solver study and does not establish a rate for PlaySoli.

Related PlaySoli guides

• Spider rules — the exact current legal-move system.
• Spider strategy — how suit continuity and space affect decisions.
• One Suit versus Two Suits versus Four Suits — what the difficulty settings actually change.
• Computer solitaire history — the wider path from mainframes to Windows and the web.
• History of solitaire — the older European context of patience games.

Sources used

• SRC-022 Morehead and Mott-Smith, The Complete Book of Solitaire and Patience Games (1949): directly inspected rules, layout, movement, removal procedure, strategy language, and caution about estimated odds.
• SRC-023 Michael Keller, “Questions about Spider and its strategy”: specialist report of pre-1949 evidence and discussion of Spider’s historical development.
• SRC-024 Microsoft, Plus! 98 availability announcement: official evidence for distribution on 25 June 1998.
• SRC-032 Michael Keller, Solitaire Bibliography: bibliographic cross-check for early and modern Spider references.

Material checked: 2026-07-17.

Disputed or unverified facts: The precise invention date, inventor, alleged 1917 description, full details of the reported 1937 magazine version, and the etymology of “Spider” remain unverified or qualified as stated above.

Editorial responsibility: PlaySoli Editorial Team.