Schotten Totten Tactics Statistics: How the Tactic Cards Are Played

What this study measures

This page is a data study, not a strategy guide. What each tactic card does is set by the rules and explained card by card in our guide to Schotten Totten tactics cards. What nobody seems to have counted is how the cards are used once the game is running: how many come out, which ones, by whom, and whether they go with the win. To find out, the site's AI played 400 games of the tactics variant against itself, with the same strength on both sides and a fixed seed. Turns follow the rulebook's order, as in the base game: play a card, claim, then draw. Every tactic card it played was written down with the side that played it, and at the end of each game we noted the winner.

Our other studies, the Schotten Totten stone statistics and the Schotten Totten formation statistics, use the base game. This one is the first to use the tactics deck, with a fresh set of 400 games. One game ended without a winner. Its cards count as played, none of them as played by the winner, and it is left out of the extra-tactic count. An earlier version of this page counted games in which stones were claimed at the start of the turn, the timing of the Schotten Totten expert variant. The site now follows the printed order, every figure below comes from a new run under it, and the comparison with the earlier run is the most telling part of the study.

Two copies of one AI are not two people. They read the board the same way and never bluff, so read the figures as what careful, consistent play produces, not as a law of human play.

How many tactics a game

In all, the AI played a tactic 1,912 times, about 5 a game. The median game used 5 of the deck's 10 cards. All 10 came out in only 4 games, and 7 games saw no tactic at all.

Tactics variant rules: Schotten Totten rulebook, transcribed by RulesPal; figures: the site's game engine, tactics variant, seed 20261011, the same AI on both sides, its search set to 200 iterations; the search only runs in the rare positions where every move risks an immediate loss, so that setting changes almost nothing.

So the tactics deck is a real part of the game, not a sideshow, but it rarely runs dry. In 223 games, a little over half, at least half of it reached the table; in 177, four cards or fewer did. Both sides lean on it about equally, since the rules never let one side get more than one tactic ahead.

Which cards come out

Each card except the Joker comes in a single copy, so it can be played at most once a game. Each player may have only one Joker on their side, so a game sees at most two.

Blind-Man's Bluff leads: it changes how one stone is judged, and the AI found a use for it in three games out of four. Mud Fight and the Recruiter follow, close together, in about two games in three. The Traitor sits at the bottom, in one game in five: it needs an opponent's clan card or elite troop worth stealing, on a stone not yet claimed, and a place to put it.

That last condition is where the claim timing bites. Under the printed order, a stone is claimed in the same turn as the card that completes its proof, so a finished formation is rarely left open through the opponent's turn. The ruses that act on the opponent's side of an unclaimed stone, the Traitor that steals a card and the Banshee that discards one, have far less to aim at. In the earlier run, with the claim at the start of the turn, the Traitor was played 165 times and the Banshee 254 times in 400 games; under the printed order, 79 and 142 times. The Recruiter, which works on your own hand, barely moved: 255 plays then, 254 now.

The Joker leads on raw count, 324 plays, but only because there are two. Each copy reached the table in about 4 games in 10, less often than five of the one-copy cards and more often than the Banshee, the Strategist and the Traitor. A card that is wild in both colour and strength sounds like one you play at once. The study does not record whether the AI held it or never drew it.

Which cards go with the win

The last column of the table above is the share of each card's plays made by the side that went on to win. With both sides played by the same AI, a card with no link to the result would sit near 50%, give or take 6 to 11 points depending on how often it was played: the rarer the card, the wider the band. Across all cards the winner made 51% of the plays.

Five cards stand clear of that band. Three lean to the winner: the Banshee, 89 of its 142 plays, 63%; the Joker, 61%; and the Shield-Bearer, 134 of its 230 plays, 58%, the last only just. Two lean the other way: the Strategist at 31% and the Traitor at 35%. Blind-Man's Bluff, Mud Fight, the Recruiter and the Spy sit within a few points of 50%, which, with nine cards measured at once, is what chance alone would give.

The Banshee is the twist. It came out in only about one game in three, but when it did, the side that played it usually won. With stones claimed as soon as they are proven, it seldom undoes a finished formation; it lands on stones still being fought over, and in our games that suited the side already in front. The Strategist and the Traitor tell the opposite story. A side that is behind reaches for the cards that move or steal: the Strategist pulls a card off a stone that is going wrong, and the Traitor takes one from a stone the opponent is winning. A side that is ahead has less need of them. The study counts who played what; it cannot tell which came first, the card or the trouble.

The extra tactic

By the rules, you may play a tactic only if you have not played more than your opponent, so the two sides always end level or one apart. In 160 of the 400 decided games they ended level. In the other 240, one side had played one tactic more:

The side with the extra tactic won about three games in five. Read it with the same care as above: the study does not show why. One likely reading is the claim timing again. A tactic that completes a provable formation is now cashed in at once, before the opponent can reply, so the extra tactic is more often a finishing blow than a last throw. In the earlier run, with the claim at the start of the turn, the same count went the other way: the side a tactic ahead lost more than two games in three.

What it changes at the table

Three practical readings, in order of confidence.

First, expect about half the deck. In a typical game 5 of the 10 tactics are played, and Blind-Man's Bluff is the one most likely among them: it appeared in three games out of four. A formation that wins on a normal stone may not win on a Blind-Man's Bluff stone, where only the total counts.

Second, a tactic that finishes a stone is worth playing. You claim right after your card, so a Joker, the Spy or the Shield-Bearer that completes a provable formation takes the stone before your opponent can answer there. In our games the side that ended a tactic ahead won more often than it lost. Each tactic still gives your opponent the right to answer elsewhere, so play one when it closes a stone, not just because you can.

Third, do not count on a rescue. The Traitor and the Banshee can only touch a stone that is still open, and with the claim right after the card most finished stones are not. The Traitor reached the table in one game in five, and, like the Strategist, mostly for the side that went on to lose. The rules of the variant are on our page of Schotten Totten tactics cards, and the base game in the Schotten Totten rules.