Refinement of point salads

May 25, 2014

About a decade ago, prof. Hiroyuki Iida of the Japan Advanced Institute for Science and Technology (JAIST) proposed a Theory of Game Refinement. This theory describes the elements that a game must have to be “refined,” which translates roughly to “entertaining to play for experts.” The theory states that the three main pillars of game refinement are:
(1) Complexity (noble uncertainty): discovery of new tactics should be possible;
(2) Fairness (draw ratio): opponents should match in playing strength; and
(3) Refinement (see-saw game): the outcome of the game should remain uncertain for an ‘optimal’ length of time.

The last pillar evidently is considered the most important one, as the whole theory is named after it. Indeed it is, as the other two pillars more or less follow automatically from it. To see why, you have to understand what the third pillar actually says. The refinement pillar states that a game has an optimal playtime length, and that the outcome of the game remains undetermined for most of that time. The optimal playtime length depends on the game itself. For instance, we may assume (since professional matches gravitated to these numbers) that the optimal length of a game of Chess is 3 to 4 hours, while the optimal length of a game of Soccer is 1.5 hours. A game is refined if victory remains within reach of each of the opponents until close to the end of the playtime.

The complexity pillar follows from the refinement pillar, as a game that is not sufficiently complex that players may devise new tactics is simply boring, and the optimal playtime length approaches zero. I call this ‘meaningful play’: the game must constantly offer the possibility to make interesting decisions. So-called ‘roll-and-move’ games, wherein you roll a die and move a pawn the number of spaces shown on the die, then do what is printed on the final space, do not have interesting decisions at all. In contrast, in Chess, in particular in the mid-game, all moves progress the plans of a player and fundamentally change the state of the game in a way that players must respond to intelligently.

The fairness pillar follows from both the refinement pillar and the complexity pillar: if one of the opponents is inherently much stronger than the others, a game which is built on meaningful decisions will be decided long before the optimal playtime length is reached. As adults usually have tactical insights while small children do not, a game between parents and their children can only be fair if tactics play no role. The reason that ‘roll-and-move’ games are common pass-times for adults with kids, is that the ‘roll-and-move’ mechanism completely levels the playing field as far as tactical insights are concerned. Naturally, such games are inherently boring to anyone with an ounce of tactical sense, which unfortunately has led to the misconception held by many adults that board games are boring and only meant for very young children.

While I have not seen the Theory of Refinement applied explicitly to modern board games, it seems to me that most designers of modern board games take the theory to heart implicitly. Modern board games are supposed to have an optimal playtime length, which is usually printed on the side of the game box. Typical playtime lengths are between 1 hour and 2.5 hours. While I have found that often these claimed lengths are a bit on the optimistic side, you know that you can finish one or two of these games in an evening and are supposed to be entertained most of that time. There are some older games (I am talking decades, not centuries, here) which have playtime lengths which are rather unpredictable, or which have long, drawn-out endgames in which everybody already knows who is walking away with the victory. However, in the games published in the last decade you seldom see that anymore.

Modern game developers usually keep grips on the length of the game by introducing a mechanism that progresses the game towards the end with a steady pace. Examples of such mechanisms are: having a limited number of turns, having a board that fills up, having a draw pile of cards that gets depleted, or having a limited amount of available resources that get used up. These mechanisms tend to work well.

After setting both a lower and an upper limit to playtime length, the problem that game developers then have to solve is how to keep the outcome of the game uncertain for almost all of that time. A recent solution to this problem, which many popular game developers apply, is a rather ugly one: victory is determined by points, and you get a LOT of points during the game, for a high variety of activities. This is usually coupled with a complete lack of insight in how many points each player scored and/or the awarding of a large number of end-game bonus points. Recently the term ‘point salad’ was coined for such games.

