It seems to me that games are built on choices. I wrote at length about making mistakes a while back (still one of my favorite articles), and how that affects game design, but I wanted to run another tangent today.
Candyland is billed as a game, but the only choice you have is whether or not to play (barring metagame choices, of course). Once that deck of movement cards is shuffled, the game is set in stone. There are no choices to make in the actual gameplay.
So, what if we make some?
How about a simple one to start with?
Instead of taking the top movement card and obeying its prescription, you take the top two and choose which one to use, and the other is ignored and discarded. (Alternate: put the unused card on the top of the deck to add a layer of memory.)
This is still pretty rudimentary, but it does give the chance for players to look ahead and make an informed choice. It’s a meaningful choice as game designers sometimes ask for, because you can only choose one of multiple options, and the choice is irrevocable, but it’s an informed choice.
So, how about removing some information?
Draw two cards and choose one without looking at either. Move accordingly, then discard both cards.
Now it’s getting interesting. It’s still a meaningful choice, but now it’s an uninformed one. It’s still enough to make the game unpredictable because of player choice, and it gives a veneer of player autonomy… but it’s still largely random. This isn’t much better than the core “game”, but the act of choosing at least starts to feel a bit more like the players have control.
So, maybe add a little Monty Hall flavor?
Let’s Make a Choice. Draw three cards, reveal one, then choose one of the three cards as your move for the turn. (Alternate: Have the other player take Monty Hall’s position as arbiter of the cards and really play this parallel to the hilt.)
Well, well. Now we’re digging into a classic game perception paradox, and really making choices matter. This is a semi-informed choice, with a bit of “playing the odds” for spice.
Layering some complexity on top of the bare bones of the Candyland game gives a lot more potential for choices to be made. Increasing the complexity doesn’t always help, of course, since giving players the choice of four or five cards with two rounds of choices is technically more complex, but in practice, it’s not really going to add much to the game. The initial addition of choice to Candyland scheme has a much stronger effect on the game than simply pushing that implementation deeper for the sake of complexity. The diminishing returns of that sort of increased complexity is something to be aware and beware of.
Alternatively, or in addition to any of these, one could splice in some chaos, and shuffle the deck after each turn. This wouldn’t have a huge effect on the actual choices as individual events, but it would make the underlying game potential more chaotic. The game state is no longer decided and set in stone at the game’s start, it’s in flux. As far as any individual semi-informed choice is concerned, that flux is largely irrelevant (unless you start putting non-chosen cards back in the deck to be shuffled instead of discarding them), but the game on the whole has more going on “under the hood”. That bit of churn adds ever so slightly to the game. (Though probably not enough to justify actually taking the time and effort to shuffle that much. The principle is more useful in digital games where “shuffling” is very low cost, relatively speaking.)
Such uncertainty imposed by randomization is a huge part of most card games, games that use dice, and even most computer games where there is a RNG under the hood fudging the predictability. That’s usually a good thing, as randomness brings the potential for even more choices to a game, if harnessed properly. When you have to constantly shift tactics and strategy in a game, it changes the choices you make. Sometimes that’s desirable, sometimes it isn’t, but most games incorporate some sort of randomization.
Of course, randomness has to be bounded somewhat (another old favorite article), lest the design get completely out of hand. Complete randomness makes choices all but useless, as a completely uninformed choice may as well not be a choice at all. Without at least a vague sense of predictability and consequence, there’s not much to a choice, and not much to be learned. Again, too much chaos pushes a game design into useless flailing.
Even too many choices can be paralyzing. As useful as choice is to making a game a useful and fun tool of experimentation and learning, too many choices can paralyze or confuse players. Too much intricacy and interconnectedness between choices can also cause trouble as players don’t really take the time to understand their own choices or don’t have sufficient feedback to understand what their choices mean.
Certainly there is room for complexity and chaos, but they must be wielded carefully. Choice is, in my mind, a backbone of gaming, but it, too, can be used ineffectively or unhelpfully (and it may not even really be choice a lot of the time). A little of all of these is perhaps necessary for a really great game, but finding the right mix is what makes game design an art… one that I appreciate and enjoy as both a gamer and a game designer.
It’s not an art that I’ve mastered, but I am learning to appreciate agency, psychology and creativity the more I dig into these things. That’s part of why I believe games have a lot to offer… if they can manage to be more than exercises in foregone conclusions, railroading players or overly random gibberish.
Tish Tosh Tesh 
I wonder if we have not reached state of over-design for many DIKU inspired MMOs, adding more complexity and new ideas but diluting the core design.
One could also debate if Tank-Heal-DPS for the dungeon finder in WoW and interchangeability of various player classes within the trinity system was a step back to the origins or taking away specials of various classes. Oh well, not going to think about that now.
