Kill Doctor Lucky Game Tactics And Strategy

Purpose And Scope Of This Post

This post goes over some tactics and strategy for the Kill Doctor Lucky game (19.5th anniversary edition, rules here) that are not obvious at first glance.  I think Kill Doctor Lucky has a lot of strategic/tactical depth that is easy to not realize.

A lot of this post is applicable to games with any number of players, but there is special attention paid to two-player games with strangers. A lot of this post is applicable to any board, but we will also analyze particular boards in order to come up with general things to keep in mind when analyzing any board.

This post is a work in progress.  I made a video of a lot of these concepts, and some sample games: game1, game2, game3.  I made a deterministic Kill Doctor Lucky game where you can play and see what the AI thinks is a good move (sorry, it's command line).

Basic Terminology

"P4" is shorthand for "player 4"; "Dr" and "dr" are shorthand for "the doctor".  In two-player games, the players are {P1, p2, P3, p4} so that it is more obvious that {P1,P3} are the normal (human) players and {p2,p4} are the strangers.  Strangers will often be referred to as "players" even though their moves are always the simple stranger algorithm.

A player is "activated" when Dr moves into a room and the next turn becomes that player's turn.

"Looting" is where a player draws a card for their action phase because no one can see the room the player is in.

"Adjacent rooms" means the rooms that are one step away from the room of concern.  On the main board, the foyer's (R24) adjacent rooms are drawing room (R1) and piazza (R23).  Adjacency is about where a player can move in one step, not about sharing a wall on the board.

"Next room" means the room that is next in doctor-visiting order.  On the main board, the dining hall's (R4) next room is the sitting room (R5).

"Riding the Lucky train" or "dr-riding" is where a player repeatedly moves to the room that Dr will move to next, thus activating that same player to take another turn.

A "sight-block" is when a player's position is preventing another player from attacking Dr because of sight.

A "ride-block" is when a player's position is preventing or limiting another player's ability to dr-ride.

A "combo-block" is when a player's position performs both a sight-block and a ride-block on another player.

"P1@R16" and P1@16 are shorthand for "player 1 is in room 16".  So if I say to "imagine {Dr@10, P1@11, P2@12}", that means to imagine those pieces at those rooms.

As I explain tactics, I'll introduce more terms.

Cards

Deck Analysis

The game has 72 cards: 24 failure cards, 24 weapon cards, 24 move cards.  The main board has 24 rooms, and each room has a corresponding weapon and move card (for the special attack or special destination).

Some summary statistics about the card types:

• Failure cards
• Avg of 2.08 clovers.
  
• 42% 1-clover cards.
• 25% 2-clover cards
• 17% 3-clover cards.
• 17% 4-clover cards.
• Weapon cards
• Avg of 1 clovers; evenly split between {0,1,2} clovers.
  
• Avg of 2.21 normal attack; 79% 2-attack, 21% 3-attack.
  
• Avg of 4.83 special attack; 42% 4-attack, 33% 5-attack, 25% 6-attack.
  
• Move cards
• Avg of 1 clovers; evenly split between {0,1,2} clovers.
• Avg of 1.29 steps; 71% 1-step, 29% 2-step.
• All cards
• 78% of cards have at least 1 clover.
• 42% of cards have at least 2 clover.
• 11% of cards have at least 3 clover.
• 5.6% of cards have 4 clover.
• Avg of 1.36 clovers.
• Avg of 0.43 steps.
• Avg of 0.74 attack.

Some notable observations and possible decisions based on these raw facts:

• In a two-player game with strangers, a stranger will contribute at least 1 clover 78% of the time, at least 2 clover 42% of the time, etc.  Keep these facts in mind when deciding how likely your next attack is to win the game, and when you are defending an attack.
• Players usually defend attacks with their failure cards first, then will resort to weapon and move cards.  If a player defends with a move or weapon card, they probably have no failure cards and only have weapon and move cards left.  Weapon and move cards have 1 clover on average, and you can better estimate how many clovers a player has left.
• Think about an opponent with only weapon and move cards defending against your attack.  If you use a 3-attack weapon instead of your bare hands, your opponent will likely have to discard an extra ~3 weapon or move cards (they average 1 clover), severely weakening their future abilities to move/attack.  That's a great deal in a two-player game, and okay in a 8-player game.  Also, it can matter greatly who attacks before who, because the earlier attack can weaken or prevent the following attack.
  
• When considering between a move that would allow looting and a move that gives non-looting benefits, you now know better what benefit to expect from looting.

Some other card-related thoughts...

If you never seem to be able to use any of your weapons' special attacks, then maybe you should focus more on dr-riding and less on attacking Dr as soon as possible.

