At first, Hannin wa Odoru looks like a social-reading game. Someone suddenly goes quiet, so maybe they are the culprit.
After a few games, a stronger model appears: the most important thing moving around the table is not a person's expression. It is the Culprit card itself.[1][2]
In a four-player game, treat the table as a tracking problem. Where could the Culprit card physically be now? How many cards does each player have? Since the last confirmed location, did anyone play Rumor, Information Manipulation, or Trade?
Seat order matters too. Let the current player be A, then B, C, D in turn order:
A → B → C → D → A
If A trades the Culprit to D, the player immediately before A in the cycle, one global one-seat movement can send the card straight back to A.
The Culprit card has gone out, touched grass, and come home.
If A sends it to B instead, it takes three ordinary one-seat movements to return. But B acts next. If B has only one card after the trade, B can play the Culprit immediately and win.
So there is no single answer to “Who is safest to trade with?” You must separate return-to-me risk from recipient lethal risk.
1. Four-player setup: there are conflicting secondary descriptions
Sugorokuya states that the 2023 package did not revise the cards or rulebook from the third edition.[1] A 2017 play report explicitly following the included third-edition rules uses five mandatory cards—First Discoverer, Culprit, Detective, Alibi, and Scheme—plus 11 cards from the remaining 27, for 16 cards total.[3]
Some newer secondary guides instead list two mandatory Alibis plus 10 random cards.[4][5]
This guide uses the five-mandatory-plus-11-random setup as the baseline because it is tied to the included third-edition instructions, while also showing Detective probabilities for the six-mandatory-plus-10-random variant. If the printed rulebook in your copy says otherwise, follow your physical copy.
The commonly documented 32-card third-edition set contains: First Discoverer 1, Culprit 1, Detective 4, Alibi 5, Scheme 2, Dog 1, Boy 1, Witness 3, Trade 5, Information Manipulation 3, Rumor 4, Civilian 2.[6]
2. What every card does
| Card | Effect | Four-player meaning |
|---|---|---|
| First Discoverer (第一発見者) | Starts the game and announces the incident | After it is revealed, that player has only three hidden cards, slightly lowering their initial Culprit probability |
| Culprit (犯人) | Can be played only when it is your sole remaining card; playing it wins | A liability early, the winning card late |
| Detective (探偵) | From round two onward, name another player; if they hold the Culprit, the Detective user wins | Stronger late; Alibi can block the accusation |
| Alibi (アリバイ) | While held, lets a Culprit holder deny being the culprit when targeted by Detective | Insurance against a future Culprit transfer; does not stop Dog |
| Scheme (たくらみ) | After playing it, you share the Culprit's win or loss | Easier to judge late when the escape route is clearer |
| Dog (いぬ) | Choose one card from another hand and reveal it; if it is the Culprit, the Dog user wins | Better as the target hand shrinks |
| Boy (少年) | Privately identifies the current Culprit holder | Perfect current-location information that can expire after movement |
| Witness (目撃者) | Privately view another player's entire hand | Can reveal both Culprit and Alibi status |
| Trade (取り引き) | Choose another player; each secretly chooses one card to exchange | Precise Culprit transfer, but the other player can also hand you the Culprit |
| Information Manipulation (情報操作) | Everyone chooses one card and passes it to the left neighbor | Culprit holder can deliberately move the Culprit one seat |
| Rumor (うわさ) | Everyone randomly draws one card from the right neighbor | Cards move one seat in the same direction, but the moved card is random |
| Civilian (一般人) | No effect | A safe discard that does not disturb known information |
The core effects are described by the publisher and detailed third-edition rules summaries.[2][6][7]
This is not normally a team game. The player who wins with Detective, Dog, or Culprit is the winner; Scheme is the special card that joins the Culprit's result.[2][6]
3. Track possible card locations, not suspicious faces
Immediately after First Discoverer is revealed, 15 hidden cards remain and exactly one is the Culprit.
The First Discoverer has three hidden cards, so their public-information probability is:
3 / 15 = 20.0%
Each of the other three players has four hidden cards:
4 / 15 = 26.7%
So the First Discoverer is not automatically more suspicious. With only this information, they are slightly less likely to hold the Culprit.
If your own four cards contain no Culprit and someone else is First Discoverer, the remaining unknown candidate slots are 11:
- First Discoverer: 3/11 = 27.3%
- Each of the other two players: 4/11 = 36.4%
Witness, Boy, and card movement then update this model.
