Tuesday, January 7, 2014

Breadth of game

I played Puerto Rico a couple times over the holidays and came to some conclusions about the problems the game has.  It does some things brilliantly like randomization and the rotating jobs system but at a very high level of play there are some issues with the game, notably breadth.  Terminology can be tricky, so I will try to define breadth of play first.  Breadth of play is a measure of how many different ways there are to be successful at the game.  Depth, on the other hand, is a measure of how complex decisions are and how difficult it can be to determine the right thing to do and PR has plenty of depth.  At a high level of play though many of the options you might think should work simply don't because everyone is either pursuing a Builder strategy or a Shipper strategy and unfortunately both rely on doing the same things.

Anyone going for Builder is planning on building all of the production buildings, a Factory, the Guild Hall, and will take a Small Market if they can.  They will definitely plan on making one of each trade good.

Shippers on the other hand build 3-4 production buildings, a Harbour, a Wharf, and will take a Small Market if they can.  They will plan on making one of each trade good if they can, otherwise they will make 4 of the 5 goods.

There are a few big problems with these two top tier strategies.  For one, they only make use of four of the pink buildings and leave seven largely unused.  (Large Market is a reasonable buy, but isn't in anybody's game plan.)  For two, the entire strategy set of 'make lots of one good and ship it' is completely missing.  Because of the immense power of the Factory and Harbour rewarding diversity and the fact that making lots of one good is incredibly costly in terms of colonists nobody pursues a strategy centered around a single resource.  Diverse production is already amazing when Trading because it makes it easy to jam other people and makes it hard for you to be jammed.  It is also very useful in shipping because you get lots of choice in making boats and no matter what boats get made you get to ship most of your goods.  What it boils down to is that the top tier strategies focus on diversification of production and use only a few buildings that reward that diversification and this is very limited in terms of breadth.

I sat down to try to figure out how the game could be made broader and results more varied instead of just reporting "Well, there were two Builders and the one that got the Guild Hall won."  What I wanted to do was fix production so that there was a real incentive to specialize in a particular good that might offset the problems with specialization and also find a way to make little used buildings better and ubiquitous buildings worse.  I really want people to think of Hospice - University - Fortress as a valid Builder strategy and Small Warehouse - Office - OMG Tobacco as a valid Shipper strategy.  Ideally of course I want plenty of cool ways to win rather than just two, and they should take more description than a single word.

First thing I reduced the cost of Hospice, Office, and Large Warehouse by one.  Then I reduced the cost of University by two and increased the cost of Small Market by one.  The really big change was that I eliminated all of the extra circles on the Large production buildings and made them support unlimited plantations.  Being able to support 5 Tobacco production with a single dude in the production building is *huge* in terms of making that strategy viable.

I tried out my changes and things seemed really neat.  The person that won (though I was all the people) had the build order of Hospice, Hacienda, Lg. Indigo, Tobacco, University, Factory, Lg. Sugar, Harbour, Guild Hall, Fortress.  Not every day you see that build order win a game, especially when the other players are competent.  Thankfully PR has reversible buildings so I just wrote on my changes and can flip them over to play the base game.  Now to find people to play 50 games against me so I can see how things play out when I get really good at this new game!

19 comments:

  1. Interesting.

    I agree - I like PR, but find it frustrating that so many of the buildings are never taken because they just aren't that good. And it's not even as if there are obscure times when they are good.

    Amusingly, when we started playing, and no one had any experience, the Hospice was our top pick and we tried to get it early and break it. So maybe it's a trap building for new players?

    ReplyDelete
  2. When I started playing I bought the Hospice constantly, convinced it was good. I lost and lost and finally had to admit that Sm. Indigo, Sm, Suger, Sm. Market are the things you buy. The problem is that there just aren't that many settler calls and the hospice is so expensive that it totally jams your chances of getting into Coffee / Tobacco first. Everyone finds that they never have enough dudes to run everything but it turns out the buildings that make dudes aren't worth buying and it is far superior just to not have enough dudes.