 

Clover Guessing

Let's work through an example of estimating clovers throughout a game...

• T1: Opponent starts out with 6 cards.
• Cards have 1.36 clovers on average.  6*1.36 = 8.16, so let's guess they have 8 clovers.
• Alternatively, we could guess a very typical hand of {2 failure, 2 move, 2 weapon}.  This approach also guesses 8 clovers.
  
• T2: Opponent defends against a 3-strength attack by discarding a 3-clover failure card.  5 total card remaining.
  
• Opponent is now down 1 failure card.  Maybe their remaining cards are {1 failure, 2 move, 2 weapon}, so maybe 6 clovers remaining.
  
• T3: Opponent defends against a 2-strength attack by discarding a 2-clover weapon card. 4 total cards remaining.
  
• Opponent is probably out of failure cards.  With 4 weapon/move cards remaining, probably 4 or fewer clovers remaining.  I say "or fewer" because as people discard their high-clover cards, they are left with their low-clover cards.  Let's guess 3 clovers.
  
• T4: Opponent defends against a 2-strength attack by discarding a 1-clover weapon card and a 1-clover weapon card. 2 total cards remaining.
• Opponent probably only has 1-clover and 0-clover weapon/move cards remaining.  So 0-1 clovers remaining is a good bet.
  
• T5: Opponents is unable to defend against the next attack and shows they have two 0-clover move cards.

So, how did we do in our clover guesses?

• T1: guessed 8, was 7.
• T2: guessed 6, was 4.
• T3: guessed 3, was 2.
• T4: guessed 0-1, was 0.

Guessing accuracy is most important when someone is running low on cards and we need to know if the next attack will win the game, so it's good we did well on our {T3, T4} guesses.

Blocking Basics

You win by attacking the doctor, but there usually is some benefit in making it more difficult for an opponent to do things like attacking or Dr-riding.  Blocking is more important in two-player games than many-player games, but still important even then.  For instance, if you're P4 in a four-player game, you're the last player to defend when P1 attacks the doctor, so the burden of defence falls disproportionately on you, so you benefit much more from blocking P1 than from blocking P3 (where you are first to defend and usually can contribute nothing to the defense).

I will talk about various types of blocking, and how the layout of rooms affects blocking.

Sight-Block, Ride-Block, and Combo-Block

Imagine you're player 2 (P2) playing a game on the alternate downstairs board and {Dr@3, P1@4}.  You can imagine P3 in R10 if you want - not important.  It's P2's turn and P2 has unlimited move cards.  What is the ideal room for P2 to move to in order to make things difficult for P1?  Remember that after P2 moves, Dr will move to R4, and activate P1, even if P2 moved to R4.

P2's turn.  What is best place for P2 to move to?

 The answer is that P2 moving to R5 causes the most headache for P1.  First of all, when Dr activates P1 in R4, P1 will not be able to attack that turn, but the same can be said if P2 moved to R2, R3, or R4.  The unique important fact about R5 is that R5 is R4's next room (Dr will visit R5 after R4).  If P2 is in R5, then P1 must do something other than attack, and P2 is guaranteed to be active next turn. Or to put it in succinct terms: P2 moving to R5 is a combo-block of P1.

P1's turn.  P2 is preventing P1 from attacking and P2 will have the next turn.

If P2 had moved to any of {R2, R3, R4}, then P2 would have sight on R4 and prevented P1's attack in R4, but P1 could simply dr-ride to a room that is not seen by P2, and then attack in that room.  P2 moving to R5 is a ride-block in that it blocks P1 from doing any dr-riding.

 

Combo-Blocking, Cheap Vs Expensive

If combo-blocks are so effective, how does anyone ever attack Dr?  Because performing a combo-block might require spending move cards, and is not always worthwhile or possible.

Imagine a game on the alternate downstairs board where P1 and P2 are just going to repeatedly combo-block each other ("dueling") if they can't attack.  Just before Dr moves, {Dr@1, P1@2, P2@3}.  So, when Dr moves, P1 will be activated, and will be combo-blocked by P2.  The picture just below shows the players' positions over the next few turns.

P1 and P2 positions over time as they duel.

Turns (Dr details omitted, but he activates someone every turn):

• P1 moves from R2 to R4, costing a move card.
• P2 moves from R3 to R5, using just the free move point.
• P1 moves from R4 to R6, costing a move card.
• P2 moves from R5 to R7, using just the free move point.
• P1 moves from R6 to R9, costing a move card.