The consequence of a ‘point salad’ design is that every move that a player makes seems to be a tactical one as it scores points, but that no players (except perhaps those who have played the game a lot) will know who is ahead until the end of the game. This might seem to satisfy the refinement requirement that the game’s outcome is undetermined for most of the playtime, but I would argue that it does not. The reason is as follows. In every game, either tactics have no influence on the game’s outcome, or the tactically best player will win the most. If tactics have no influence, decisions are not meaningful, and the game is not refined. If the tactically best player usually wins, it means that it is known which of the myriad of point-scoring actions in a point salad game is the most beneficial, and only for the tactically weak players the outcome of the game is unknown until the final tally.

Naturally, there are nuances in this argumentation. If a point salad game is so well-balanced that no dominant strategies exist naturally, a tactical fight between strong players of such a game might be interesting and entertaining. Strategies are very hard to balance, though, and it is common that, within a relatively short time after publication of a point salad game, players have identified the one or two dominant strategies of the game, i.e., the strategies which allow the collection of just enough extra points to lead to victory. It still means that every player at the end of the game has so many points that the ending scores are pretty close, but the knowledgeable player knows that the final ranking was inevitable.

Still, despite the rampant criticism on the point salad mechanism, point salad games tend to be quite popular and sell well. I can explain the popularity as the result of the many colorful components of such games, the fact that they usually are not overly competitive or punishing (if you cannot do action A for 7 points, you do action B for 3 points now and 4 points later), and the fact that players can easily socialize with each other around the game table rather than sit in silence burning their brains.

From that perspective, as an activity, point salad games can be entertaining. But as games they seldom are refined.


Substance over form

May 16, 2013

Stefan Feld has been appointed headmaster of the Uwe Rosenberg School of Crappy Game Design.

Let me rephrase that: in recent years, several concepts have become popular in the design of modern board games that I believe lead to weak designs. Two noteworthy and popular designers who apply these concepts are Uwe Rosenberg and Stefan Feld, but they are definitely not alone.

Before I am going to discuss these inferior concepts, I need to give an explanation of what I think makes a good game. In my opinion a good game is a game that has tactical and strategical depth following from a parsimonious ruleset. To put it in simpler terms: A good game is easy to learn and still offers plenty of opportunities for exploration. It follows that a good game has little randomness (because much randomness makes choice less relevant) and has interaction with opponents (because without, a lack of randomness would allow straightforward calculation of the best moves). In this vein, Go approaches perfection in game design: with only four rules it can be taught in 5 minutes, but it offers a lifetime of tactical and strategic exploration.

Now, besides offering some challenge, modern board games also must be fun, or at least interesting, to play. And fun can be incorporated through different means, for instance by adding a captivating and visually good-looking theme, or by offering players a lot of diversity in choices. So-called ‘Ameritrash’ games usually have overly complex rulesets and a lot of randomness, but are so theme-heavy that they may become a lot of fun to play if the theme appeals to you (for me personally, Battlestar Galactica is the core example of a game that sins against all my mantras for good game design and is still an enormous amount of fun to play).

In recent years, I have observed a trend in modern board game design that has become highly popular but which I think is detrimental to the design of good games. This trend really took off with Uwe Rosenberg’s Agricola, which is why I attribute it to him. Let’s call it the Rosenberg Style. The Rosenberg Style has the following features:

  • The game looks good, with a rainbow of colored wooden or cardboard bits, to provide an instant ‘Wow’ factor leading to quick sales.
  • The game has a crapload of cards, where each card represents a different rule, so that players believe that it has a lot of diversity (alternatively, for ‘cards’ read ‘action spaces’).
  • The game has a myriad of different ways to score points.
  • The game obfuscates the final score, so that all players will always have the feeling that they “just might win”.

The Rosenberg Style might lead to interesting games. Agricola, for instance, is a game that I enjoy. I don’t think it is a great design, but because of the tension that the game provides between scoring and (basically) eating victory points, combined with the fierce battle between players over the scarce resources, provides a nice challenge.