    ReplyDelete
  3. Really?! You never tried the "I make only corn and ship it on my private wharf" strategy? It actually sounds to me like you haven't played PR enough, since you are discounting one of the top-tier strategies entirely. Factory might be rewarding, but it takes a LONG time to make it work well. Corn shipping starts immediately and you shorten the game by taking vps/captain really early in the game. It is also independent of ALL the buildings, but warehouse and hospice are high on your list. The game also changes depending on the number of players. You probably played consistently with the same number of players (I would guess 3, as 4 and 5 are more balanced). Your changes likely make the game worse and unbalance it more than the problems you are trying to fix.

    ReplyDelete
  4. @Steve Your assumptions turn out to be wrong. I have played around 400 games of Puerto Rico, so I suspect that I have played it enough. I have 'only' played ~50 games of 5 player, ~150 games of 4 player, and ~200 games of 3 player. I don't have a world champion plaque on my wall but I do play with someone who does and I consider myself *slightly* better at PR than him. (Which is not to say I consider myself the best in the world, certainly not, but I am pretty dangerous at PR.)

    The idea that you can just get corn and ship it is a nice fantasy but when playing PR against opponents who know how to play everyone drafts corn over everything else. You can get a lot of corn by drafting Settler and first picking corn constantly, in which case your opponents are going to walk away with the game by drafting the good jobs with bucks on them. Simply put, everybody drafts corn hoping to get as much as possible and if they get a lot then they ship it. This is a given, so the only way you get to play "ship all the corn" is if your opponents just stand back idly and let you corner a valuable resource. Since they all need corn for their factory or their harbour... good luck with that.

    ReplyDelete
    Replies
    1. Well, of course, but that means they're playing defense to that strategy. Which makes it a viable strategy if they're not. PR is one of those games which you win if you do the thing that others are not. If 3 people are building, your shipping will probably win, because the 3 are fighting over the same stuff. And I actually meant make *mostly* corn. Overspecializing is bad in PR. It just surprises me that your 2 top strategies insist on 3-4 production buildings, because that hasn't been my experience at all. I've seen wins with 3 or even 2 production buildings. I've seen the Large Market strategy work (and players build it into their plan). I'll admit that the Hospice comes after the Construction Hut and Hacienda for useful small buildings (so it might be better cheaper), but none of the things you mentioned (Sm. Indigo, Sm, Suger, Sm. Market) are required for a win. I commented because my experiences playing haven't shown the diversity strategy to be necessarily better. Sure it wins sometimes, but it's hardly as definitive as you suggest.

      Delete
    2. I would say that the top Builder strategy insists on all the production buildings eventually, but 3-4 to get the Factory engine going. (Whether it is 3 or 4 depends on competition for Guild Hall.) The best Shipper strategy obviously picks up the small production buildings so as to diversify and maximize the Harbour and usually picks up either tobacco or coffee to be a good seller (and more shipping, too). Getting all five goods as a Shipper is much less common but is super strong if you can manage it.

      You say that if there is a 3 builder / 1 shipper split that the shipper wins and I don't agree with that. There are a couple reasons. First, the builders all desperately want to pick builder and focus on quarries to constantly spew out cheap buildings to end the game on buildings as quickly as possible. They are going to focus on mayor and settler secondarily and avoid craftsman like the plague. They are also going to cooperate to jam the boats with whatever goods the shipper can't make. The shipper is not going to have people calling craftsman / captain for them so they will have to do it themselves and that is a huge problem in the endgame when you need every shipping phase you can get before the game ends on building space. My experience in a game where three people are trying to build and one ship is that the shipper gets blown out and whoever gets the Guild Hall wins the game handily.

      In a game where there are three shippers and one builder the reverse obviously happens. The builder gets the Guild Hall by default, which is great for them, but in the endgame there are so many craftsman / captain calls that the one shipper who managed to get both a harbour and a wharf blows everyone out. The builder just can't make the game end quickly enough with everyone else cooperating to prevent builder / mayor phases.