Notice how P1 used 3 move cards but P2 didn't use any.  Why?  Because P2 was repeatedly in rooms that only require one step to go to the next-next-room and P1 was repeatedly in rooms that require two steps to go to the next-next-room.  So, if two players are going to start dueling, the one who starts in R3 instead of R2 is going to be at an advantage.

So, here's something that has a lot of consequences that are not obvious: whether a room can go to the next-next-room in one step (a "next-next-cheap" room) is an important attribute of the room.

Also, notice that P1's move to R9 combo-blocked P2 in R7 even though R9 is not R7's next room.  R9 has sight on R7 and R8, so P1 being in R9 still prevents all attacks before R9.  Combo-blocking is often about moving to the room just after your opponent, but not always.  If P1 had moved to R8 instead of R9, then P2 could have moved to R9 to again combo-block P1 without using a move card.  But, because P1 moved to R9, then P2 would have to use a move card to move to R10 to combo-block P1.

 

Next-Next-Cheapness, Implications On Earlier Rooms

Let's imagine another scenario where P1 and P2 both start out in next-next-cheap rooms, and they know whoever attacks next will win the game...

• It's P1's turn.
• Dr@11, P1@11, P2@12. (So P2 has sight on Dr.)
  
• No one has any move cards.
  
• Who will get to attack first?
  

P1's turn.  {Dr@11, P1@11, P2@12}.  Who will win the duel?

The answer is P2.  Here's how things can go:

• P1 moves to R13.  Dr moves to R12, activating P2.
• Note: if P1 moves to R14, P2 will be able to attack in R12.
  
• P2 moves to R14.  Dr moves to R13, activating P1.
• The disadvantage of P1 is now apparent.  P1 does not have the move cards to move from R13 to R15.  P1 staying in R13 or moving to R14 still leads to P2 getting activated, and P2 can then dr-ride to attack somewhere else.
  

So, it's good to end up at R14, which gives advantage to R12 and disadvantage to R13 and R11 in a two-person duel.  Multi-person duels are more complicated and depend on the number of people.

 

Next-Blind Room Pairs

Next-Blind Room Pairs And Looting

Most rooms have sight on the next room.  That means that if you are dr-riding, you usually can't loot any room.  There are exceptions.

On the main board, there are three pairs of rooms where a room does not have sight onto the next room, and therefore you can loot while dr-riding.  These room pairs are: {R10, R11}, {R18, R19}, {R23, R24}.

Main board; arrows indicating next-blind room pairs.

Imagine you're playing a game on the full main board; all of your opponents are in R13; you and Dr are in R17. You could dr-ride to R18 and attack in R18.  Another possibility is dr-riding all the way to R11, which draws 3 cards, then attack Dr in R11.

On the alternate board, the next-blind room pairs are:

• Full board: {R15, R16}, {R28, R1} (when Dr switches floors)
• Upstairs only: {R28, R16}
  
• Downstairs only: {R15, R1}

Alt board; arrows indicating next-blind room pairs.

 

Next-Blind Room Pairs And Blocking

Usually, if your opponent and Dr are going to be in some room, you can combo-block your opponent by moving to the next room.  Next-blind room pairs mess up combo-blocks.

Let's imagine it's P1's turn and {Dr@17, P1@17, P2@18}.  P1 can't attack because P2 has sight on P1.  What are some things that P1 can do to make things difficult for P2?

If P1 moves from R17 to R19, P2 can attack in R18.

P1 could ride-block P2: if P1 moves to R19, then P2 can't dr-ride to other rooms.  However, P1 will not have sight on R18, and P2 will get activated and will be able to attack Dr in R18.

P1 could sight-block P2: if P1 moves to a room that has sight on R18 (ex: R4, R17, R18), then P2 won't be able to attack in R18, but P2 can dr-ride to another room and attack Dr there (ex: R21).  And it might be a long time until P1 can take another turn or get close to Dr.

So, because R18 is not seen by its next room, you can't combo-block R18 by yourself.  If P3 was in R4 (which has sight to R18), then P1 can move to R19 and P2 will be combo-blocked (sight-blocked by P3 and ride-blocked by P1).

So, there are advantages to being R18, a room that can not be seen by it's next-room.  That has implications for advantages and disadvantages of being in rooms before R18.  For instance, imagine...

• it's P2's turn
• P1 has no move cards
• Dr@15, P1@16
• P2 has unlimited move cards
• P2 can move anywhere you wish.

Can you choose a room for P2 that will allow P2 to attack before P1 gets to attack? Or at least prevent P1 from attacking in one of rooms {R16, R17, R18, R19}?

 

P2's turn, Dr@15, P1@16.  Where can P2 move to prevent P1 attacking over next 4 turns?