However, more often than not it leads to games that are simply inferior. Rosenberg’s latest, Ora & Labora is just terrible. There are so many ways to score points in that game, and you will score so many points during it (which are only calculated at the very end), that you never have any idea who is ahead. The game comes with hundreds of cards, all of which will be played during the game, each card representing a different building with different ways of using it. You can use all the cards on the table, even those in front of other players, which are very hard to see, and of which even the exact positioning matters. It comes with several hundreds of little two-sided cardboard chits, of which each player has dozens of different ones in their resource pool, and which have a tendency to accidentally fall over to their other side, which has a big impact on the game. The game just handles badly. Even if it would be a good game otherwise, its physical attributes alone make the design just a plain awful.

The weird thing, to me, is that Ora & Labora is quite popular, to the point that many people call it superior to Agricola. I have been wondering why that is, and I have concluded that Ora & Labora is more an activity than a game, and what many people look for in a game is something to do rather than a challenge to overcome. Because the scoring of Ora & Labora is more or less invisible, there is no need to worry about it. Because you have so many possibilities, there is always something that you can do. If you can’t quickly find an action to perform, you take some resources — that’s never bad. If you want particular resources that don’t seem to be available, you take some action — those are never bad. Or you construct a building — there are several points right there. If you want, you can set a strategy at the start or during the game — “this time I will try to make lots of whiskey and beer and get points by turning them into relics!” And if that doesn’t pan out, you just change to another strategy on the fly. It is not a bad thing that a game allows switching strategies, but in Ora & Labora you can do it because it does not seem to matter.

To be frank, I believe that in Ora & Labora strategies actually do matter, and that there are definitely ways to play the game that are superior to other ways. But to get to that level of thinking, you must have played the game so often that you know all the cards, and you can follow them on the table. It will probably take at least half-a-dozen plays to get some ideas in that direction, and several dozens to reach the minimal level of knowledge needed to be a strong competitor. But Ora & Labora got rave reviews from the outset, from people who played it once or twice. Why? Because it is an activity. You do stuff. It does not matter if you really know what you are doing. You always have choices, you always score points, you have a feeling that you are in control because you have lots of resources, and at the end someone wins. Maybe with 300 points, but as you had 285 points, you did quite well, didn’t you?

It is my firm opinion that in a game choices should matter. I loathe roll-and-move games (like Monopoly) because there are no real choices to be made. The Rosenberg Style often (but not always) leads to games where choices do not matter much as they are balanced to provide every player with about the same amount of points, regardless of what they do. The main saving grace of these games is that they hide the fact that choices do not matter under a thick layer of diversity and colorful components. (I should point out here that I think that Rosenberg’s own games that use this style tend to be more interesting than those of most of his colleagues.)

I think that one reason that we see so many new games using the Rosenberg Style these days, is that it is relatively easy to design games this way. You have a game with a board that allows players to get resources, and buildings that convert resources. You give it ten types of resources and create 200 buildings, all along the line of “The Smithy turns 2 Iron and 3 Gold into a Horseshoe and 7 Nails”. You allow players to construct those buildings using resources, and use them with workers. Each building is worth points, some resources are worth points, and quite a few buildings allow turning resources into points. At the end you add up all the points. And there’s your mechanism.

The main trick, where the designer is concerned, is to devise a good theme to give those buildings interesting-sounding names and convince the publisher that they should hire a good illustrator and invest in nicely-colored wooden bits. You might think that the building design is pretty tough to keep balanced, but it is not really that hard, because of three reasons: (a) it is usually relatively easy to come up with some rules of thumb to decide what the building should cost to construct; (b) as you allow every player to use every building, any player can make use of an overly strong building; and (c) you want a little bit of unbalance (though not too much), so that players feel that they need to fight over some buildings.

Rosenberg Style games are often forgettable, in the sense that people rave about them when they come out, but no longer play them two years later. Ora & Labora is already on its way out. And why not? These games are similar to action movies that hinge on special effects: once you have seen them, they no longer hold interest, and the effects are dated in a few measly years.

My main annoyance with Rosenberg Style games is their brief popularity. I do not play games on my own. I need others to play with me. And if these others are keen to switch from one Rosenberg Style game to another, I never get the chance to really explore a game that deserves exploration. Fortunately, in the groups where I play there are enough people who want more from their games than just an activity. But even they get swayed now and again.