      Delete
    3. Well, I've played more 5-player games than you, but fewer 4 and only a handful of 3. There's a reason that I guessed (correctly!) that you played more 3-player games than anything else. Your breadth analysis is tilted heavily towards a 3 game, as some of the problems you listed simply don't happen in larger games. More players=more complexity=more breadth. That is also the *reason* I haven't played a lot of 3-player games. The 3-player game is, as you said, kind of broken, but I think its a mistake to assume that the 4 and 5 player games have the same problems. Perhaps you should try the expansion if you find PR lacking?

      Delete
  5. I was kind of meh on these changes until The Steve mentioned corn shipper as a 'top-tier strategy', which it is when playing with poor players (which I suspect is the real difference between The Steve and Sky's experience with the game).

    The issue is that as a reasonably good player (I've made a final table at WBC but am significantly worse than Sky) I will draft a corn over a second anything under virtually any conditions, and there are only a few things I would draft my first of before corn (my first money good (including sugar) if it's exclusive and I can build the production building already). This means I 'counter' the produce a ton of corn strategy by default.
    Why do I do this? Because as Sky says: 'Everyone finds that they never have enough dudes to run everything but it turns out the buildings that make dudes aren't worth buying and it is far superior just to not have enough dudes.' Since I play the game short on dudes corn is better because it's man efficient. The first coffee is good because it gives me a good sell, an additional good, something to jam up a boat with, etc, but the second one isn't twice as good as a corn.

    So now I like these changes because it devalues corn relative to other goods, and might lead different strategies. I'd be in for giving it a spin next I'm in TO.

    ReplyDelete
  6. @Steve

    You suggest that I said the three player game is broken. I didn't say that, rather I said that the game, when played with top tier people, lacks breadth. In fact I like the three player game better than the other games though I find that (and nearly everyone I play with agrees) the five player game is the worst option by a long shot and is way too random.

    I tried the expansion but I wasn't a huge fan. Taking out standard buildings does shake up the game some so it is a way to brute force breadth but I ended up not liking the expansion buildings very much. I also found that if certain buildings were omitted there often was only one decent strategy and everyone struggled to do exactly the same thing - you can't play a selling game if there aren't good buildings for selling. If I wanted to mix and match expansion buildings with baseline buildings I could probably make a really nice set though and that is an idea worth exploring.

    @Pounder

    Well then I guess I need to get you to TO soon so we can test this out!

    ReplyDelete
  7. That "way too random" factor is the breadth that you were complaining about in your post. I assume you mean by that that the strategies you mentioned as untenable can win in the 5-player. In the 5 player, many strategies are competitive. In the 3, only 2. This is the problem I have with your post. I mention that the 5 has few of the problems you complain about, and you blow it off as "random". Do you mean that better players will not still be better in 5-player? Do you mean that I can just pick a random role every turn and win? Do you mean that the game magically converts to 100% luck and 0% skill? What exactly do you mean by "random", and what makes it bad? As someone who enjoys the 5-player game, I hope you can explain your dislike better than "its way too random".

    ReplyDelete
  8. The issue I have with 5 player PR is that so often the best player does not win. You can set up for a particular building or strategy and get blown out or just have the person after you enjoy calling craftsman and there is really nothing you can do to avoid getting smashed. Because there are so many turns in between each of yours (sometimes 8 of them) and because buildings are so limited I found 5 player PR to be just about taking the best thing available each turn rather than building a plan and following it.

    To put it another way I found that in 5 player if player A and player B are in the lead then the game usually devolves into A thinking "Well, I win if C does this one thing and I lose if he does the other thing and I can't influence that at all". It feels similar to Settlers in that way - if I am going for longest road and some guy decides to go all in for longest road against me I lose because I either commit too much to beating him or just cut my losses and don't win anyway. In 3p or 4p PR though I can control the board more and react to people doing unlikely or unfortunate (for me) things and recover if I am good enough. In 5p it too often goes "Well, Bob built the last Tobacco Storage so I guess I get boned. GG!"