I think that no matter where P2 moves to, P1 can attack Dr in a few turns.  In order to prevent P1 from attacking immediately, P2 must sight-block P1.  In order to prevent P1 from dr-riding to an attack-able room, P2 must ride-block P1.  R17 is the only room that comb-blocks R16, so we must place P2@R17.  Unfortunately, P1 can move to R18 without using a move card, and then Dr will move to R17, and it will be P2's turn.  P2@R17 is combo-blocked by P1@R18 and we previously showed how a single player can't combo-block a player in R18.  So, P1 will either attack in R18 or R19.

So, the advantage of R18 causes a disadvantage for R17 and an advantage for R16.

Many-Player Games

Stealing Activations

If an opponent is ride-blocking you, it may take a painful amount of move cards to turn the tables and ride-block them.  Sometimes, it is easier to prevent Dr from activating them.  Imagine...

P1's turn.  What can P1 do about P2?

• P1's turn.  Dr@2, P1@2, P2@4, P3@22.
• P1 moves to R4.  Dr moves to R3.
• P2's turn.
• If P2 stays in R4, P1 will get activated.
• So, P2 moves to R5 in order to combo-block P1.
• Dr moves to R4, activating P1.
• P1 has good options, like moving to R6 or R13.
  

There are two key ingredients to this technique:

• P1 moved to the next-next-room (R4) so that Dr would advance to R4 after P2's turn, meaning P2 won't win any activation contests in R4.
• This maneuver only works against the player right after you in turn order.  If P3 and P2 were swapped (so P3@4), P1's maneuver would not steal an activation from P3.
  

Also, note that this maneuver could be done even if P1 and Dr started out in R23.  P1 would tight-ride until R2, then move to R4.

TODO: More Techniques

Two-Player Games

Rule Variation For Attacks By Strangers ("Attacking Ally")

The official 19.5th anniversary edition rules for two-player games say that if a stranger successfully murders Dr, then the stranger wins (and presumably both human players tie).  I much prefer to play with the "attacking ally" rule change: if a stranger successfully murders Dr, then human to their left wins.  This way, a stranger to your right is both an ally in terms of giving you turns and murdering Dr.  Your allied stranger still defends against your attacks, and you still block each other from attacking via sight; the only change is who wins when a stranger murders Dr.  I prefer this "attacking ally" rule change for the following reasons:

• With the original rules, I don't like the scenario where you and your human opponent are having a battle of wits, and you've managed to be cleverer and get in a lot more attacks than your opponent...then your allied stranger stumbles upon Dr, attacks, and you end up tying.
  
• With the rule change, you end up caring more about all the pieces on the board, and are encouraged even more to do deep planning.

The following tactics are ones I discovered while playing with the "attacking ally" rule change, but they are still relevant with the official rules.

 

Rule Variation For Players Moving Strangers ("Nosy Strangers")

The official rules also allow a human player to move a stranger during the human's move phase.  I love the extra possibilities and clever planning this allows, but I think there should be one restriction.  The "nosy stranger" rule: you can't move a stranger that can see you and then you attack in the same turn.

With the official rules, strangers are very bad at sight-blocking.  If a stranger has sight on a room you were hoping to attack in, the solution to that problem is usually trivial.  Just get activated by Dr, then with your free move step, move the stranger out of sight, and then attack.  Here's a main board example:

• Start out with Dr@15, P1@16, p2@21, P3@7, p4@22.
• Let's pretend it's extra important for P3 to make sure that P1 doesn't attack in R16.  Can p2 help at all?
  
• It's P3's turn.
• P3 does not have enough move cards to move P3 to sight-block P1, so P3 moves p2 to R17 in the hopes of making things difficult for P1.
  
• Dr moves to R16, activating P1.
• P1's.
• P1 uses its free move step to move p2 to R4 (or right back to R21 if you wish).
• P1 attacks.
• Dr moves to R17.
  

 P3 moves p2 to R17 in order to combo-block P1, which doesn't work.

I'm not endorsing the choices of P1 and P3; it's just an example to show how "I'll use a stranger to sight-block my opponent" is often trivial to counteract with the official rules. In a related example, imagine at the start of P1's turn, P1 and Dr are in R13 and p2@17; p2 is no obstacle to P1 attacking in any of the rooms other than R17 itself.  P1 can simply dr-ride to any room before R17, then use its free move step to move p2 out of sight, then attack.

So, if strangers are very weak at blocking attacks, it's a bit like they don't exist or matter as blockers, and the board easily becomes "too big" where it's too easy to attack.