That’s fine in a sense: I can be entertained by “just an activity” too. As long as I get to experience games designed on substance rather than form with some regularity, I’m good.


God plays dice

January 26, 2010

Don’t worry, this is not a post about quantum mechanics. It is about randomness in games. Specifically, I want to discuss  the different kinds of randomness you can have in board and card games.

Typically, a game that works with dice is a game with randomness. In general, I do not enjoy randomness, and I am not alone in that. Still, there are games with dice that I enjoy. The reason is that there are different kinds of randomness. Broadly speaking, we can distinguish two: non-determinism, and imperfect information.

Non-determinism means that certain decisions of a game are resolved in a random manner. Imperfect information means that aspects of the game are unknown to one or more players. At first glance, one would say that only non-determinism amounts to randomness, while imperfect information does not. However, imperfect information can be the result of randomness, and that is the kind of randomness that I can enjoy!

Let me give an example of randomness: the game of Risk. In Risk, a player might decide to attack another player with a certain amount of troops. He then rolls dice to determine the damage that the troops do. His opponent defends by rolling dice. The two dice rolls determine how many troops are removed from the game. This is a typical example of non-determinism: the randomness gets applied after the decision to attack has been made.

On the opposite side of the coin we have a game such a Poker. Poker is a game of imperfect information. Each player knows his own cards, and open cards of his opponents, cards on the table, and the bids that the opponents made. They do not know the hidden cards of the opponents. Therefore, they have to make a decision in a situation with unknowns. There is no non-determinism involved, however; it is not as if the best hand will be decided by a die roll or something. The best hand is the best hand in the current situation, and the fact that part of that current situation is unknown makes the game interesting.

Some would argue that Poker has non-determinism in the dealing of the cards. That, however, is not non-determinism, even though it concerns randomness. The initial game situation is set up randomly, but that all happens before the first decision of a player. Therefore, Poker is purely a game of imperfect information.

There are no dice in Poker, but there are many games with dice that are imperfect-information games. An example is Backgammon. In this game a player rolls dice to determine his possible moves. He then decides which moves to make. The player has complete knowledge of the state of the game when he makes a decision, and the outcome of the decision is purely the result of his decision. However, whether or not it was the best decision depends on the situation in which the opponent must make a follow-up decision, and that situation will be partly decided by the rolling of dice.

To give a very simple example to distinguish non-determinism and imperfect information: suppose I have a game in which I have two pawns that move along a track. How far a pawn moves is determined by a die. Now, if the game rules specify that I should first decide which pawn to move, and then roll the die to determine how far it moves, that is non-determinism. If the rules specify that I first roll the die and then determine which pawn to move, that amounts to imperfect information. Personally, I’d say that the second game is more interesting than the first, as it may allow deeper tactics. Although I probably would still rather play Arkham Horror.

I recently stated in a newsgroup that the difference between imperfect information and non-determinism is similar to the difference between science and religion. In science, we examine in what shitty situation we are and take steps to get out of it. In religion, we place our trust in fate to get us out of shitty situations. I thought this was a fun observation but I got a warning from the moderator: I had not surmised that my (fairly innocent) remark might offend religious people.

The moderator was probably right. Still, the response brought me a slightly deeper insight in why I prefer imperfect information over non-determinism: it might be because I am a scientist. I don’t mind being thrown into a shitty situation in a game, and I don’t mind that my opponent gets into a less shitty situation. It is the situation I have to work with and which I have control over. It is a challenge to overcome, and the shittier the situation, the sweeter the victory if and when I manage to achieve it. On the other hand, winning because the dice fell favorably just feels hollow.

Well, yeah, I can relish in the knowledge that Fate was on my hand, but that means that Fate, not I, won the game. And frankly, if Fate really must win a game, I say let her pick her own game group.