4. One Detective has been used. Are you safe? Usually not
Under the baseline setup, the mandatory Detective is already included. The remaining 27-card random pool contains three more Detectives, and 11 cards are sampled from that pool.
| Total Detectives in the game | Probability |
|---|---|
| 1 | 19.1% |
| 2 | 45.1% |
| 3 | 30.1% |
| 4 | 5.6% |
There is an 80.9% chance of at least one additional Detective. The expected total is about 2.22 Detectives.
For the two-Alibi-fixed, 10-random variant, the corresponding probabilities are 21.5%, 46.2%, 27.7%, and 4.6%; at least one extra Detective appears 78.5% of the time, with an expected total of about 2.15.
Either way, seeing one Detective discarded is not a green light for the Culprit.
5. Why Detective becomes so strong against a one-card hand
A one-card player creates two simultaneous emergencies.
First, if that card is the Culprit, they can play it on their next turn and win.[2]
Second, a one-card hand cannot contain both Culprit and Alibi.
At two cards, Culprit + Alibi is possible, so a correct Detective accusation can still be blocked. At one card, that defense is physically impossible.
A particularly strong Detective timing is therefore:
- you have evidence the player holds the Culprit;
- they have exactly one card;
- no Culprit-moving effect has occurred since the last reliable location information.
“One card left” means both no Alibi shield and lethal next turn.
6. Early Detective has another problem: Culprit and Alibi often start together
If the baseline deck is sampled randomly and dealt randomly, the probability that the initial Culprit holder also starts with at least one Alibi is about 45.5%.
In the two-fixed-Alibi variant it is about 52.9%.
An early Detective must therefore solve two problems: identify the correct holder and get past the Alibi possibility. Late in the game, discarded Alibis and one-card hands remove that second obstacle.
7. Dog is a 1/k shot once the holder is known
If you know with certainty who holds the Culprit and that player has k cards, Dog succeeds with probability 1/k.[6][7]
| Target hand | Dog success |
|---|---|
| 4 cards | 25% |
| 3 cards | 33.3% |
| 2 cards | 50% |
| 1 card | 100% |
Dog also bypasses Alibi because it reveals a specific card rather than asking the holder to deny the accusation.
If the holder itself is uncertain, multiply the probability that the player holds the Culprit by 1/k.
8. Rumor gets more dangerous as hands shrink
Rumor makes each player draw one random card from the right neighbor.[6] If the Culprit holder has k cards, the chance that the Culprit moves is 1/k.
| Culprit holder's hand | Culprit moves |
|---|---|
| 4 | 25% |
| 3 | 33.3% |
| 2 | 50% |
| 1 | 100% |
This is why “I know the culprit and they have one card left” can be destroyed by a Rumor before your Detective turn. Their one remaining card must move.
Boy and Witness information also has an expiration date: after Trade, Information Manipulation, or Rumor, downgrade old location knowledge.
9. Rumor and Information Manipulation move cards in the same seat direction
With seats A → B → C → D → A, Information Manipulation passes a chosen card to the left neighbor, while Rumor draws a random card from the right neighbor.[6]
From the card's perspective, both produce one-seat movement:
A → B → C → D → A
The difference is control. Information Manipulation is strategic; Rumor is random. Do not model Information Manipulation as a coin flip. A rational Culprit holder chooses what to send.
10. Who should receive the Culprit in a Trade?
B, the next player: the Culprit is three directional steps away from returning to A. That is excellent for return distance. But B acts immediately, so a one-card B can win instantly, and a B holding Trade can immediately send the problem back.
C, opposite: two steps from returning, and acts after B. It is the middle seat on both risks.
D, the previous player: D's own turn is latest, reducing immediate recipient lethal risk. But the Culprit is only one directional step from A. One Rumor can randomly return it with probability 1/k, and Information Manipulation lets D deliberately pass it straight back.
| Recipient | Directional distance back to A | Recipient's next turn | Main risk |
|---|---|---|---|
| B | 3 seats | immediate | recipient lethal or immediate Trade retaliation |
| C | 2 seats | middle | balanced risk |
| D | 1 seat | latest | one global move can return the Culprit |
The right choice depends on hand counts and remaining movement cards. If movement cards are still plentiful, D is dangerous. If movement cards are mostly gone and D has several cards, delaying D's turn can be attractive. Never hand a one-card B the Culprit unless their immediate win is acceptable—for example, after you have already played Scheme.
