How much of Settlers of Catan is Luck? | Hacker News Reader

How much of Settlers of Catan is Luck?

malcolmong.com

27 pointsmalcolmong44 comments

If you like Catan, try these games:

http://boardgamegeek.com/boardgame/35677/le-havre <= economic engine building game set in a costal french port

http://boardgamegeek.com/boardgame/43111/chaos-in-the-old-wo... <= Point scoring game where doing a faction specific objective gets yours further ahead. Very balanced, but via active player action (like refusing to trade with a guy at 8-9 points in Catan)

http://boardgamegeek.com/boardgame/68448/7-wonders <= civ building game with small military elements

http://boardgamegeek.com/boardgame/34635/stone-age <= stone age building/civ game

Yspahan http://boardgamegeek.com/boardgame/22345/yspahan <= point scoring game about running shops in an "idealized" arabian bazar.

Luck is a factor, but interpersonal dynamics and alliances are most important, in my experience.

How to win: always be a moderate second while limiting the routes for first to win, concurrently reserve at least two separate routes to victory for yourself, and don't let the game get too lopsided. Then as the end draws near, finish it in a turn.

It would be interesting to have a Catan-AI competition. Does the Catan server allow bots?

What bugs me about Settlers isn't that it's got too much luck, it's that I can pretty much tell who is going to win (or at least one of two) based on the initial settlement placement. Or at least who's going to lose. And it's pretty much completely no fun to play a game where the result is preordained that you'll lose.

Speaking of settlement placement: There aren't enough rolls in a game of Catan for the distribution of rolls to come out as you'd expect them. I've played a game where more 12s were rolled than 6s (though OMG were there a lot of 8s). It's not bad dice -- it's just the normal result of a "random" distribution when you're not generating that many results.

So a HUGE part of the strategy that the article misses is to maximize your DISTRIBUTION of numbers. I don't care if there's an 8/9/10 combo available for your second settlement; if you already have 8 & 9, you're much better off with a 3/4/5 combo (assuming you have none of those). Otherwise you're stuck with a feast or famine situation; sure you might be lucky, but you might instead be resource-starved for large parts of the game.

I actually picked up a "Deck of Dice" [1] to use with Catan, which is 36 cards that have a regular distribution of results of rolling two dice, but haven't given it a try yet to see if it makes Catan feel different. It seems like it might play a lot different; I really should give that a try... :)

[1] http://www.boardgamegeek.com/boardgame/5460/the-deck-of-dice

I disagree. Going for the low-variance option is a good idea if you do not want to lose, but not one if you want to win.

Consider a game where one player gets two items for every roll above 7, another two items for every roll below 7, and a third one item for every roll that is not 7. With any not perfectly even distribution of rolls, one of the first two players will have the highest production, and the other of the two the lowest. There is no set of throws that will give the last player the most produce.

Except that rolling high vs. low vs. a balance of both all have equal probabilities.

A good distribution of numbers is a dominent strategy in catan due to the robber taking half your goods if you have more than 7 cards when a robber is rolled.

Late game overclustering (aka, being on too few numbers) is very brittle to rolls of 7 and additionally to the robber stopping production there as well.

I find that it's not so much a balanced distribution of numbers, but that you don't have to cash in 4 for 1 for a resource that you need.

Also, you don't need a deck of dice - just draw on an old pack of cards with a permanent marker :)

Ports are good, sure. I'm not saying it's the ONLY strategy. Just that it's a major bit of strategy that happens before the game even starts. You can certainly choose a spot near a port at the start, though I wouldn't recommend putting an initial settlement on a port. You need the most resource production you can manage early on.

Already got the deck of dice. I don't actually have any extra old packs of cards -- I have a couple, but I'm not a big card game player. Well, I used to play Magic, but writing on Magic cards just seems wrong. ;)

I've played a lot of games of Catan against the computer on the Xbox 360, which gives stats, and I found that is almost all cases the player who had the most resources (produced - lost/stolen) won. To some extent this makes sense since a better economy will produce more resources, but it didn't seem to matter what the resources were or whether they were used directly or 4-crunched into something else.

What I took from that is if you're choosing settlement spots or robber targets you'll do pretty well just maximizing the number of resources you produce or block (which, for the robber, means multiplying the number of pips by the number of resource gained if the number rolls)

That the player with the most resources wins seems obvious to me, because you have to buy all wining points yourself.(Cities, Towns, tech cards, streets, etc) There is no other way.

But you have a bunch of different strategies to get the most resources.