Protection racket

November 29, 2009

The abbreviation IP stands for “Intellectual Property”, and IP rights are government-granted, exclusive rights to exploit some intangible assets, such as music, art, and designs. Due to the ease by which digital material can be copied by computers and spread by means of the internet, IP rights have got quite some attention in recent years. So much, in fact, that a whole branch of lawyers has been bred to deal specifically with IP rights.

I am very much in favour of a system of IP rights. If someone creates a new and unique work, that person should be able to exploit the fruits of his efforts in a way that pleases him or her. If someone makes a successful design, and later finds someone else has copied that design and gains considerable profits from it, it is only just that the orginal creator shares in those profits or even can prohibit making the profits.

Individuals and companies have enjoyed the benefits of IP rights for quite some time. But the rise of the Internet seems to have struck fear in the hearts of many rights holders. Or rather, it seems that IP lawyers have seen it fit to strike such fear in the hearts of IP holders. Two cases:

Some time ago I had a website where on my “about” page I had a cartoon which I had copied from a book of cartoons. This cartoon was quite apt at describing me personally, and I liked it a lot. I knew that copying from a book is allowed if it is short and an adequate reference is made, so I added a caption to the cartoon, stating exactly from which book it was, and who the author was. At some point I received a so-called Cease & Desist (C&D) letter from the author’s lawyer, stating that I had illegally copied a cartoon, and that I should remove it immediately. I complied. However, I also wrote a nice email back stating that I was not going to act against the wishes of the author, but that I believed that I was actually in my rights in copying that one cartoon, and I thought that having that cartoon in that specific place would actually help the author in selling more books. Naturally, I did not receive an answer. I suspected, however, that this particular C&D letter was not sent on orders of the author, but by an over-zealous IP lawyer who had been hired to scan the whole Internet and threaten anyone who was “ripping off” the author. Frankly, I did not see how that lawyer would be worth his money by threatening me, as I certainly would not cost the author any sales.

The second case I wish to offer is a lot bigger. There is a great website about board games, called BoardGameGeek (BGG). This is a community site for and by board gamers, on which thousands upon thousands of games are reviewed, commented, discussed, and reported upon. It contains a wealth of material on every board game in existence. Now, a group of IP lawyers, on behalf of the game publisher Games Workshop (GW), has sent a C&D letter to the maintainers of BGG to remove any and all files on their games from the website.  The maintainers complied, and removed hundreds of files which supported the GW games. They removed helpsheets, user-created scenarios, and rule summaries. They removed everything that the community needed to enjoy the GW games.

The big question is “What does GW hope to accomplish?” Do they strive for more profits? Well, then this is not a good action to achieve that. Profits are gained by selling games and game materials. Games are sold to gamers, who are often part of gaming communities. This holds in particular for GW games, for which many communities exist. Such communities want and need the community-created materials which are no longer available. Will the members of these communities be happy with GW’s actions? Of course not. They are outraged. There is no doubt in my mind that GW lost many sales by sending these letters.

I can think of no good, business-related reason why GW would act this way. But they seem to think they made a good move here. The only reason I can think of is that they were approached by some IP lawyers, who, on the lookout for more business, made the publishers afraid by telling them they had discovered many of their “IP rights” up for grabs on the interwebs, and that they could help GW to mitigate the damage if GW would only pay some “protection money”.

It is not hard to alienate your customers, which is what GW has done. And they will have a hard time winning them over again. The only winners here are the IP lawyers.

It is good to have IP rights. And having rights means needing some protection for them. But it also means having the ability to graciously allow the world access to them. Not arbitrarily, but judiciously. If it helps you reach your goals, why not?


Longest game ever

October 8, 2009

I recently finished the longest game I ever played. It took about 50 days from start to finish. It concerned the popular board game Battlestar Galactica played by forum.

We all know that usually tie-ins to popular books, movies, or games are not very good. Harry Potter books are OK, Harry Potter movies based on the books are not not so good, and Harry Potter video and board games based on the movies are pretty bad. But there are exceptions. One such exception is the Battlestar Galactica board game. On the famous website BoardGameGeek, on which thousands of boardgames are reviewed, this game actually ranks in the top-20 (in comparison, Settlers of Catan‘s rank is between 40 and 50).