My proposed rule change of "you can't move a stranger that can see you and then attack in the same turn" makes it so that you care much more about the position of strangers, and you have to plan ahead more rather than getting away with "I can ignore strangers and then use my free move step on a stranger at the last moment".

I'll be playing with the "nosy stranger" rule in my following examples, but the demonstrated techniques are still relevant without the rule change.  And the multi-attack techniques are probably even stronger without the "nosy stranger" rule.

 New Way To Dr-Ride: Loose-Riding

In many-player games, where all players are human, dr-riding usually requires a player to move to exactly the next room that Dr will move to.  For instance, imagine {Dr@2, P1@2, P2@12, P3@12} and it is P1's turn.  P1 has to move to an empty R3 to take another turn without any opponents taking any turns.  If P1 moves to R4, then Dr will move to R3, and it'll be P2's turn, and P2 will probably do something that disadvantages P1 (like move to R5, combo-blocking P1).

The existence of strangers means that Dr moves twice instead of once before it becomes a human opponent's turn (assuming the Dr doesn't activate anyone along the way). Reminder: in a two-player game, {P1,P3} are human and {p2, p4} are strangers.

So, imagine a two-player game with Dr and P1 in R2, everyone else in R12, and it's P1's turn.  If P1 moves to R3, then it'll become P1's turn again, just like in a many-player game.  The new development is that P1 can move to R4, and P1 will be activated in R3 before human P3 gets to have a turn.  We'll call this technique "loose-riding" for when a player moves to the next-next-room and a stranger gets to take a turn before the player is activated again.  And "tight-riding" can refer to moving to the next-room and no one gets to take a turn before the player is activated again (previously we just called this "dr-riding").  Let's look at what happens if P1 decides to do a bunch of loose-riding on the main board:

• Setup: Dr and P1 are in R2 and it is P1's turn.
  
• P1 moves to R4 and loots.  Dr moves to R3.
• p2 moves to R11.  Dr moves to R4, activating P1.
  
• P1 moves to R6 and loots. Dr moves to R5.
• p2 moves to R10. Dr moves to R6, activating P1.
• P1 moves to R8. Dr moves to R7.
• p2 moves to R9. Dr moves to R8, activating P1.
• P1 moves to R10. Dr moves to R9, activating p2.
• p2 can not attack and moves to R8.  Dr moves to R10, activating P1.
  

P1 loose rides from R2 to R10; p2 is activated by Dr in R9, but is seen by P1 in R10.

From the example you can see that P1 was able to loot in {R4,R6} and would not have been able to do so with tight-riding.

Also, note that when P1 loose-rides, p2 moves "closer" to Dr.  If P1 had tight-rode to R10, p2 would have made no moves and would have still been at R12.  Because P1 loose-rode, Dr and p2 encounter each other in R9.  However, because P1 was combo-blocking p2, p2 made no attack and then P1 was activated without P3 having a turn.

So far, loose-riding has the benefit of letting you loot more often, but has the disadvantage of moving the stranger-after-you closer to Dr.  Sometimes moving the stranger-after-you is actually an advantage.  For instance on the main board, if P1 does one loose-ride iteration while p2 is in R21, then p2 moves to R20, which means p2 can see fewer rooms and it'll be harder for P3 to move p2 to somewhere useful with a single step.

Sidenote: you can combine loose-riding with next-blind rooms for more looting.  Imagine the main board with P1 and Dr in R16, everybody else in R1.  P1 can move to R18, loot, get activated after p2's turn, then move to R19 to loot again.

Multiple Attack Planning

In two-player games, the ability to move strangers allows for some very interesting multi-turn plans to attack Dr multiple times, often without your human opponent taking a turn at all.

In these plans, it's important to distinguish between your allied stranger and your opposing stranger.  The allied stranger is the stranger to your right and takes a turn just before you.  The opposing stranger is the stranger to your left and takes a turn just after you.

Your allied stranger taking a turn has a few advantages for you, even with the official rules:

• If the allied stranger attacks, then the burden of defending disproportionately falls on your human opponent.  You can probably pass (discard no cards), and your human opponent will probably have to defend against the entire attack, minus a clover or two contributed by the other stranger.
  
• After your allied stranger takes a turn, you will take a turn, unless Dr activates someone else.
• Bonus! When Dr moves to your room after your allied stranger's turn, you're guaranteed to be the one activated, no matter who else is in the room.
  

In the following multi-attacks, it'll be common for a human player to dr-ride until they get to the room they want to attack in.  Dr will activate the player in that room, and a good use of the player's free move step is to move a stranger (usually their allied stranger) to set up further activations or attacks.

