Disorderly Conduct

A fundamental question you must answer for your design is "what will a player do on his or her turn?" In most cases, players will have individual turns. Yes, there's simultaneous play, like in Wonders, or there's real time play, like in Paper Route, but generally most games have player turns.

There are typically three options to give a player on his or her turn:

  1. Give the player one Action. Ascending Empires does this beautifully.
  2. Give the player several Actions. These are done in a strict order. Dominion is a great example of this (Action, Buy, Clean).
  3. Give the player several Actions. These are done in any order. Alien FrontiersFarmageddon, and Empire belong here.

Let's go through every option. There are good things and bad things about each. Here's a bit of foreshadowing: I prefer option 3.


Give the player one Action: This is the cleanest of the three options. Essentially, you give the player a variety of choices, but they can only choose one. Ascending Empires is one of my favorite games and it does this perfectly.

You can build, move ships, research, recruit men, or mine. For many of these, you must meet the conditions before you can choose the Action, which further simplifies your decision. You could also argue Wonders does this as you select one card out of a fairly large hand.

One Action has a few VERY strong arguments in its favor. Firstly, the pacing of a game with this structure is typically fantastic. Taking only one Action means you can't have any combinations (two is required), which greatly simplifies the decision and reduces analysis paralysis. Games with one Action are also simpler and more straightforward. They ask "what one thing do you want to do most this turn?"

Games that use this can also be presented very easily. You can use a tableau that shows the five options (Ascending Empires, Glory to Rome). Or, like Wonders, you tell the player "Pick one card" and they can just look to the cards.

There are some downsides to this mechanic. If not properly tuned, a game where you take one action at a time can feel plodding and tedious. You need to make actions decisive and "big" enough that the player doesn't feel like he's trying to drain the Titanic with a bucket. The one action should feel like a step, not a baby's crawl.

The other issue is that by having one Action, you remove the ability for combinations. Yes, this is a plus, but it's also a minus (depending on your point of view). In my experience, combos, even simple combos, are one of the best ways to let a player feel clever. Combos give players that "ah ha!" moment and with one Action you'll need to deliver that elsewhere.


Give the player several, strictly ordered Actions: This is arguably one of the most common ways to let a player take his or her turn, and for good reason. As a designer, you can present the player with several options and far more flexibility, yet still give the player a comfortable and rigid structure.

One of the first things I learned in design is that players want to be told what to do. They find it comforting. No, I'm not saying people are sheep. But, nothing is more overwhelming than being told "do all of these things however you want!" Gah! Instead, you're saying "Do all of these things, but step by step."

Dominion is an absurdly popular game. A big reason for this is how accessible the game is. Yes, learning 10 new cards every game isn't easy. But, knowing that on every turn you're going to play an Action, Buy a card, and clean up your hand eases things along. ABC. ABC. ABC.

Designers should be careful with the number of steps. Yes, you can (and should) have a reference card that walks you through the steps. However, I've played some very simple games that had 6 steps on every player turn and, even though the order was strict, it was overwhelming.

Multiple actions with a strict order give the player a broad turn, let the game move along quickly, and do so within order and with a reasonable pace. Analysis paralysis isn't too bad because players know they must do step 1, step 2, and step 3. It's a routine and people like a routine. But, the downside is that combos may be more rigid and you limit the option for player creativity.

If you have a complex game and you're trying to figure out how to simplify it, try testing the prototype with a strictly ordered turn system.


Give the player several Actions with no order: Be warned! Use this only at your peril, for it is fraught with all sorts of nonsense you must resolve.

With this system, you give the player an array of options and tell them that they can not only do many of them (if not all), but they can do them in any order, and sometimes multiple times. This immediately sets you back:

  1. Your game is now more prone to analysis paralysis players.
  2. Your game is more overwhelming and less accessible.

You will need to understand and accept this. It also means that you'll need to scale back your complexity elsewhere and reduce other opportunities for players to stall and over-analyze.

One of the hardest things for players to grasp with Farmageddon is that they can do things in any order. Every demo I have ever given is immediately broken up by a "how much can I plant?" or "When can I play Actions?" or "Can I plant again?" The game would be FAR simpler if I said this was the order:

  1. You must harvest if able.
  2. Plant (optional)
  3. Play Action card 1 (optional)
  4. Play Action card 2 (optional)
  5. Fertilize at least 1 time.

