Skip to Content

Citadels - chat transcript 26 February

No replies
Scurra's picture
Joined: 09/11/2008

The Rules in two minutes

Scurra> Anyway, we're going to talk about one of my favourite games today
Scurra> And I'm going to call it Citadels, since that was the original name for it
zaiga> It's shorter too
Scurra> And easier to spell
zaiga> The Dutch edition is called "Machiavalli" btw
Scurra> Can we have a show of hands first - who's played it and who hasn't?
Torrent> i own it, I bought wooden coins, I even brought it to EUrope with me.. but I still havent play it
zaiga> I have played it
DavemanUK> heard it, analysed, but not 'played' it
SVan> i have not even a clue about it
zaiga> Well, according to Scurra it has only two rules, so I'm sure he can explain it in 5 mins
Scurra> A couple of weeks back I claimed you could teach someone Medici in five minutes. Well I reckon you can teach someone Citadels in about two.
SVan> so far i like that
Torrent> my big worry is the dreaded, "OK but HOW do you play" sort of question
Scurra> Exactly. And *that's* what makes Citadels great
Torrent> not just what the rules are, but basic strategies
Torrent> and a difficult thing for me to bring to my 'gaming group'
[Scurra> In brief, the game consists of a series of rounds. Each round a player will select a role (in a first phase) and then has a turn (in a sequential phase)
Scurra> The game is about building "districts" and it ends when someone (or more than one) has built eight of them
Scurra> And that's it
SVan> *sounds easy enough* although i'm sure it isn't

It’s all about the role selection

zaiga> It's all about the role selection, though
Scurra> What's key here is that the role selection is semi-hidden. Unlike Puerto Rico, say, where you essentially announce your choice at the start of your turn, in Citadels you make your choice in a sequence.
Scurra> To simplify it (by excluding expansion options!) there are eight roles. Each of the roles has a special ability - and it's those that are the key to the game.
Scurra> With varying numbers of players, different numbers of the roles will be available (and I must say that I think this is a much better game with five or six than with just four)
Scurra> The key to the game is thus twofold: calculating what other people have taken, and disguising what you have taken
sedjtroll> well, unless you're talking about the expansion guys, there're always the same number available
Scurra> Yes, but some are "removed" early with fewer players
sedjtroll> The important point is that they are removed face up
sedjtroll> Scurra: I have to disagree- I think Citadels is much better with 4 people because the figuring out who has what is better
Scurra> OK Seth, we'll maybe come on to that
SVan> so in the way that you choose roles is somewhat similar to Puerto Rico; basically the choice you make has a impact on all the players?
Scurra> Only in a general sense, Svan
zaiga> They are removed face down, aren't they?
Scurra> Zaiga, one is removed face down
sedjtroll> i.e. with 4 players, you start with 1 face down and 2 face up, then the players choose the roles in turn (secretly choosing)
Torrent> Verraeter is the other game that it is similar to.. even acknowledged by Bruno
Scurra> Oh, the role selection wasn't original - but I think it clearly works better here because the game is more compact
zaiga> The game IS the role selection
Scurra> Not quite, Zaiga, but pretty close.
zaiga> The choices in the other phase are rather limited and obvious imo
Scurra> Well, maybe we'll get to that!
SVan> so people don't know what you choose? for the roles i mean
Scurra> It's a drafting mechanism. Svan
zaiga> SVan: you can never be sure what role someone has chosen...
SVan> ok
sedjtroll> I like Citadels and I like how the Role Selection can be compared to PR
Scurra> The start player (usually the player who chose the King the previous round, but not always!) gets first pick of the roles
Scurra> And so they know what the face-down card is - because it's the missing one! As the cards go around the table, each subsequent player has less information. Until the last player has just two roles to choose between…
zaiga> And guessing what someone has chosen is an important aspect of the game
Scurra> (In a seven player game the last player gets to choose between the last role and the face-down role btw)

