This week’s dimension of gaming is Downtime. We’ll define downtime simply as the portion of the game for which you’re not actively playing. Downtime can happen during other players’ turns, during setup or cleanup, or during intermediate steps of a game like dealing cards for a new round. This dimension of gaming exists at the intersection of Game Length and Player Count: longer games increase the risk of downtime because as the length of a game increases, the length of time a given player is not doing anything increases. And as games are scaled up to include more players, designers need to keep in mind not only how the mechanics balance but also how adding more players affects the player experience, especially as it relates to taking a longer time before any individual player gets to take his next action.
Downtime is a great opportunity to discuss the paradox of choice, which plays out in the context of game design as analysis paralysis.
The Paradox of Choice
When presented with multiple options, game theory suggests that a person, as a rational actor, will make the optimal choice. Given the choice between a hot dog and a hamburger at a cookout, a person who really loves hamburgers will pick the hamburger almost every time. With more options, it seems reasonable that a decision-maker could be even happier: suppose someone else at the cookout leaned slightly toward the hot dog camp but wasn’t crazy about either option. The addition of grilled chicken as a third choice would allow that person to optimize the decision he had to make.
Now let’s transform that cookout from a reasonable one, with a few choices, to the world’s biggest cookout. In addition to hot dogs, hamburgers, and chicken, now we’re throwing everything we can think of onto the grill. Steak, corn, tuna, mushrooms, kabobs, peppers, even pineapple. And for each one, you can pick from plain, dry seasoned, or marinated. That’s suddenly many, many more options to pick from. Even the guy who really loves hamburgers might have a hard time deciding what he wants to eat. Maybe this sounds unusual but we observe this situation in games as well.
But wait! If a rational actor has more choices, he’s better equipped to make an optimal decision, right? In principle, yes. In reality, it’s going to be much more difficult to make that optimal decision with so many false and suboptimal choices in the mix. Pressure to make a decision and conflation from all of the different factors may even cause the person to make a “wrong” choice.
That’s the essence of the paradox of choice: even if the decision arrived at is objectively better, actors presented with more choices tend to be less happy with their decision. Most of the underlying explanation for the paradox of choice involves the evaluation of opportunity cost. Eating a grilled steak probably means I can’t also eat a bratwurst, because I’d be too full for both. Even if a steak is the best choice for me, I may be so caught up on what the brat might have tasted like that I fail to enjoy my steak. Finally, with so many options on the table (or the grill), a person might have an unreasonably high expectation for their choice: if I really decided against steak and sausage and tuna for these mushrooms, they had better be the best mushrooms I have ever tasted.
One of the most adverse effects of the paradox of choice in terms of game design is analysis paralysis, familiar to players and designers alike. The paradox of choice leads seamlessly into analysis paralysis because more choices–and more internal doubt about whether your decision was the best one–leads to players taking more time to analyze each and every one of those choices.
Paradox of Choice in Game Design:
The more choices a player faces at one time the less content they might become with their choice. A few reasons why players might be less satisfied with a game simply due to the sheer quantity of choices they face on their turn:
Increased expectations (“I’ve got a fist full of cards so there must be one that works best for my strategy.”)
- We can accept that our single tile in Carcassonne may not always be adequate but if we had to consider a dozen tiles our mindset would quickly change.
Additional effort and psychological stress (“Let me double check my cards to make sure I didn’t miss a better option.”)
- Evaluating multiple options leads to a heightened “weight” associated with the decision.
Increased opportunity cost (“This card is likely better if this game lasts three more turns, but it could end next turn…”)
- The way we value things depends on what we can compare them to in games. Even when we make a choice it is easier to imagine the attractive features of rejected options (the features we did not choose). This leads to…
Increased buyer’s remorse (“Now that Josh is taking his turn I have twenty minutes to sit here and think about what I’ve done. Maybe the other card would have been better…”)
- With so many alternatives it is natural to consider how other choices might have been better even after we’ve finalized our decision.
Lastly, in light of all of the above factors, we are more likely to blame ourselves when our decisions don’t meet our expectations. When we reflect upon our decisions later we may not consider all the variables we had available at the time. (“I had all these options, I clearly messed up, I should have picked better” vs. “I only had three options and I did the best with what I had”).
Each of the above factors can contribute to downtime but it will require the presence of one more thing for turns to come to a screeching halt. If you’ve been waiting on us to finish this turn, that final ingredient is…
Even if you don’t know it by name, every single person who has ever played a strategy game has seen it: the guy just before you in the turn order contemplates the board for an eternity before finally making a move. He’s in the tenacious grip of analysis paralysis. It’s likely that the game has offered him so many choices that he can’t possibly hope to pick out the right one, or he’s terrified of making the wrong pick because of how badly it might doom him in future turns, or he would be able to make a choice but has to do some banal number-crunching first, or the guy before him in the turn order made such an unexpected move that it threw off his strategy. Maybe it’s some terrible combination of all of the above.