Battlestar Galactica is an SF television series, spanning four seasons, in which the human race is virtually extinguished by robots called Cylons. The last remaining humans are fleeing with a fleet of spaceships, led by the Galactica, in search for the legendary planet Kobol to restart human civilization. The twist of the series (in which it differs from the 1980’s series with the same name) is that the Cylons can appear as humans, and there are actually Cylons posing as humans on board. Some of the Cylons do not even know their own nature, until they get a wake-up call.

The Battlestar Galactica board game, created by Corey Konieczka and published by Fantasy Flight Games, ties in to the first season of the television series. The game is for 3 to 6 players, but is best with 5 or 6. It is a cooperative game, which has two teams: the humans and the Cylons. The loyalties of the players are unknown at the start of the game; they all appear human. Someone might know that he is actually a Cylon, but it might also happen that someone thinks he is a human, while halfway through the game he finds out he is actually playing for the Cylon team.

In the game, each of the players fulfills the role of one of the 10 major characters of the series. They move on the ship, collect skill cards, perform actions, and try to deal with crises that pop up every turn. The goal for the humans is to reach the planet Kobol by performing several faster-than-light jumps. The goal for the Cylons is to stop the humans from completing their goal, either by destroying the ship, or by dropping one of the key resources of the ship to zero.

The game is quite complex because it offers a myriad of choices, and the game’s mechanics are not very engaging. These are both minus points. However, where the game shines is in its atmosphere. It is all about the interactions between the players. The humans know that there are Cylons amongst them, and if they can smoke them out, they can try to make them harmless. The Cylons, on the other hand, try to stay hidden while not helping the humans too much, until they see their way clear to do some big  damage. In the meantime, the crew tries to stay alive under a bombardement of crises. The atmosphere of the game is one of paranoia. There are dangers lurking from every side. Sometimes the humans luck out, sometimes things go wrong — but if they went wrong, was that just chance, or was a Cylon actively messing with them?

I really like this game, but I cannot bring it to the table as often as I would like. It is quite a long game, usually spanning 2.5 to 3 hours. And while most players like it, the complexity might seem daunting for new players. Also, some players are turned off when they realize the game ties in to a television series that they have not watched. Actually, the game can be enjoyed by anyone, though it does come alive more for those who have some experience with the series.

On BoardGameGeek people have started to play Battlestar Galactica through forum posts. A moderator creates a forum thread, and invites five or six players from a list, who then have in-game discussions and post their actions in the thread. The moderator sends players their cards and executes random factors when necessary. It works really well, and the game is a joy to play in this fashion. It is, however, different from the tabletop game, as it is much easier for players to track all in-game information. This makes a Cylon’s job quite a bit harder, as messing with the humans will give his identity away much faster. Still, the Cylon team wins about half the games played this way, so it seems balanced.

Reading the threads is also a nice way to get familiar with the game. A few threads which I recommend:

If you want a list of all the games, visit the BSG_PBF Wiki. If you would like to read a thread without knowing the identity of the Cylons, you should watch out, as that information is found on the Wiki. If you own the game and wish to sign up for a PBF experience, you can do that through the Wiki also. Owning the game is a must, as Fantasy Flight Games does not allow the publishing of card texts in the threads.

Highly recommended, both the game and the forum-playing of it. And OK, also the first season of the television series. So say we all.

Addendum (July 3, 2010): In the meantime, I finished two more games of Play-By-Forum Battlestar Galactica: BSG 56 – Battlestar Isoroku, and BSG 72 – Battlestar Lincoln. Both were good fun, but if I have to pinpoint my favorite of the four games, it is Battlestar Isoroku. It was exciting, tense, and quite funny. I think it will make a good read. Should someone take me up on that, I challenge you to guess who the second Cylon is before it is revealed.


Board gaming etiquette

August 15, 2009

One issue troubles virtually every board game that supports three or more players: kingmaking. Kingmaking means that a player who is not going to win, with his actions may hand the final victory to one of the other players.