Guessing and Second-guessing

sedjtroll> The reason I don't like more players is because the later players really have no way to determine who picked what ahead of them. With only 4 players there's enough info to make a realistic guess.
sedjtroll> 2 and 3 player games need to use variant rules in which each player gets 2 roles- those are interesting games !
Scurra> Seth, I completely disagree. With four it is far too easy to completely shut someone down
zaiga> Sedj: but the first player doesn't really know what will be picked after him, so it works two ways
sedjtroll> how do you mean shut someone down? The game is about figuring out what people picked, so if you can't reasonably do that then it's a silly game. with 6 players you can't do that. With 5 it's a stretch but doable, with 4 it's almost easy.
sedjtroll> 7 players is just silly - though we played 7 players with all 9 guys, including the expansion people - it was very silly, but quite fun
zaiga> Citadels almost seems a different game with different number of players
Scurra> Indeed I agree with that.
sedjtroll> it turns out if you remove the Thief from the game, people can stock up TONS of money
Scurra> With four you *are* able to work out what people took, and play accordingly. But with five or six you start having to work on probabilities and instinct
sedjtroll> Zaiga: the first player doesn't know what's picked after him... true. That is equally bad with 6 players.
Scurra> And I find that is a much better reflection of the game
zaiga> It depends on what you like most then
Scurra> As someone said, Faidutti designs games that are chaotic but not random
sedjtroll> (as in as bad as the later players not being able to discern what the early players picked)
Scurra> I find four player games not chaotic enough (but seven players is, indeed, random!)
zaiga> Hmm, I didn't care much for the Faidutti designs I played
zaiga> You do a lot of stuff and then someone wins
Scurra> I think you generally like or don't like that sort of thing
sedjtroll> Scurra: I guess it's preference then. I'd rather be able to figure out the roles, not have to flat out guess.
zaiga> Where winning never really seems to correlate to what you did during the game
Scurra> Seth, and my argument is that you don't need to guess! There's usually enough evidence for you to make a reasonable assessment
zaiga> Scurra: what kind of evidence, for example?
sedjtroll> Scurra: well you can say "hmm... the Architect would really benefit player 4 over there, but there there's a facedown card and 3 players before him and it doesn't LOOK like they'd want Architect, but if thy did choose it then my Thief will suck."
sedjtroll> And evidence such as "that player likes to choose Assassin" only helps some of the time.
Scurra> I don't claim that it applies every turn. But I do think that sometimes it can be important
sedjtroll> I don't claim that it applies ANY turn ;P
zaiga> I never choose an obvious role if I have some money...
zaiga> To avoid getting robbed by the Thief
Scurra> That's an example of evidence right there
sedjtroll> heh
Scurra> The thing about Citadels is that you find yourself spiralling into a horrible second-guessing game - which I find exhilarating while others hate it. And that is, I think, the major weakness of the game
Torrent> i guess that is the 'Bluff' side of things
Torrent> it ends up being a mechanic of its own that people have opinions on
Scurra> If you play with a "mixed-ability" group (i.e. some who have played and some who haven't) then the game does run the risk of collapsing in a heap
Scurra> (I've played in a game in which one player basically never had a real turn, for instance)
sedjtroll> Certainly some people are better at figuring out what other people might have taken. They should do well at this game
Scurra> (what's odd is that I am dreadful at most other games of this type, such as Basari, but really good at this one)
sedjtroll> In the game where a player never has a real turn, that player is at least in part to blame
Scurra> Yes, but they hadn't played before, which was the problem. There were two newbies and three experienced players in that game and it fell apart badly
Scurra> If you play Citadels with a group of five people who all know how to play, I reckon it's about as good as gaming gets.
sedjtroll> A lot of games have the problem where if it's your first time you might simply get owned... the first playthrough should be something of a learning experience, and if the other players are nice they'll walk you through it and give you tips. 5 players is good. I think 6 is too many
Scurra> In most games I'd agree. The problem with Citadels is that you can't do that "hand-holding" thing!
sedjtroll> maybe if you use 9 roles (one extra face up at the beginning)
zaiga> At least in other games you can still *do* something, even wehn you can't win