Trading is a pretty big aspect in a real Settlers game and the value of your resources depends on the overall resource output of the game field. E.g. When your the the only one controlling a iron field, it becomes more valuable for the others.

It's true that the person with the most resources will probably win, but a key factor getting to that point is ability to best ramp up production to produce those resources.

Early in the game it is important to try to get that next settlement or city. If you can place a robber to block important sheep or whatever is in short supply then that can be better than just blocking a high number of pips. When placing a settlement sometimes it is better to spread out rather than circling around one hex with high pips (and being vulnerable to a robber knocking out your entire production chain).

Another aspect of placing the robber on an opponent's "high pip" space is that they become motivated to play soldiers and possibly move the robber to your own high pip space. If you instead place the robber on a ho-hum tile you might not block as many potential resources, but the robber will stay away from your tiles for a longer period.

Meta-question: board games like chess and go are entirely based upon skill, but are there board games similar to Monopoly or Catan where you have the board, pieces, cards that grant abilities, etc. which are based entirely upon skill?

You'd have to eliminate dice entirely and make movement solely the choice of the player, within a defined set of rules. Cards wouldn't come up randomly but would be earned based on objective criteria. Conflicts for position/status/etc. need to be settled objectively as well.

I can think of a number of computer games which might fit the criteria, but I'm unaware of what's out there in the physical board game world.

This isn't an answer to your question, but I think it would be pretty easy to come up with a few rule changes to make Catan deterministic.

You could get rid of the dice rolls by saying that a tile with a certain number on it will produce a resource every X turns.

The robber could either be completely eliminated, or you could just have players take turns moving it at some interval.

The closest one I've run into is Agricola: it's almost pure skill ('almost' because there is an optional set of 'occupation' cards that get randomly distributed at the beginning of the game). The theme is building out the best farm by the end of a set number of turns.

The main thrust of the game is for players to take turns choosing actions from a limited common pool that is reset at the beginning of each turn. One of the actions is 'go first next turn'. Everyone ends up competing for actions. The game seems complicated at first but is actually fairly simple. Great fun!

If you like Agricola, try La Havre. Same guy, better game, similar amounts of luck. http://boardgamegeek.com/boardgame/35677/le-havre

You've hit my hobby (but I dislike the games you're talking about).

I'd contend chess's outcome is based purely on the decisions of the players, but there are often close calls in that game as well, which it is reasonable your opponent will zig or zag, etc, where your chances of winning come up entirely on if you happened to believe if he was zigging or zagging and did moves to counter zigging or zagging appropriately several turns ago.

Multi-player games still have the chaos of multiple people making decisions. So yes, there is no technical randomness, there is still chaos and uncertainty:

Lifeboats: http://boardgamegeek.com/boardgame/249/lifeboats <= People place various pawns representing their factions, in boats, then vote which boat to move, then vote which boat springs a leak. If your boat doesn't have enough room for everyone due to the number of leaks, you pick someone to drown. If your boat has more leaks than people, everyone on your boat drowns. Points are tallied by how many of your people survive the journey and to which island they get to after the above is repeated several times. There is a trump card, but everyone starts with 3 of them, they are not random.

http://boardgamegeek.com/boardgame/483/diplomacy <= period of time of negotiation, everyone writes down secret orders, orders are resolved simultaneously using a very strict ordering.

Stratego: http://boardgamegeek.com/boardgame/1917/stratego <=Advance people in a chess like manner with player markers being partially hidden

Blokus: Lay down tiles and try to last as long as possible http://boardgamegeek.com/boardgame/2453/blokus

Gemblo: Same as blokus, harder to find in the US, done better: http://boardgamegeek.com/boardgame/19427/gemblo

Antike: http://boardgamegeek.com/boardgame/19600/antike only luck is who you get as your place on the board. It's okay, but it has a certain snowball quality to play some games. Somewhat Risk like

Hive: http://boardgamegeek.com/boardgame/2655/hive chess, without a board, with bugs. Fun, short, but non-random

1830: http://boardgamegeek.com/boardgame/421/1830-railways-robber-... a game about being a robber baron. Interesting stock market mechanic, etc.