However, the game would be far less fun. Farmageddon isn't a terribly deep game. It's not a brain burner. But, over and over again I've seen the joy in people's faces when they see they can plant, insure, and destroy a crop, then plant again to take up the field. Or, they can plant, steal a card, Bumper another crop, harvest, then not fertilize because they burned all their cards.

I made the choice that flexibility would vastly improve the game even at the expense of accessibility. It was a compromise I made and had to work for, but one I've never regretted.

A very similar game to Farmageddon is Gamewright's The Big Fat Tomato Game. This game does many things like Farmageddon, but you must do everything in order. The game comes with reference cards to walk you through the 6 steps. It's not a difficult or complex game, but my friends and I had to use the reference cards for the entire game. If I brought it out again, I'd still have to use the cards. Sometimes the confusion is just shifted to another place!

Order of operations allows for exciting combos and player creativity. If you have a game with cards, you'll see players coming up with new combos that you hadn't conceived. This has been the best part of developing Farmageddon and Empire for me. I personally love playing Alien Frontiers because every dice roll is a "holy crap look at all this cool stuff I can do!" moment.

Really, it all comes down to the game you're trying to deliver. It is greatly determined by your target audience. If pace and simplicity are your priorities? Use One Action. If you need a bit more flexibility but want to hold players' hands? Multiple structured Actions. If you want to enter the wild west and you're ready to scale back elsewhere to let it happen? Give multiple action unstructured a chance.

Contribute to the discussion! Do you have other examples? Other pros and cons I failed to mention? Use the comments below.


Excellent break-down as usual, Grant.

I would add that all of this can be stratified (and perfectly, in many cases) to help a designer see how much decision complexity her game entails. We can do that by enumerating every combination of choices the players can make on their turn.

If a player takes just one action and has five actions to choose from, she has exactly five choices. This is a linear scale (1 * numChoices).

If a player can choose A or B then 1, 2 or 3, then I, II, III or IV in that order, she has 24 choices. This is a multiplying scale (numChoices1 * numChoices2 * numChoicesN).

If a player can choose between three actions three times, in any combination or order, she has 27 choices. This is an exponential scale (numChoices ^ numActions).

Where this becomes muddied is when the availability or reasonability of an action is conditional. Using Dominion as an example, whether you play one or more actions is entirely dependent upon drawing one or more action cards in your five-card hand and further modified by the possibility of those action cards drawing you more and giving you the extra actions with which to play them. There are 17 cards you can theoretically buy during the buy phase, but 1 you never want, 2 you rarely want and the rest are all restricted by the number of coins you have to spend that turn. If you've got 6 coins and 1 buy, you're really only going to consider the cards that cost 5 or 6, narrowing your choices down considerably. With two buys, you'll likely spend 8 coins to buy a single 7 or 8, or else a 3 & 5 or two 4s. The clean-up step doesn't involve any choices, so you're really just multiplying your Action choices with your Buy choices. Many of your actions will be directed toward improving your buys, and bigger buys are almost always better, further simplifying what looks like a very large number of choices on paper.

I think this is a good start for what could actually be a 2-part post. I'd like to see your take on simultaneous actions and what role public vs. private information plays in how many actions to allow players.

I hadn't thought of that! I'll write that one tonight! Good idea.

An early version of my current work-in-progress suffered from very slow game pace. It was kind of a cross between Options 1 and 3, because each round, each player would have an opportunity to take any one of eight possible actions. The opportunity to take one action passed around the table clockwise, and when it came back to you, you could again take any one of those eight actions. Most actions typically cost money, or could only be done once, or had some other constraint, so eventually everybody would pass. Eventually. But until you got to that point, every time a player's opportunity game up, he would re-evaluate all eight options. "Can't do that, can't do that, I could do that, I don't want to do this, ..." It would take forever to finish a round. Early playtests took over 4.5 hours to finish.

I completely restructured the turn order, so that every player would have one opportunity to take Action A, then everybody would have an opportunity to do Action B, etc. Decision-making became much simpler and more straightforward. A few game-play subtleties were lost with respect to clever re-ordering of actions (like timing an auction when everyone else was low on money), but overall the game didn't suffer. There is one phase that acts like your Option 2, where each player has an opportunity to do all "Ship operations" in order - load ships, then move ships, then build ships.

The result was that game duration was cut by more than half with very little loss of decision flexibility. So as you said, game pace can be seriously affected by the option you choose.