Side-foray into specifics

sedjtroll> So which do you think is worse- the Assassin or the Witch?
Scurra> The Witch is much better than the Assassin. But you have to know how to play her, whereas you can point and shoot with the Assassin (IMO)
sedjtroll> It seems like the Witch is a 'nerfed' assassin, Better for the Assassin if you 'miss', and better for the Witch if you 'hit'. In either case the Witch is easier on the target.
Scurra> However, I don't think we should get into detailed "strategy" discussions here
Scurra> What I was interested in was the overall impact of the role-choosing mechanic
sedjtroll> I was more talking about the choices made in making the expansion guys but you're right, that's not really Role Selection


sedjtroll> I like the Role Choosing thing a lot
Scurra> I find Citadels to be an interesting example of a one mechanic game where the mechanic is so strong the game would almost work regardless of the other bits
sedjtroll> I like it in Citadels, I like it in PR... I even suggested it for Everest in the GDW!
zaiga> What I dislike about the role selection is the downtime
Scurra> And there's the other weakness. The AP is horrendous.
zaiga> It's only really interesting for the player who's turn it is
sedjtroll> is it really a 1 mechanic game? Or does the 'power card' effect count seperate? I guess the Role Selection is nothing without the power of the cards.
zaiga> Well, not necessarily AP, just normal downtime
sedjtroll> ... another reason a smaller game is better
Scurra> Isn't the role-selection a prime example of AP?
zaiga> I mean, you really want to take some time chosing a role, doing the second-guessing thing, but the other players just sit and wait
DavemanUK> unlike Basari where the actions are simultaneously chosen (which I like)
zaiga> Hmm, not really I think.... the problem is more that others have very little to think about, when it's not their turn
sedjtroll> I'm not familiar with Basari
Scurra> Hmmm. We'll have to talk about it in a few weeks time then
zaiga> At least in, for example, E&T you can try to plan ahead during someone else's turn
Scurra> OIC - yes, you can do very little in the downtime
SVan> sometimes it's nice to have downtime in a game to plan
Scurra> And can only really analyse the situation once you get the cards.
zaiga> In Citadels, there's very little to think about, because it is such a chaotic and tactical game
sedjtroll> Could you combine Role Selection with some sort of board position? Even PR doesn't really do that
Scurra> Svan, the argument here is that you can't plan
SVan> i see
zaiga> You might get killed or robbed, or get a whole set of new cards
SVan> you really can't do anything without knowing what your card is then? instead of do i mean plan
Scurra> No, although you can make some good guesses as to what other people are thinking
sedjtroll> well, it could- you could determine what you want to do, and when the roles come up you can see which will help you do it
zaiga> At least in PR you can look ahead and otherwise you can still try to talk an opponent into selecting another role, because the selection is open
Torrent> you can kind of plan what you would want, but without the cards you don't know your exact choices
SVan> so does open or closed seem better?
Scurra> Svan, you could never play Citadels "open"!
zaiga> Well, open would not work in this game of bluff
SVan> i mean in general
zaiga> I don't think you can say one is better than the other
sedjtroll> depends
sedjtroll> I'd like to play a game of PR where the role selection were secret. That would be cool
SVan> it certainly be an interesting game
Scurra> That's why we need to talk about Basari sometime. It has an elegant approach to that
Scurra> But Citadels works because the roles have continually changing importance in a way that, say, the roles in PR don't (at least not quite)
DavemanUK> sedj, you'd like my 'version' of PR with hidden simul role choice
Scurra> Ho yus. I'm looking forward to the test run of that, Dave
DavemanUK> indeed, there's next to no downtime from AP now
DavemanUK> incidentally, I would spend the 'downtime' in Citadels studying my opponent's faces, micro-expressiona and body language to determine the role they had chosen

Building Districts

Scurra> But I do think that the Districts do actually have some significance in the game - more than just a filler aspect