11. Culprit strategy changes completely with Alibi
Without Alibi, moving the Culprit away through Trade or Information Manipulation can protect you from Detective, but you must avoid creating an opponent's immediate win.
With Culprit + Alibi, holding can be strong. A hand such as Culprit, Alibi, Civilian often wants to discard Civilian first and preserve Alibi. Civilian suddenly becomes useful because it reduces your hand without altering card-location information.
Once you reach Culprit + Alibi, you cannot play Culprit yet. After you finally discard Alibi, the gap until your next turn is your vulnerable window.
For everyone else, a player who discards Alibi and drops to one card deserves immediate attention.
12. Boy, Witness, Scheme, and Civilian
Boy gives perfect information about the current holder, but movement can immediately make it stale. It becomes more convertible into a win after movement cards have been spent.
Witness is extremely strong with four players: no Culprit means that player is cleared for the moment; Culprit without Alibi creates a Detective target; Culprit with Alibi suggests waiting, using Dog, or watching for the Alibi to leave.[6]
Scheme is often easier to justify late. Early, you are betting on a Culprit future you barely understand. Late, fewer Detectives and movement cards plus a short Culprit hand make the escape route easier to evaluate.
Civilian is not simply “bad.” It is a do-nothing safely card. In a game where many effects destroy your own information, that is real utility.
13. What to count in the discard pile
You do not need to memorize everything. Start with:
- Detectives
- Alibis
- Rumors
- Trades and Information Manipulations
Then add Dog, Boy, and Witness when comfortable.
With only 16 cards used in four-player games, “I haven't seen Dog yet” does not mean someone has it. Under the baseline random selection, the probability that at least one copy enters the game is approximately:
| Card type in random portion | At least one included |
|---|---|
| Dog | 40.7% |
| Boy | 40.7% |
| extra Scheme | 40.7% |
| Civilian | 65.8% |
| Witness | 80.9% |
| Information Manipulation | 80.9% |
| extra Detective | 80.9% |
| Rumor | 89.6% |
| extra Alibi | 89.6% |
| Trade | 94.6% |
This is why “Dog has not appeared, so somebody must hold it” is bad four-player reasoning, while expecting at least one Trade is much more reasonable.
14. The practical turn checklist
On your turn, ask:
- Who can still physically hold the Culprit?
- When was its location last confirmed?
- Which movement effects occurred since then?
- How many cards does everyone have?
- Is anyone at one card and about to act?
- Who has already discarded an Alibi?
- How many Detectives and movement cards appear to be spent?
- Is your own win route Detective, Dog, Culprit, or Scheme?
The key shift is simple:
Do not merely guess the suspicious person. Maintain the possible location of the Culprit card and use your winning card one move before lethal.
For four players, Hannin wa Odoru is less like Werewolf and more like parcel tracking—except the parcel is the criminal.
References (7)
- すごろくや「犯人は踊る」製品ページ。第三版の変更点、2023年新パッケージでカードデザイン・説明書の改訂なし、内容物32枚。 sugorokuya.jp
- SUPPE / すごろくや「犯人は踊る」案内。ゲーム開始、探偵、アリバイ、取り引き、犯人の勝利条件。 suppe.sugorokuya.jp
- GIGAZINE「『犯人は踊る』をプレイして…」2017-09-21。第三版の同梱説明書に従う4人構成として、第一発見者・犯人・探偵・アリバイ・たくらみ各1枚＋ランダム11枚を記載。 gigazine.net
- トマラボ「『犯人は踊る』のルール解説＆レビュー」。4人構成をアリバイ2枚固定＋ランダム10枚と記載する近年の二次資料例。 tomatablog.com
- 黒豹ボドログ「犯人は踊る」。同じく4人構成をアリバイ2枚固定＋ランダム10枚と記載する近年の二次資料例。 kurohyo0803.com
- Kumax「“犯人は踊る” のルール、カード一覧」。第三版のカード枚数、各カード効果、4人構成、うわさ・情報操作の方向。 kumakumax.hatenablog.jp
- りりぃのボードゲーム体験記「犯人は踊る」。探偵・いぬ・少年・情報操作・うわさ・取り引き等の効果整理。 boardgame-diary.com