Analysis paralysis is never a good thing. It’s not good to be the player who’s agonizing from it because you stand the chance of ten minutes of brain-burning and still getting it wrong. It’s even less good to watch someone else suffer through it because it’s the downest of downtime… and you know very well it could be you next. Analysis paralysis can arise from a whole host of underlying issues, but we’ll boil them down into three in particular.
Players new to a particular game might have no idea how to play it.
While every game of course comes with some learning curve, certain games are much better than others at showing new players how they should go about playing the game. The first time you’re dealt six cards at the beginning of Race for the Galaxy, how in the world are you supposed to guess which four of those might be cobbled together into a winning strategy? In April and May, we wrote an article series on the topic of approachability, the tendency of a game to remove barriers to entry for new players. Because we discussed approachability and the “new player problem” in so much detail earlier in the year, I’ll only briefly summarize new players as they relate to downtime here.
In general, if you can establish a clear purpose (win the game by controlling the most territories at the end of turn seven), make sure there’s a sense of clarity relating to the cause and effect of the players’ decisions (conquering territories gives you resources, which let you build more armies, which in turn let you conquer more territories), and assure your players that they’ll be able to get where you tell them they’re supposed to go (close-to-optimal play results in the ability to conquer at least a territory or two in the first turn), then you’re in great position to solve the new player problem.
Certain players are overanalysts.
In strategy games, much like in the NFL or in any competitive endeavor, you play to win the game. And as Alex talked about in one of our very first articles, both the result of winning and the process of building a good strategy can motivate people to play games. It’s a different story when you’re watching someone build a winning strategy; probably nobody is motivated to play games because they like watching other people move wood around on cardboard. But there will always be players who are so motivated by achievement that they will put their worker casually on the “gather ore” spot, look at it for a few seconds, grab it back, (to their credit) apologize, and twirl it idly for the next five minutes while sighing and muttering things about “value” and “percent.”
Aside from discouraging this sort of player to play your game–and especially if you’re designing a “gamer’s game” for the “hardcore gamer” audience, that might not be the wisest approach–there are plenty of design choices to mitigate the innate tendency of the most competitive and perfectionist gamers to play exactly optimally. Plenty of the options we’ll look at later on can help to curb this problem but unfortunately there isn’t a panacea here; some players are just prone to analysis paralysis in games.
Some games are especially paralytic.
More complexity might be a good thing or a bad thing, depending on the gaming experience that you’re trying to create. But more complexity is almost always associated with more (and/or tougher) decisions to be made, which correlates strongly with downtime. By throwing a multitude of decisions at a player, even experienced players or those not normally ones to ponder whether to draw one card or two might find themselves at a loss. Games with role selection or worker placement can be especially tough in this regard because players will naturally want to weigh opportunity costs very carefully. This is one of the primary focuses of this article as it seems to be in the hands of a designer and on the minds of players but difficult to avoid and frustrating to experience.
Causes and Composition of Downtime
Let’s discuss a hypothetical and generic (but not very good) game, Strategic Kingdom, that lasts 60 minutes, plays with 4 players, and is of average complexity. As a player, you’d expect to be “active” for 10 to 20 of those minutes. And let’s say a typical game lasts 10 turns. You’re “playing” (making decisions, physically performing actions, interacting with the components) for 1 to 2 minutes at a time and observing for 3 to 6 minutes. That’s not too bad.
The most mathematically obvious ways to increase downtime in Strategic Kingdom (remember, it’s not a very good game) are to increase the player number and to increase the game length. If we increase the number of players and keep the total game length and time per turn (a proxy for complexity) constant, the number of turns per player necessarily needs to decrease. By upping the player count to 6, now your turns are lasting for maybe 45 to 90 seconds. The downtime between turns has increased to about 4 to 7 minutes. Now players are noticing downtime and it may even be a common complaint.
Alternatively, we can increase the game length to accommodate the increased number of players. A 90-minute version of Strategic Kingdom with the same number of turns per game and length per turn now features off-turn downtime of 5 to 10 minutes per round, very likely totaling an hour over the course of the game.
Although we’re considering the Downtime dimension to exist at the intersection of Game Length and Player Number and Scaling, Complexity is a hidden confounding variable that subtly affects Downtime as well. If we increase the complexity and double the turn length (but keep the overall game length fixed), now you’re taking 5 turns that last 2 to 4 minutes over the course of the game. So is everyone else. Your average downtime between turns is now somewhere between 6 and 12 minutes. One of the important results of this analysis is that not only actual downtime but also perceived downtime matters: 45 minutes of downtime divided into several short stints is much more bearable than the same time in half as many longer blocks.
Fluctuating Levels of Complexity
Why does increasing complexity increase the turn length? Part of the answer is that you have more decisions to make, as I mentioned earlier. In both Lewis & Clark and Concordia, the initial hand contains only a few cards with relatively straightforward mechanics. Decisions actually become easier in turns 2 through 5 or 6 as a hand decreases in size, and players have many fewer options. But over the course of both games, players increase their hand size through both acquiring new cards and recovering their played cards. By the end of either game, it might become necessary to make a decision among one of a dozen different cards to play–and how to optimally sequence the remaining eleven.