Torrent> remind me what is the winning condition.. the most spent on buildings?
Scurra> Yes (with special case bonuses)
zaiga> I would spend the downtime getting more beer
Scurra> sounds like a plan
DavemanUK> true, zaiga...Citadels is a lot harder than Basari (where some of our players give themselves away by looking at the resource they are choosing an action for)
Torrent> in this way it is almost poker-like.. the bluff aspect anyway
DavemanUK> yes indeed Citadels does capture everything I like about Poker
Scurra> Anyway, the thesis tonight was that for a game with (as I said) basically two rules (well, two-and-a-half really), Citadels is surprisingly deep for such a small game
Scurra> The trick is that the interactions between the roles makes the game spin in some very interesting directions
DavemanUK> yup, it's why the game works so well for the abstract-minded players, they can just 'see the numbers' without the gloss getting in the way
zaiga> I don't agree that Citadels is deep, I fnd it very shallow
Scurra> I was just about to ask you why you though the district building part was obvious
Darkehorse> I dislike citadels..
Scurra> I know *you* don't like it, Darke
zaiga> Because it is almost always best to simply spend as much money as possible on building
Torrent> if iremember correctly.. the only point of money is TO build. It cant be used for anyuthing else
DavemanUK> why, is the risk of holding money too great, or the oppurtunity loss?
zaiga> Handmanagement is impossible, because of the Wizard (or whatever the handswitching guy is called)
zaiga> Dave: both
zaiga> Two buildings of a cost of three is better than one that costs six
Scurra> But a building of three will get knocked down, where a building of six won't
Scurra> I agree that you want to spend your money - but it's not always obvious how to do it efficiently
zaiga> Just a minute
zaiga> First of all, it is not efficient for the soldier to tear down a building... it costs money and it doesn't gain you anything And you run the risk of losing money, through the Thief
zaiga> So, better spend it now
DavemanUK> he who dares wins though
Scurra> I'm not sure I agree with you about the Soldier/Warlord not being efficient
DavemanUK> are most games decided on who hoarded money and didn't get 'thiefed'?
Scurra> and the whole deal with the thief is for that reason
zaiga> Why, what do you gain when you build a six cost building as opposed to two of 3?
Scurra> Usually a special effect!
DavemanUK> mechanic-wise, they both provide tension otherwise it'd be a dull game
Scurra> But I have often knocked down three-cost buildings at the right time...
zaiga> But two buildings give you an edge in the race towards building 8 buildings
zaiga> Why would you tear down a 3 cost building? It costs 2 gold and doesn't gain you anyhting?
Scurra> I'm not disputing that. I'm just saying that you have to be in a position to do it, and be ready to choose between defending them or doing something else
DavemanUK> possibly, in the right group, it is merely the 'threat' of the thief rather than the 'execution' of it that drives a game aka 'The Boss' in Tikal
Scurra> Yes, it can be a swing of three points (and sometimes the three-point colour bonus too.)
Scurra> The assassin is more dangerous than the thief for those purposes though

Fixing the Game

zaiga> Perhaps the choices during the buidling phase would have been more interesting if triangular numbers were used for scoring., ie. 1 gold =1 point, 2g = 3pts, 3gold = 6pts, etc
Scurra> You'd have to up the "finishing" bonuses too I think. But if all the buildings were reduced (so that nothing cost more than four, say) then the Warlord would become very very strong
zaiga> Perhaps, or just let the game end after X rounds
Scurra> I'm not sure I would like a fixed round version of Citadels. It would skew some of the roles
zaiga> Scurra, nothing that can't be fixed
Darkehorse> Isnt it true though that when you arent able to get the role you want, sometimes you have to choose the least of the available evils?
Scurra> Darke, no, sometimes you have to take the most because people are expecting you to take the least
Scurra> How would you rebalance the Architect so that he was still worthwhile? without being unusable (since he'd be shot or robbed *all* the time instead of just *most* of the time
zaiga> Is that the guy that can build 3 things in a turn?
Scurra> yes, he's the guy who gets two cards *and* can build three things (kind of like getting a free go, really)
zaiga> Why would he be unbalanced?
Scurra> I think he'd be unusable since the eight buildings thing would have gone away - and one of the tricks of the architect is trying to sneak him in when you think you can get away with it
zaiga> Perhaps, if he could build things for 1 gold less or such
Scurra> oh, now that would be an interesting change.
zaiga> Or simply swap him with one of the expansion guys
Scurra> Yeah, the Diplomat would probably work better in that context
zaiga> Of course, the whole game was build around the 8-building ending...
Scurra> playing with that does risk undermining the whole structure
zaiga> Yeah, I wouldn't change Citadels... but it's sometimes interesting to think about it... so you can use those ideas in your own games
Scurra> Which is why we're having these chats...
zaiga> Exactly
Scurra> (Hey, that seems like an ideal place to stop, you know..
SVan> The End
zaiga> Next week: Taj Mahal
Scurra> (cue fifteen minutes of credits)
Darkehorse> Produced by.....
Darkehorse> Edited by....
zaiga> First hairdresser....
Scurra> Animal wrangler...
zaiga> Second hairdresser....
zaiga> Assistant to the second hairdresser
Darkehorse> Note: no designers were harmed or even ridiculed during the making of this chat.
SVan> All rights reserved

Syndicate content

forum | by Dr. Radut