Bounce Attacks

A bounce attack is where you attack, your allied stranger gets activated (maybe even attacks), then you maneuver so that you can make another attack, possibly without your human opponent ever taking a turn.  Also, if you manage to attack, then your allied stranger takes a turn, then you get another turn to make movements favorable to you, I would still consider that a bounce maneuver (you're bouncing turns between yourself and your allied stranger).

Imagine a game on the alternate downstairs board...

• P3 just attacked Dr in R15, then Dr moved to R1, activating P1.
• The strangers start at p2@13, p4@3.
• P1 wants to attack multiple times before P3 takes another turn.
  

P3 just attacked, and now it is P1's turn. Dr@1, P1@1, p2@13, P3@15, p4@3.

 

P1 could do this...

• P1 does not use their free move step.  P1 attacks.  Dr moves to R2.
• p2 moves to R12.  Dr moves to R3, activating P4.
• p4 attacks.  Dr moves to R4.
• P1 moves to R5 (via the stairway). Dr moves to R5, activating P1.
• P1 tight-rides all the way to R10. Dr moves to R10, activating P1.
  
• P1 moves p4 to R12 (not labelled in picture) and P1 attacks. Dr moves to R11.
  
• p2 moves to R11 and can not attack. Dr moves to R12, activating p4.
• p4 can not attack and moves to R11. Dr moves to R13.
• P1 moves to R1, ride-blocking P3.  If P3 decides to attack in R15 again, P1 could probably do another bounce-attack.
  

Notice that P3 attacked once, and then P1 attacked, p4 attacked, and then P1 attacked again.  Was it because I chose a rare scenario that gives an opportunity to bounce attack?  No.  The core ingredient of a bounce attack (or any multi-attack) is that a player has the ability to dr-ride, and has a lot of space to do so before Dr activates their human opponent.  Once P3 attacked in R15, that gave P1 the opportunity to...

• Dr-ride up to P1's allied stranger and attack before the stranger can see Dr (in this example, the dr-riding was not needed, P1 attacked immediately).
• Get P1's allied stranger activated (possibly attacking), which leads to P1 getting the next turn.
• Then more dr-riding by P1 to get another attack or set up a favorable board state.
  

You want to attack such that your opponent can't swiftly start attacking and dr-riding to set up attacks before it's your turn again.  One way to safely attack is that after you attack, your allied stranger gets activated, then you use your following turn to give your opponent as little dr-riding space as possible before the Dr will activate you again.  (In the example above, p4 getting activated in R12 allowed P1 to move to combo-block P3.)  Also, once you've exhausted your opponent of move cards, there are scenarios where you can attack and your opponent can't attack or dr-ride at all before you get another turn.

Also, notice that when P1 first attacked in R1, they could have used their free move step to move p4 or p2 (maybe if P1 had moved p2 to R14, things would have gone even better).  Later, P1 attacked in R10, P1 used its free move step to move p4 to R12.  This movement of p4 both prevented p2 from attacking and set up a future activation of p4 so that P1 can take another turn.  Whenever Dr activates you and you are thinking about attacking (thus you won't be moving that turn), consider using your free move step to move a stranger.

Repeated Stranger Attacks

Once you attack, your human opponent is going to get a turn soon unless you've set up your allied stranger to get activated.  So, it's a struggle to keep your opponent from getting a turn after you attack.

One of the great things about your allied stranger attacking is that the next turn is going to be yours (unless someone gets activated right after the attack, which is rare).  You can use this turn to get yourself into a good position, but what if you use your turn to move your allied stranger so it'll get another attack?  And then another, and another?  Here's a main board example:

P1 uses 1 move card and p4 attacks 8 times.

• Dr@16, p4@22, everyone else in R14.
• P1's turn.  P1 does nothing.  Dr moves to R22.
• p4 attacks in R22. Dr moves to R23.
• P1 moves p4 to R24. Dr moves to R24, activating p4.
• p4 attacks in R24.  Dr moves to R1.
• P1 moves p4 to R2 (using up a move card).  Dr moves to R2, activating p4.
• p4 attacks in R2.  Dr moves to R3.
• P1 moves p4 to R4.  Dr moves to R4, activating p4.
• p4 attacks in R4.  Dr moves to R5.
• P1 moves p4 to R6.  Dr moves to R6, activating p4.
• This pattern continues, with p4 attacking in {R6, R8, R10, R12}, for a total of 8 attacks.
  

This is an extreme example, but it is not hard to set up at least two stranger attacks in a row.  The major ingredient is to get your allied stranger to attack in a next-next-cheap room.  Once your allied stranger attacks, Dr will move to the next room, and then you move the stranger to the next-next room with your free move.  If your stranger can attack in R2, maybe they can attack in R4 as well.  Even if the stranger is not able to attack that second time, it might be valuable in bringing Dr closer to you or setting up something to your advantage.

