The best games give players a constant feeling of choice. I personally enjoy being confronted with options such that each one feels like an excellent decision, but involves sacrificing other equally good options. I like the feeling of mastery I get out of making a tough decision and seeing the rewards (and penalties) associated with that decision later in the game. It can also make the off turn more exciting, if the decisions that other players are making are more interesting and trying to derive what hidden information (e.g. the contents of their hand) can be inferred from them making particular decisions. Choice is great!

For me the game that manages this most effectively is Race of The Galaxy. In this game most costs are paid by discarding things from your hand, while that same hand is the option of things that you might develop over the course of the game. Taking anything will be at the cost of permanently closing off not just one, but a host of other options. Many of the options have effects that are related to your other choices, powerful technologies offer points for having collections of specific types of card. Military forces allow you to conquer particular kinds of planet for free, but are useless without them. Trade powers allow you to get more out of your goods, but you’ll need something else to produce the goods in the first place. The choices all feel important and are woven together in such a way as to influence each other, the result feels excellent to play.

I think that this touches on an important principle, which is that choices need to be partially but not fully comprehensible in order to be meaningful. If you want a player to choose whether to take a ‘flarg’ or a ‘bilge’ with no concept of what these things are they don’t really have a choice, sure one will turn out to be better than the other, but with no more information the player might as well flip a coin. You’re wasting everyone’s time asking for the choice, it’d have been better for the game to just assign them one at random. A total lack of understanding negates any choice.

On the other hand if you ask a player if they want “+1 point” or “+3 points” that’s just as bad (unless your game has some very unusual rules). Anyone would take the greater number of points, again rendering the choice meaningless, it’d have been quicker to have the event just provide three points and not insult your player’s intelligence. A total understanding negates any choice.

The understanding problem may be more nuanced than it appears. For instance, it’s clear that +3 points is more than +1 points – but how does “+1 point every time you score points from now on” compare to “10 points right now”? The answer is that it depends on a host of other factors: Will you score points more than ten more times? Can your opponents influence how many opportunities you’ll get once they see which choice you’ve made? Is there an advantage to having a higher score in the short term or does it only matter at the end of the game?

In some cases knowledge of the game will resolve the situation for you. What was an interesting choice on your first game, becomes a boring choice after the second or third replay, because as you know more about the distribution of cards in the game some of the questions are automatically answered which can make the decision trivial.

In other cases a knowledge of mathematics can undermine a choice. It might seem interesting to be offered a choice like “This level I can get a +2 to agility checks or the ability to reroll failed checks”, but depending on the system there’s often a mathematically superior choice.

These factors can combine, for instance in the equation above, if Y (The target number) is variable depending on other situations in the game then there might not be a simple solution – instead the superior choice depends upon the rest of the game. In this instance a solid game knowledge and grasp of mathematics might combine to make a choice obvious, but it remains meaningful for players who lack one or both of these things.

This creates a trap for game designers, as producing a game that provides meaningful decisions to players who’ve really enjoyed their game and understood it on a deep level might lead to producing a game that offers choices that seem meaningless and arbitrary for players who’ve don’t have that level of engagement.

I once had a series of lectures on artificial intelligence by a professor who’d plow through the material at breakneck speed, covering all sorts of high level concepts. Inevitably students would get left behind. When they expressed their confusion she’d stop and try to bring them up to speed, but the material she explained in detail seemed to be at random, rather than related to their problem. Sometimes she’d try to alleviate the confusion by going slowly through the simple stuff, binary addition and the like, other times she’d try to relate it to cutting edge research that was technically beyond the scope of the class. Confusion reigned. To this day I’m convinced that the problem wasn’t that she was stupid – I think she might have been a literal genius – to me it looked like she had no comprehension of the notion that someone might not instantaneously understand this stuff and wound up at an absolute loss for what to tell someone who didn’t get it first time. My point is that it’s very easy to ruin something for everyone by assuming that they’ll get something that’s not as obvious to them (as people who’ve played once) as it is to you (as someone who’s spent months working with the details of a design).

Some games handle this very well by passing off decisions that affect the mathematics to players. For instance I’ve played (far too much) Dominion. Each game starts with a collection of cards being selected and over the course of the game you choose which cards to buy with the intent of having the greatest number of victory points once sufficient card supplies have been depleted. If you were playing alone there is very often a mathematically superior approach, which will generate the most victory points in the shortest period of time, however once you add other players into the mix the effectiveness of your strategy varies depending on your opponent’s actions. Using a chapel to thin out your deck by destroying cards you don’t want becomes a less powerful strategy if your opponent is using thieves, pirates, swindlers or other options that depend on them pulling the right cards off the top of your deck. Building a combination that would work up to generating a huge number of VP each turn suddenly doesn’t work if your opponent is running some Village/Smithy/Workshop combo that will deplete the decks quickly and end the game prematurely. The mathematics are solvable, but the decisions (and thus the game) remain interesting because the actions of your opponent influence your choices.

It’s interesting to note that the game is probably still more or less solvable. Both players could engage in some minimax like reasoning and would wind up playing the same strategy of each other and making no meaningful decisions, however the complexity of the calculation is such that most players would be unable or unwilling to do this, allowing the game to remain interesting. Similarly, Chess has been solved, but that hasn’t stopped anyone having fun playing it.

This solution isn’t workable for cooperative games like the one I’m currently working on (Wizard Academy) as there is no opposition to provide such challenges. I suspect that this underlies the relatively common criticism that a lot of cooperative games are simply puzzles and once solved they can be discarded. So I’ve had to try to pull together some different elements to solve the problem.

The first is that the problem is obscured. Initially it’s unknown which 28 spells the players have to work with this game or which 10ish disasters they’ll have to face, but the combination of this information determines the optimum strategy. Rather than being able to derive an optimal strategy and apply it every game, the situation changes fluidly enough to require players to find solutions to the mess that they’ve found themselves in this time. It may well be that some or all of these situations do have a mathematically optimal approach, but as with the examples above the decision tree is deep enough that it doesn’t make the decisions feel trivial.

The second is that Wizard Academy is partly an experiment in overloading working memory limits (this is the sort of thing that happens when a psychologist is allowed to design games). In playtesting so far, it’s rare for all possibly solutions for even serious problems to be identified by the players, due to the limits of information that can be simultaneously processed. This serves a dual purpose of interfering with the process of trivialising decisions and making the game cooperative in a purer sense. In a lot of cooperative games a single experienced player is able to identify the best moves and give instructions to the other players, reducing them to little more than waldos for moving pieces around. In Wizard Academy playtests so far I’m always pleased to see that even inexperienced players will sometimes produce a solution that the other players simply hadn’t noticed. That seems like a good experience for cooperative games to have.