As games are about interaction between players, usually the potential for kingmaking is unavoidable. In some games the possibilities a kingmaker has are bigger than in others. For instance, in a wargame (such as Risk), if one player consistently attacks a specific opponent, regardless of the current game state, that particular opponent is very unlikely to win. Neither is his attacker.

Personally, I do not like to have kingmakers in my games. I enjoy neither losing nor winning because of a kingmaker’s actions. In my view, good boardgaming etiquette should prevent players from kingmaking. But what kind of behaviour should the etiquette prescribe?

Think of the following situation (which everybody has encountered). You are playing a game, and you are in a good position to walk away with the final victory. The other players notice this, and suddenly everybody is attacking you. You get beaten down, while the until-then-second player is making the victorious move.

This feels fair to many people, but is it? Basically, the player who performed best until just before the very end, ends up in second place, while the one who did good but not best, ends up being the winner. And that because of, lets be fair, kingmaking actions. The players in third and lower positions had nothing to gain by attacking the leader; they had no chance of winning anyway. The only reason for their attacks was to postpone the game’s finale.

My own behaviour in board games is to try to finish in as high a place as possible. If I am in first place, I try to maintain that position. If I am in second place, I try to dethrone the first-place player and take over his position. If I am in third place, I am striving to gain second place and from there climb to first. Etcetera. Until recently I thought this was proper behaviour, but now I am less sure.

I recently played a game of Small World (which I alluded to in a recent post) with three players. One of my opponents showed a keen interest in attacking the other one of my opponents. He did so without regard for his own position. After the first two rounds, it was clear that I was probably going to get the final victory, while the victim was in second place, and his nemesis in third. Yet, the attacker did not lay off.

I told the attacker that he was playing a stupid game. He should attack me, because I was in first place. But then I realized that that is exactly what my personal gaming etiquette prescribes to do not. As the attacker was now in third place, and the victim in second place, the best thing for the attacker to better his own position was actually to attack the victim. And as the best option for me to maintain my position was to attack the person who threatened me the most, I should also attack the victim. The game was going to be about second and third place, no longer about first place, and the victim was going to have to deal with all his opponents’ attacks.

As the game was not turning into a direction which I liked, I decided for myself just to go for “as many points as possible”, regardless of who I had to attack to get them. Furthermore, I encouraged both other players to try to level the playing field by doing something similar, or if need be, attack me.

But what about the etiquette? Evidently, the behaviour that I prefer is not appropriate in all situations. It may be appropriate in the final gaming round (and frankly, usually that is the moment that kingmaking becomes a problem), but it probably is not during the part of the game that leads up to it.

What should good gaming etiquette achieve? In my view, it should achieve that the final ranking is according to the players’ relative strengths during that particular game. This is what I would call “fair”. Note that that does not mean that always the same player should win: many games have a chance factor, and a player’s strength might be that he was lucky. Or, the tactical situation might have been such that the player who is usually the best, was unable to deal with it well, and thus during that particular game he was not the best player.

And what rules should gaming etiquette prescribe to achieve such fairness? I have not yet formed a firm opinion about that. Not kingmaking is one of the rules, at least for the final gaming round. Trying for a personal best score is another one, but it is a shaky rule, as a personal best score is calculated relative to the other players’ performance, so focussing attacks on one particular opponent could be defended.

Anyway, such rules of good gaming etiquette do not preclude explicitly the behaviour of the attacker in my game of Small World, who focused his attacks on one particular opponent. What counts against his behaviour was mainly that he seemed to do it out of some form of personal spite, but in itself gaming etiquette is about actions, not about reasons for actions.

Still, I do not like this kind of behaviour, and I rather not play with people who show it. Although in this case I forgive it, as the attacker was a pretty young player. He still has time to learn.


An umpire’s nightmare

July 13, 2009

Four friends of mine became the Dutch boardgame champions. They scored the most points in four different games out of nineteen teams. One of them even proved to be the best player all around, with three wins and one second place. Because of their success, they are now invited to compete in the European Championship Boardgames, which are held in Essen in October.