Of these, lifeboats is a little fun, but honestly, most interesting game mechanics are there to make a game playable more than a few times (which is one thing the above games lack), and most of them have some degree of randomness. The next crop of games has a minimal random element:

Power Grid (Random element: which power plants will be up for auction): You all play Monte Burns esque robber barons over a electrification empire. Very "brain burner" to many people's minds: http://boardgamegeek.com/boardgame/2651/power-grid

Puerto Rico (Random Element: which plantations are available this round): You play plantation owners on Puerto Rico during the days of the sugar trade. The theme is distasteful to some http://boardgamegeek.com/boardgame/3076/puerto-rico

Chicago Express (Random Element: Seating order): Robber baron type scenario, but playtime is <1 hour after the first game or so: http://boardgamegeek.com/boardgame/31730/chicago-express

When I said "like Monopoly or Catan" I wasn't looking for games similar to them in style or gameplay -- I meant like them in the sense that they have features such as cards and other additional game elements beyond just the board and playing pieces.

Come to think of it, Battleship is purely skill based though it's quite simplistic in that, like chess, it only has a board and pieces.

Thanks for the list. I'll check them out.

I wasn't intentionally presenting ones similar in style or play, just in amounts of uncertainty.

Settlers of Catan has piles of luck compared to many other games that only get uncertainty from the chaos of a bunch of other players moves.

The die rolls, especially using the wooden dice that come with the game, are very likely to come out very biased and not very even for a given game, as well as the very rare powerful cards buried in a sea of knights in the dev card deck.

Luck is not bad. Luck is merely a thing you can put in a game or not, just like 'it is played on a field' or 'do you wager real money on the outcome'.

Games with too little luck make unskilled players unwilling to play them with skilled players (such as chess).

Does Catan have more luck than many other designer/german board games? More so than many ones originally published in Europe and a few of the US ones (Princes of Florence, Power Grid, Automobile, Chaos in the Old World all have far less luck), but many other games, especially "Classic" board games, have far more luck, such as Diplomacy (chaos based uncertainty is still something uncertain), Monopoly, Scrabble, and Risk.

If you gasp at catan having lots of luck, think of this tidbit: When's the last time you have seen people who know how to play texas holdem play straight up 5-card draw with no wilds? Other than a brief interlude, probably not often: The game is too skill based. Whereas the much closer odds for texas holdem have MANY people willing to play a round, even drop 10k play against the pros in the WSOP.

If you're looking for more games in the vein of Catan, check out this list (some good wargames in there as well, read the description before buying):

http://www.boardgamegeek.com/browse/boardgame

Could you explain how 5 card draw is more skill based than Hold 'em? This is not in a combative tone, I really want to know.

Yeah, I think the Hold'em players skip 5 card draw because it is less of a skill based game.

Starting hand strengths are MUCH further apart in holdem and 5CD.

You can start with the nuts (Straight flush to the Ace) in 5CD, but in holdem, AA is only about 450:1 better than a random disconnected offsuit (2s 8h).

Tight play is overwhelmingly rewarded (you can start with a hand thousands of times better than your opponent). So in a 5CD game against bad players, the good players fold repeatedly until they get a strong hand, and take all the money, repeatedly.

Against people of similar skill, draw (with the antes most people play with) is more of an actual game, and less of a horrible train-wreck of watching bad players getting fleeced. But in a similar holdem game, a guy can sit down, play every hand, and will often turn out okay.

*My comments are about cash games. For tournaments, the tournament dynamics dwarf a huge quantity of the strategy in the actual game.

I guess I don't see how that's more skill based. In fact it seems more luck based to me – your play is dictated more by literal luck of the draw than by your skill level. You could be the best player in the world, and if you are continuously dealt trash, lose. In Hold 'em, the skills of bluffing and reading are much more important than the luck of your hole cards.

There's good game design around luck (randomness) and bad game design around luck.

Monopoly has a lot of randomness, and it is hard to design a strategy to mitigate bad luck, which is probably why it isn't worth playing more than a few times.

Settlers of Catan and Texas Hold'em Poker both involve randomness, but a good strategy will mitigate bad luck (for example , in settlers, avoid holding a lot of a commodity to not be crippled by the monopoly card) and exploit good luck.

Monopoly's biggest issues are that it eliminates people, and people refuse to play with the auctions and other game rules as written (and do stuff like letting people survive, etc, prolonging it longer).

To each his own, for sure, but that game takes forever without the auctions.

Any proof on those dice being biased? I am truly interested in this now, surely some geeks tested various game company's dice.

A wooden die that has carved pips with no active attempt to prevent bias (aka, a weird weight in the middle, etc), will exhibit a bias from the fact some faces have more pips (which are pits in the wood) than others. This means some sides are more likely to be pulled face down, and end up face down.