I have an example of a bad minigame where you have the choice, but where it does not matter at all. Reminds me of MMO dialogue where you have two choices to answer, but the same thing happens nevertheless.
Star Trek Online’s “Dabo” is a lot like Roulette. But it does not matter if I bet on three places/symbols/numbers/icons, whatever we want to call them, or one only (in fact it is always better to bet on 3).
Because there is no way to reduce, increase or spread the risk like in Roulette by just betting on red/black, odd or even, a combo of both or in numbers.
Then… ah well, let’s stop here. I asked on the forums if such a “game” can be called a “game” at all. One somewhat interesting but also embarrassing answer was that it is a “mini game”… 😉
The biggest challenge in game design is indeed balance, finding the middle way between complexity and intuition, randomness/chaos and predictability. i’ve often asked myself how i would balance these issues in wow and also, how randomness and challenge are connected – if they are connected at all.
it’s an intriguing topic and i believe that the answer is probably found by looking at real life models, what people consider ‘challenging and fun’ or ‘complex yet motivating’ about a system or game.
It’s not a good example, because Candyland is designed for very young children, and doesn’t require reading skills. Adding choices then wouldn’t help the target audience. I think the point is more to get them used to colors and the idea of a journey to the end.
Adding choice in your example is sort of meh too. In Candyland, and in other boardgames, there is never any point to pick the worst possible choice from three cards. The only way there is if you let another player pick for you, and either that’s just another form of randomness (face down) or picking the worst possible choice (face up.)
Choice only matters when there are multiple choices with no good or bad, even slightly. For example, MMO classes only work when there really isn’t any advantage to picking one class or another in a set. Tanks play different, but achieve the role the same. When there is difference, when one tank is a 6 on a d6 and one is a 2, that’s when choice doesn’t help.
Longasc, pfft, “minigames” are games, too, and should be designed as such. 🙂
You’re probably right on the DIKU mold, though. There’s only so much it can do, and overdesigning is one way to try to carve out a niche. I’d rather challenge basic assumptions and design choices, m’self, but that doesn’t seem to be the common approach.
Syl, there are indeed a lot of factors to balance. That’s why I write on occasion about understanding the basics, because if we can’t get those right, all the layers on top won’t solve structural problems. I do tend to believe that the middle ground tends to be the best place to play, and in letting players meander a bit according to tastes.
Dblade, did you happen to check out the video link about the difference between choices and problems? There are indeed crucial differences between choices and problems, and to nitpick, I’m really just presenting problems here for Candyland players.
They aren’t high end gaming, no, but they are more than merely following directions, they are about looking at the board and understanding what those colors mean and choosing the best move. That’s not a huge logical leap to jaded veteran grown ups, but it’s certainly age appropriate to the audience. The Monty Hall variation pushes a little further into probability estimation and even reading other players if you play the parallel to the hilt.
You do offer a further step for the game’s evolution, though; make a good reason to sometimes pick the “wrong” color. Maybe if you land on the other player, they go back or something, occasionally balancing the desire to get ahead with the desire to put the other guy back. I didn’t bother digging into that layer of basic mechanics; changing how the player moves actually work. It’s definitely another good direction to explore.
I think it’s worth, before all this, deciding whether your designing A: Skill challenge game, or B: Exploration game.
I went and watched the link.
Here’s my gameplay variant.
You get a card. You have a choice. You can either keep the card and not move at all, or move with it. If you get so many of a certain color, you win the game, no matter where you are. So to choose the worst possible card, you add different win conditions. That way a player will choose between going for 10 red cards, or moving to the exit, or other win conditions.
You could balance it many ways, by keeping the red cards hidden and letting movement people use extra cards but be forced to put them in a discard where completionists can use them.
I think that equal but different win conditions would satisfy what you mean by choice. If you mean just problems, I don’t see much difference. Movement is movement, and people will always choose the quickest.
Yeah, I recommended the video because it underlines what you’re arguing. You’re right, there’s a difference between the two. Your variant would be another good way to go; alternate win conditions are a good idea and a reason to choose “nonoptimal” cards. The tension between a good move now and a piece of the puzzle for a huge move later scratches that “choice” itch.
The variants I suggested are all simple problems in his parlance, but they are more than the game initially has. Problem solving is something that kids need to learn. Yes, someone who has figured the trick out and learned to read the board will always take the best move. Learning how to do that is something kids can still learn.
The Monty Hall version requires a bit more thought and guessing about probabilities, though. It’s not as good as an alternate win condition, but it’s a stronger bit of problem solving than the other options. That’s sort of the point; there’s a spectrum of options to play with here, and it’s good to know the differences between options.
[…] Candyland, Choice, Complexity and Chaos […]