    So I guess my position is that 5p PR isn't a game I enjoy very much (and by and large this is the consensus of people I have spoken to about it) but that it does end up having more breadth just due to the limited number of buildings.

    ReplyDelete
    Replies
    1. And so you finally see my point. Your issue is with PR "problems" and yet the problems you cite are specific to the 3p game, not the 5. Sure, the "diversity" strategy might be overpowering in 3, but it certainly isn't in 5, due to lack, which is the game's balancing mechanism. We never talked much about the 4 player game, which is about half-way. The diversity strategy isn't as good as the 3, but it's stronger than in the 5. You applied your statements to the game as a whole, and many of them were either not true or exaggerated in my experience. (ie "no matter what boats get made you get to ship most of your goods") That's how I could tell that you played 3p a lot.

      Delete
  9. What I would like is for you to address specifically why you think the 5p game is better baseline than in the way I suggest. Baseline the Hospice, University, and Large Warehouse are complete garbage and the Office is situational at best. Making those more competitive has to open up new top tier lines of play. I spent a few days rereading PR strategy guides to see if people tended to agree with me and everyone does, regardless of how many players are in the game. Diversity is the single best strategy (obviously you get all the corn you can, everyone does) and those who end up losing out on Factory / Harbour lose the game the great majority of the time because those buildings are so powerful. In 5p making a ton of a single resource is a terrible idea and indicates that you are going to get beat by good players. Sure, a good player playing a suboptimal strategy can win against scrubs but if you are playing top tier people you need to play the best strategy and that is remarkably narrow. To be fair the restricted nature of buildings means it is *less* narrow in 5p but my changes still open up more possibilities for competitive tactics. So again, there are my reasons why I think my changes are good and help to open up new avenues of optimal play. Why do you think they are bad specifically?

    ReplyDelete
  10. Hmm, shall we have a look at your changes? The winner had Hospice, Hacienda, Lg. Indigo, Tobacco, University, Factory, Lg. Sugar, Harbour, Guild Hall, Fortress. In other words, a diversity ->factory, harbour, guild hall win. Wait, isn't that your main bitch? "Well, there were two Builders and the one that got the Guild Hall won." How exactly has that fixed things? The *exact same* buildings that you claim were overpowered...are still overpowered, and still won. Not only that, you completely *broke* a third of the game. Remember this? Count the number of empty spaces on buildings, and *add that number of colonists*. Well, you just eliminated a huge chunk of those empty spaces. Did you change the number of colonists to suit? Is the end condition *ever* going to be running out of colonists with your changes? You complain about your persistent lack of dudes, but that lack is the reason to take the mayor. What do your changes do? Make it *less* optimal to take the mayor, and get the extra guys, and speed the end of the game, because you made the dudes more efficient. The reason the diversity strategy is no longer optimal in 5p is because of lack. Will you make all 5 production types? No, building lack. Will you make lots of money with your factory during production? Not always, there will be resource starvation. Is trading better for you? Not 25% of the time, when you're last, or even second last to trade. *Will you ship most of your goods during shipping*? No, you will ship ONE type of good, and the rest will go to waste, because of lack of ship space. These limitations are the thing that makes other strategies&buildings viable, and this is the reason 3p lacks breadth. I might buy your argument that the hospice and uni suck, but the solution is already there - sub in expansion buildings. But any warehouse in 5p is never garbage, because it prevents blowout when the person to your left takes captain.

    ReplyDelete
  11. The 5 player game is so random I don't think you can call any strategy in it viable. You can't make a strategic plan of any kind because you don't have the ability to influence the game enough to make it work.