That is why in vegas you use those certain red dice at a craps table http://en.wikipedia.org/wiki/Dice#Precision_dice

I play catan with craps dice.

If you play Cities and Knights (Catan expansion), the contribution of luck to a victory decreases by quite a bit, methinks. The game is much longer (play to 13 or more points) and involves a heck of a lot more strategy. There are more ways to win. You can also play with cards instead of dice (or like me, build a quick web-based cards prototype because I was too lazy to go physically buy them)

Here's a quote from Klaus Teuber, the designer of Catan, that speaks to his philosophy of luck in games:

"When I play a game, I play intuitively. I don't like games where you have to think too much, where luck plays no part and the better player always wins. Chance makes things more exciting. If I know exactly what is going to happen tomorrow, things get boring. It's the same in a game. I prefer it that way."

What makes Settlers fun is that there is a good deal of randomness (luck), but you can play the game in such a way that you more often than not luck is on your side.

For instance, having access to all five resources leaves you in a better position for trading and buying, having diversity of numbers guarantees more often than not you'll collect a card, and having access to ports means you can have favorable trading.

There's also a good deal of diplomacy involved, because if someone decides to shut down your resources with the robber or refuses to trade with you because of some ribbing you can really limit your chances to win.

For those interested in playing Settlers of Catan, but don't have the game or enough players handy, it is available on BSW, a German site that also has an English interface. The quality of play is generally quite high & competitive, for this game and many of the others on the site.

http://brettspielwelt.de/Spiele/Siedler/

There are also a number of the other card/boardgames mentioned in the rest of this discussion which are available on BSW.

Short answer: a fair amount. Anecdotal story: The one time i played, having never even heard of the game before, I learned as I went along and ended up winning. I'd say my final few plays were the only ones based more upon strategy and less upon "Is this a fair play? Yes? Well then here's my play." So a good deal of luck is involved because a novice can't beat pros in a game based entirely upon skill.

apart from chess, I don't know of other games based entirely upon skill :)

Go =)

tic-tac-toe and go also require as much luck as chess.

Chess/Go/Checkers are one category of game where luck plays no part, but not the only one. The thing they all have in common are 2 players and perfect information.

Then there are games like Stratego, which have no 'luck' but require you to make decisions based on imperfect information.

And games like Diplomacy, with no random element but a social dynamic. I'd assert that this game has no 'luck', but someone else up thread defines luck a bit differently and says that it does. :)

It has uncertainty, which comes out the same as luck to most players.

You came to your conclusion based off a sample size of 1 from an anecdotal story...

How much of it is luck depends heavily on where you place the plateau of what "everybody knows".

For example, it is a skill to use the heuristic that one should not place the robber near one's own towns, or that one should try and build additional towns and roads. If that is considered something that not everybody knows, the skill factor likely is over 50%.

I find that, in Settlers of Catan, if you're winning - it's skill. It's only luck when you're losing ;)

If the luck element bothers you, just use a deck of 36 cards, numbered from 2-12: 1x2, 2x3, 3x4, etc. (Either draw your own on a deck of cards or use two decks of playing cards). Shuffle, and deal from the top instead of a dice roll. When the deck is exhausted, reshuffle.

How much of it is your friends teaming up to help somebody else win just to make you lose?

It can be argued that making sure that you're not that losing player is a skill as well. There is a game called Family Business that is almost entirely based on this concept. Each player has a gang of mobsters, and on each turn you get action cards. The game revolves around the "hit list". You can put enemy mobsters on the hit list with action cards. When there are 6 or more mobsters on the hit list, a "mob war" starts, and on each turn the topmost mobster on the hit list is killed. The mob war continues until the hit list is empty. There are also action cards for removing a mobster from the hit list, instantly killing an enemy mobster, blocking a card that puts a mobster on the hit list, etc. The key to winning this game is convincing your friends to collectively screw over another friend, and to convince them that it is a bad idea to kill your mobsters by using the right tit-for-tat strategy and using intimidation and bluffing about which action cards you have to take revenge.

Similar strategies can work for Settlers of Catan as well. When a player does something bad like block you with a road and you're playing at a disadvantage, then you can sacrifice this game for scaremongering and do everything in your power to make the player that did it lose (block him with roads and always use the robber on him). Next game people will be more careful ;)

plug: my Settlers of Catan strategy guide: http://blog.oarsum.com/post/263227690/the-weasel-guide-to-se...

Might come in handy for some...

Reply on news.ycombinator.com