Notice how next-next-cheap rooms are at the core of yet another technique.

Even if you're playing with the official rules where strangers can win instead of a human, this attack is still helpful for weakening your opponent or turning a loss into a tie.

Strategic Implications Of Multi-Attacks

A weapon card helps with one attack; a move card can decide who gets to attack or even multi-attack, so I value move cards more than weapon cards.  (But, even I admit that using a weapon's special attack is often a game-winning blow.)

Sometimes you should decline an attack opportunity because it opens up a multi-attack opportunity for your opponent.  In many-player games, you should take almost all attack opportunities you have (unless the attack is too costly in terms of required move cards).

The bigger the board, the easier multi-attacks are.  So, watch out for multi-attacks more on the 19-room main board (with a wing closed) than the 13-room alternate upstairs board.

Taking Over A N+3 Room

If P1 and Dr are in R4, and P3@5, then P1 moving to R5 will result in P3 being activated, not P1.  Likewise, if P1 and Dr are in R3 and P3@5, then P1 moving to R5 will still result in P3 being activated in R5.  But, if P1 and Dr are in R2 and P3@5, then things are different...

• P1 moves from R2 to R5.  Dr moves from R2 to R3.
• p2 moves down (doesn't matter where p2 as long as they didn't start in R3 or R4).  Dr moves to R4.
• P3's turn, starting in R5.  If P3 stays in R5, then P1 will get activated.  If this is on the alternate downstairs board, then P3 doesn't want to move to R6, so P3 might be forced to "retreat" to R10 or R11.

So, if your opponent is in R5 and you really want to take a turn in R5, then when you and Dr are in R2, that is when you move to R5.  The critical relationship is R5 - 3 = R2.

To put it down in the general sense:

• IF you and Dr are in the same room on your turn...
• AND your human opponent is 3 rooms after you (ex: they're in R5, you're in R2)...
• AND there are no strangers in the rooms between you and your opponent...
• THEN you can move to your opponent's room, and when it is their turn, they won't be activated if they stay in the room.

Using Strangers To Skip A Human's Turn ("Stranger Override")

Imagine this scenario on the alternate downstairs board:

• P1's turn.
• P1 has no move cards.
  
• Dr@10, P1@10, p2@13, P3@11, p4@12.
• How can P1 prevent P3 from making some great plays once Dr moves to R11?  In fact, how can P1 prevent P3 from doing anything?

P1's turn.  How can P1 prevent P3 from making strong plays?

P1 has no move cards, so P1 can not move to any of the {R12,R13,R14} rooms that would ride-block P3.  Of course, P1 moving to P3's room R11 won't help.  But, moving p2 to R11 will help a lot.  Let's see how that plays out...

• P1 moves p2 to R11.  Dr moves to R11, activating p2 instead of P3.
  
• p2 moves to R10. Dr moves to R12, activating p4.
• p4 moves to R11.  Dr moves to R13.
• P1's turn!
• P1 could move to R15 to do one iteration of loose-riding, but that doesn't matter for this example.

Wow, P3 didn't get to take a turn at all.   There were two critical ingredients for P1's strong "stranger override" play:

• Moving the opposing stranger to the same room as the human opponent.  p2 is just after P1 in the normal turn order, which means that p2 will always win activation contests when the Dr moves at the end of P1's turn.  So, p2 is the only piece that P1 could move to R11 to get activated instead of P3.
• Someone friendly needs to be in the room after the human opponent.  After p2 gets activated, p2 will move down a room and Dr will move up a room.  p4 being in that next room (R12) prevented turn order naturally progressing to P3.

The second ingredient might seem hard to set up, but the friendly piece after your human opponent can be can be you instead of your allied stranger.  Like so...

P1's turn. Goal is to prevent P3 from taking a turn.

• Dr@11, P1@13, p2@3, P3@12, p4@4.
• P1 moves p2 to R12. Dr moves to R12, activating p2.
• p2 moves to R11.  Dr moves to R13, activating P1.
• P1's turn.  Things look great for a multi-attack.

Sometimes it's worthwhile to do a stranger override to just delay your human opponent's turn, rather than totally skipping it...

P1's turn.

• P1's turn.  Dr@6, P1@18, p2@24, P3@7, p4@19.
• P1 moves p2 to R7.  Dr moves to R7, activating p2.
• p2 moves to R6.  Dr moves to R13.
• P3's turn.  P3 would need to use 3 move steps in total to move to R13 or R16, which they might not have.