    ReplyDelete
  12. Random elements are those which are not chosen by players. Dice are an excellent example. In Monopoly, the dice are a huge element of the game, and consequently, Monopoly is largely luck. There is only one random element in PR, and that is the choice of Plantations. Just because most people cannot predict the moves of 4 others and how that will affect their position does not, in fact, make it random. The 5p game is not more random than the others. However, you DO have less control, as you control 20% of the moves, rather than 25% or 33%. You also do not get fewer moves/game in PR, as the colonists and VPs change proportionally to the players. It is this reduction in control that you (and Sky) dislike, and not "randomness".

    ReplyDelete
  13. @Steve

    Not having enough guys is a game constraint, a challenge to be overcome. This is a *good* thing, not a bad thing. A big component of PR skill is to figure out which buildings and plantations to activate and which to leave empty and if floods of colonists come in then those choices are removed. My design means that the hospice and university are better so people can choose to spend resources to get more colonists as a solid tactical choice but doesn't remove the pressure of not having enough colonists as a form of pressure in the game. That pressure can be frustrating to work around but finding the best path around difficult constraints is the essence of interesting game choices. The number of incoming colonists will go down with decreased open slots to be sure but it will go up with increased usage of Hospice / University. Will it result in exactly the same rate of colonists coming in? Probably not, but I challenge you to demonstrate that the old rate was the right rate and that the new rate is the wrong rate.

    You judge the game I played with myself as if it is representative, and you also are doing it wrong. Just because the player who won had Factory / Harbour doesn't mean those buildings and the strategy they represent is just as dominant. The player that had them was solo in Tobacco while the others fought over Coffee. That same player happened to have multiple quarries and picked up the Large Sugar as the cheapest way to build out the game asap. That player might have done something else but when builder is called and you have a university, a manned sugar plantation, and four bucks worth of buying power you buy a large sugar mill. The Guild Hall was a consolation prize because the Residence *and* City Hall got bought first by another player. So much for Guild Hall dominance. The Harbour this player bought was reasonable but only managed about 4 points so it was an acceptable but unexciting purchase. The player won because of being solo Tobacco and being able to make lots of it cheaply and getting some great positioning on critical plays.

    Subbing in expansion buildings is a possibility. However, I felt that simply writing off the basic buildings was wasteful and that it was better to make them competitive than to dismiss them entirely. Also it should be noted that the Guest House is well designed but many of the expansion buildings have overpowered combinations (Forest House/Haci) (Union Hall/Warehouse). It is certainly true that you can increase breadth just by judiciously including expansion buildings but I think my changes actually increase it more.

    Oh and by the way the statement "any warehouse in 5p is never garbage" is absurd. Sm. Warehouse is a fine investment, Lg. Warehouse is rubbish.

    ReplyDelete
    Replies
    1. Sorry, I'm pretty much done talking about PR. I could actually prove mathematically that removing building spaces breaks the game-ending mechanic, but I'll leave that to you as homework. Later.

      Delete
  14. @Anonymous

    You are attacking a straw man here. Neither Nick nor I is claiming that there is more dice rolling type randomness in 5 man PR. What we are arguing is that the win rate of a superior player in 5p is drastically closer to what would be expected if every player made completely random decisions.

    If, for example, every player acts randomly then in 5p you would expect a 20% win rate and in 3p a 33% win rate. However, in the case where you have a one superior player and the rest moderate players you tend to see a win rate of more like 80% in 3p and 35% in 5p. Simply put, in 5p the results tend towards the results one would expect with random play and in 3p the results trend towards the good players winning. That is what we mean by random.

    Another example is a game of pure skill where everyone plays for an hour moving pieces around a board and at the end the player with the lowest score picks someone else to be the winner. Sure, there is some skill in not being in last place but I will still deem that game to be exceedingly random (and also crappy!) and that randomness is not mitigated by the fact that a human is making a choice. Just like in the example above a superior player will have a win rate roughly 100/(1-N) and that isn't something I enjoy.

    ReplyDelete