Two-suit Spider: a gap in the published research, measured
Two-suit Spider sits between the gentle one-suit game and the notoriously hard four-suit one. You still play with two packs, 104 cards in ten columns, but now the cards are split evenly between two suits. A card may still go on any card one rank higher, but only a run in a single suit can be moved as a group, and only a same-suit run from King to Ace clears from the table.
We looked for a published figure for how many two-suit deals can be won and did not find one. The peer-reviewed study from the University of St Andrews, by Blake and Gent, covers the four-suit game, where a solver that sees every card wins 98.487% of deals. For two suits, the best we can offer is our own measurement of this site's game, with an honest account of what it does and does not show.
Our computer player, the same one that won 52.8% of one-suit deals, won just 5.7% of 4,000 two-suit deals.
The evidence, laid out
| Measure | One suit (ours) | Two suits (ours) | Published |
|---|---|---|---|
| Games won by our no-peek, no-undo player | 52.8% | 5.7% (95% range 5.0–6.5%) | None for two suits |
| Suits completed per game, out of 8 | 4.96 | 0.82 | — |
| Games reaching at least four completed suits | 56.9% | 6.3% | — |
| Winnable with every card visible | — | Not computed | Four suits: 98.487% (Blake & Gent, 2026) |
What this table does not say is that only one two-suit deal in twenty can be won. Our player is deliberately limited: it cannot see face-down cards and never takes a move back. The St Andrews figure shows that even the four-suit game is almost always winnable with full information. The fall from 52.8% to 5.7% measures how much harder two suits are to play without that information, not how many deals are impossible.
Rules for two-suit Spider here
- Ten columns: four of six cards and six of five, with the last card of each face up.
- A 50-card stock, dealt one card onto every column at a time, five times in all, with no redeal.
- Any card may be placed on one of the next rank up, whatever its suit.
- Only runs in one suit may be moved as a group.
- Any card or run may fill an empty column.
- A King-to-Ace run in a single suit is removed automatically; clear all eight to win.
- Try to deal with a column empty and the game will warn you first.
Why the second suit hurts so much, and what to do about it
Mixed runs are traps
With one suit, every run you build can later move as a block. With two, a heart placed on a spade breaks the run: you can no longer move it together. Each mixed link you build is a small knot you may have to untie later. Prefer same-suit placements whenever there is a choice.
Empty columns are rarer and worth more
Our player emptied a column before the first stock deal in just 0.5% of two-suit games, compared with 1.3% at one suit. When it did, it won 15 of 21. Few things change a two-suit game as much as a clear column early.
Wins take longer
Our player's winning two-suit games took 154.6 moves on average, against 117.5 with one suit. Budget more time and more thought per game.
Untangle before you deal
Each deal buries ten cards. Before dealing, try to leave as many columns as possible ending in same-suit runs, so the new cards land on something you can still move.
Use Undo generously
This is where you can outplay our computer. Explore a line, see what is uncovered, and step back if it closes more doors than it opens.
Two-suit Spider on this site
Two-suit Spider on solitaire-free.uk is free, with no download, account or email address, in any up-to-date browser on computer, tablet or phone. Big clear cards with zoom, Hint, unlimited Undo, Restart, Auto Play, drag or tap to move, and wins, time and moves shown for the game in progress. Sound, fullscreen and table colour can all be adjusted.
Want the gentler version? Go back to one suit. Ready for the full challenge? Try four suits.
Common two-suit mistakes
- Taking the first available placement. An off-suit placement is legal but locks the run. Look for a same-suit alternative first.
- Dealing with loose ends everywhere. Ten new cards land at once. Tidy what you can first, so they land on runs you can still move.
- Using an empty column as storage. It is your best tool for reordering. Use it to swap runs around, then try to clear it again.
How we ran the research, step by step
The solitaire-free.uk team ran this in September 2026. With no published two-suit figure to test against, we have set out the method in full so you can judge the result for yourself.
- Rules from the site's code. The same Spider rules as the one-suit game, with the pack split evenly between two suits (we used spades and hearts). We assumed that dealing from the stock is not allowed while a column is empty; the site warns you, but we did not test exactly what it does.
- Programming the game. Spider was rebuilt in Python and given a computer player.
- Deals the player had never seen. The player's scoring was set using seeds 1 to 30, and the 4,000 test deals used seeds from 100,001 onwards, so the result was not tuned to the test.
- The player. Before each move it searches up to 250 positions ahead using only visible cards, then plays the most promising line. It never undoes. At two suits this player is weak, and its score is a floor, well below what a patient human can do.
- Running it. 4,000 deals.
- Published comparison. None exists for two suits. The St Andrews four-suit figure (98.487% with every card visible) is the closest reference, and it tells us impossible deals are not what holds players back.
- The margin of error. A 95% range is where the true figure very probably sits given how many games we played: for 5.7%, between 5.0% and 6.5%.
- What the figure cannot tell you. It is not how many two-suit deals can be won, and it is not a typical human score. It is how a simple, cautious player without Undo gets on.
Behind the numbers: the code
Our code runs on Python 3.11 and uses nothing outside its standard library. spider.py handles all three suit counts, and a shared runner ran the deals in parallel across 19 processes on a 20-core machine. For two suits the pack is two sets each of spades and hearts. The player uses a best-first look-ahead over visible cards only; it scores positions by rewarding same-suit links and empty columns and penalising face-down cards, then plays towards the best one. Each position it considers is a real, legal Spider position, but any card it would uncover is treated as unknown. The 4,000 games took about fourteen minutes. Percentage ranges use the Wilson score interval.
The team used an AI assistant, Claude by Anthropic, to help write and check the code and to draft this page. The AI did not generate the statistics. They come from running the simulation, and we checked them against the published research before the page went live.
Frequently asked questions
How many two-suit Spider games can be won?
We could not find a published answer. Four-suit Spider is about 98.5% winnable with perfect information, so the true two-suit figure is very likely high as well. Our 5.7% is a no-undo computer's result, not a winnability figure.
Why is two suits so much harder than one?
Because only same-suit runs move together. Every mixed link locks cards in place, and our player's completed suits dropped from 4.96 per game to 0.82.
Should I master one suit first?
It helps. The habits that matter most at two suits, such as early empty columns and uncovering cards, are easier to learn with one.
Does it matter which two suits are used?
No. The rules treat both suits the same way, so the difficulty is identical whichever two are chosen. Our simulation used spades and hearts.
What else is in the same difficulty range?
If you enjoy a stiff, thoughtful game, try Yukon or Solitaire Turn 3. The games page lists everything.
By Admin, solitaire-free.uk. Study date: 18 September 2026.