Reminder: if you're playing without the "nosy stranger" variant, then P1 moving p2 to R6, R22, or R23 would not have prevented P3 from attacking in R7.  P3 would just use their free move step to move p2 back out of sight before attacking.

Using Humans To Skip A Human's Turn ("Human Override")

Very similar to a "stranger override", you can have your piece in the same room as your human opponent and steal the activation from Dr.  The setup is a little different.

P1's turn.  Moving p4 to R11 will prevent P3 from taking a turn.

Example 1:

• P1's turn.  Dr@10, P1@12, p2@8, P3@12, p4@3
  
• P1 moves p4 to R11.  Dr moves to R11, activating p4.
• p4 moves to R10.  Dr moves to R12, activating P1.
• P1 is free to do a lot of Dr-riding before the Dr gets back around to P3.

The crucial ingredient was that P1 moved his allied stranger p4 so that p4 would take a turn and then Dr would next move to the room with P1 and P3.  P1 will always win activation contests right after p4's turn.

To cheaply set up a human override, you don't always have to already be in the same room as your human opponent.  If you allied stranger is in the room before your human opponent, you can move yourself to the same room as your human opponent.

P1's turn.  P1 moving to R12 will prevent P3 from taking a turn.

Example 2:

• P1's turn.  Dr@10, P1@3, p2@8, P3@12, p4@11.
• P1 moves to R12.  Dr moves to R11, activating p4.
• p4 moves to R10.  Dr moves to R12, activating P1.

Stranger Loop: Unlimited Stranger Attacks

With stranger overrides, you can execute an unlimited loop of your stranger attacking.  The basic ingredients are:

• For every room on the board, you must be able to go to the next room in one step. (ex: main board with any wings closed prevents stranger loops; alternate downstairs board allows stranger loops).
  
• Your human opponent is in a room that you can leap-frog over in one step (ex: your opponent is in R4 and you can move from R3 to R5 in one step).
• The opposing stranger is in the same room as your human opponent (sometimes you do clever things or use a move card to set this up).
• Your allied stranger is in a room that your opponents don't have sight on.
• When your opposing stranger moves down a room, that room won't have sight on your stranger.
  
• Your allied stranger is in a room that you can go to their next-next-room in one step and not see your allied stranger (ex: your allied stranger is in R12, you can go from R11 to R14 in one step, and R14 can't see R12).

If you're playing with official rules where you need to kill Dr yourself, then you do enough stranger attacks to exhaust your opponent of cards and finish up with your own attack.  Or, if you're behind, you can turn a probable loss into a tie or win.  Also, some stranger loops allow you to loot (draw a card) once or twice each iteration of the loop, so stranger loops can help you gain cards while your opponent loses cards.

• Some loop examples
  
• AltDown board.  P1's turn.  Dr@11, P1@11, p2@12, P3@12, p4@3.
• P1 moves to from R11 to R13.  Dr moves to R12, activating p2.
• P1 leapfrogs over p2 and P3 in R12 to do stranger override.
  
• p2 moves to R11.  Dr moves to R13, activating P1.
• Stranger override just happened.  P3 did not get a turn.
  
• P1 Dr-rides from R13 to R2.
• When P1 moves from R15 to R1, P1 loots R1.
• P1 moves from R2 to R5 and loots.  Dr moves to R3, activating p4.
• p4 attacks.  Dr moves to R4.
• Turn naturally progresses to P1.
  
• P1 moves p2 to R12.  Dr moves to R5, activating P1.
• p2 was moved to P3's room (R12) to set up next stranger override.
• P1 Dr-rides to R11.
• P1 repeats above steps.
• Note: Dr and P1 don't have to start at R11.  If it's P1's turn, Dr and P1 can be in any room.
  
• AltDown board.  P1's turn. Dr@2, P1@2, p2@15, P3@1, p4@12
• P1 Dr-rides to R11.
• P1 moves from R11 to R14.  Dr moves to R12, activating p4.
• p4 attacks Dr.  Dr moves to R13.
• P1 moves p2 from R15 to R1.  Dr moves to R14, activating P1.
• P1 moves to R15.  Dr moves to R15, activating P1.
• P1 moves to R2.  Dr moves to R1, activating p2.
• p2 moves to R15.  Dr moves to R2, activating P1.
• P1 repeats above steps.

Here are some two-player boards and how easy it is set up a stranger loop:

• pretty easy: AltUp
  
• somewhat easy: AltDown
  
• medium: LairNorth
• hard: LairSouth, HauntedFull, HauntedUp
• impossible: HauntedDown, Main (with at least one wing closed)