The four games they have to play are Chicago Express, Diamonds Club, Power Grid, and Small World. I don’t know Chicago Express (but in general, I am not a big fan of Queens Games). I have never played Diamonds Club, but it was part of the Dutch boardgame championship, and my friends told me it is a very good game.

I think that Power Grid, which is one of my favorite games, is a truly excellent choice. It is a tough game, rather unforgiving, with very little chance involved, and lots and lots of player interaction. The game is already pretty seasoned, so there will be no unclarities in the rules.

With Small World, however, I predict there will be serious problems.

Small World is a modernized version of the game Vinci. It is a light war game, in which players invade a small map with armies. An army consists of between five and twenty units (but will usually have about ten units). Each army has a “race”, and a “special ability.” The game has fourteen races and twenty special abilities available. When a player has to select an army (which he does one or more times per game), he chooses between six random combinations of a race and a special ability. So, in one game there might be “flying sorcerors”, “swamp ghouls”, “merchant elves”, and “hill amazons,” while in another game the fight might be between “merchant giants”, “seafaring tritons”, “commando skeletons”, and “spirit elves”.

The random combinations make the game great fun. The key to winning is selecting the most powerful combination available. However, while some combinations are powerful per se, others might be powerful specifically against the combined powers of the opponents. Some combinations are only powerful if employed in exactly the right way. And some combinations might seem to be powerful, but are actually quite easy to defeat if the opponents find the correct approach.

The big problem of the game is the interaction between all the race powers and special abilities. The rules of the game, which seem clear from the outset, actually fail to answer many of the questions that crop up during gameplay. Often, the rules are unclear because they use terms that are not well-defined. For instance, a rule says that a race that has the “pillaging” special ability gains a point for each “non-empty region” that it conquers. But what makes a region “non-empty?” The rules don’t say. This question has already been posed to the publisher, and the answer is that a region is “non-empty” if it is occupied by an active or non-active unit, so that one is settled.

But questions like this crop up all the time. Last night I played the game and we raised a question that hinged on the definition of the word “you.” Is “you” the player, or the race that he plays? You would probably guess that it refers to the player, but the rule made much more sense if it would refer to the race. In almost all cases, it would not actually matter whether you refered to the player or the race, but in the particular situation we were in, it did.

Such issues arise because a game designer cannot acknowledge all possibilities of all combinations of all races and special abilities on one board. That is simply too big a search space for a human to grasp. It is the same with video games, which often have lots of rules, and designers are expected to foresee every tactic that a human player can employ with these rules. That is impossible in practice, and therefore, even after many man-years of playtesting, games still contain bugs and exploits in their rule sets.

Normally, in boardgames such issues are no big deal. Players simply agree on a house rule, and gaming can continue. But in serious competitions, there must be a set of undisputable rules that cover all possibilities, and all players must be aware of those rules.

Currently, competitors for the European Championship Boardgames are training to excell in Small World. They will be discussing rules interpretations. If they cannot decide on how a rule should be interpreted, they might consult the organizers, who will ask the Small World publisher for a judgement call. So from this exercise we might actually get a better set of rules.

However, I predict that many teams will interpret game situations differently, without asking the organizers because they feel that their interpretation is the most logical one. And thus disputes will arise during the actual Championship games. At that point, the rules will be interpreted and changed ad hoc, to the dismay of all the teams that are playing at that time. This will be chaos.

Small World is a really bad choice for a competition. Why was it chosen? Because it was sponsored by Days of Wonder, the publisher. And Days of Wonder sponsors it, because it want to promote this particular game. But the organizers should have refused. The fact that a game is nice in a friendly, recreational setting, does not necessarily make it suitable for competition play. And Small World definitely is unsuitable.

My friends now have to root out all the issues that they can possibly have with the Small World ruleset. I do not envy them. Playtesting should be done by the publisher, not by the players. Still, that is more or less practice with video games too.