An even more dramatic example is Alien Frontiers. Players start with a few dice in Alien Frontiers, which generally limits the possible uses of actions on the board. Therefore, early turns are relatively simple and quick. As players inevitably increase the number of dice available to them (adding a fourth, fifth, and sixth die), there is suddenly and exponentially more complexity: adding a fourth die doesn’t just add one more possibility, it adds at least four, because of how the new options may interact with the first three dice. Furthermore, cards come out as the game progresses, making every decision more complex because the dice can be manipulated and straightforward turns are less likely to occur. Finally, unlike in Concordia or Lewis & Clark, players can’t plan ahead because they roll at the beginning of the turn and must decide their course of action then.
Probably a more important part of the influence of complexity on turn length (and therefore downtime) is that the other players also have more decisions, which can substantially alter the game state before you have the chance to take another turn. Simply increasing the player number can have the same effect. In a two-player game of Castles of Burgundy, your opponent can take as many as three (but much more often only one or two) tiles from the board and prevent you from taking those tiles. There’s a chance that actions could derail your own plans, of course. But that chance is much greater in a four-player game, where more than ten tiles can disappear before you have any idea what’s going on. We’ll return to Castles of Burgundy later to see how it mitigates this problem.
Aside from complexity, one additional design component that can create downtime is the presence of “cleanup” or intermediate scoring phases. If watching another player take his turn is an uninteresting break in the action, then watching another player count up his points is positively dreadful. One of the few games that sticks out as having an obtrusive cleanup phase is Bora Bora, in which the tiles on the board are swept clean in exchange for new tiles after each of six rounds. Ideally players will total their scores simultaneously during these interim periods or divvy up the responsibility to replenish tiles but these are periods of downtime are worth reconsidering if possible.
Other than the sheer quantity of available choices, there is another catalyst of analysis paralysis which is the necessity of planning ahead in games. This need to plan ahead can be driven by the structure or objective of the game or by a specific mechanic. These ideas are typically thrown into the giant vat called strategy, but the need to speculate future events or plan many turns in advance are designed principles that encourage downtime.
The most clear example is probably Chess, a game admired for provoking thought since successful play is so closely associated with a player’s ability to plan many turns in advance. The possible permutations of how a game can play out can be seemingly endless in Chess to the point that competitive environments like tournaments often mandate a timer.
One game that has gotten a lot of attention recently, Five Tribes is a fascinating example of a game that encourages downtime in nearly every way imaginable. A successful turn in Five Tribes is akin to solving a puzzle by identifying and performing the best scoring opportunity. At the beginning of each round players bid sequentially using victory points to determine turn order. To properly value a bid on turn order, it would appear that players need to identify several of the highest-scoring moves and bid accordingly to maximize total score. Once turn order is set the first player takes the best move they were able to find. But wait, as a result of this first move the board is now changed for the second player, potentially opening up a new lucrative action. It isn’t uncommon for the third or fourth player in turn order to find a board so different that they need to spend a few moments to reassess the best possible actions.
Five Tribes is deliberately designed not to be a game where every turn can be planned perfectly ahead of time. Planning your turn during another player’s turn is one of the most basic means of mitigating downtime. The downtime in Five Tribes can be excruciating, if you’re the type of player who absolutely needs to have an optimal turn planned before you take it, and you realize there is no way to accomplish that task. (On the other hand, the downtime here can be liberating, once you realize that you actually cannot use it to further your position in the game, and you start to use it for pure socializing with other players.)
The redeeming quality of the Mancala-like mechanic in Five Tribes is that it is an end-point driven decision which shaves down the necessary downtime considerably. Players can find a congregation of meeples in a color they like and work backward to find a starting point nearby. Consider for a moment if players had to do use a more traditional trial-and-error approach; individually look at each location and play out the various possibilities until they find best scoring option before testing the next location the same way. It sounds silly but quite a few games are structured to reward this comprehensive method of planning a turn, most notably our earlier example of Chess.
Solutions for Downtime
Perhaps the “holy grail” of mitigating downtime is to remove it altogether. The method of simultaneous actions has been used in a variety of games, and the principle is simple: if every player takes his turn or action at once, then nobody is left in the cold waiting for the turn order to come around to him. The simultaneous action idea is one of the (borderline) defining concepts of 7 Wonders, which along with Sushi Go!, is probably the most salient example of a simultaneous-action game that has been released recently. 7 Wonders plays out in a three-round draft, and the entirety of the decision-making in the game happens while selecting cards. The reason the game works so well is that every player is making the choice about which card to draft at the same time, and the flow of the game is naturally limited by all of the players making their draft selections.
Simultaneous actions are also possible in role-selection games like Puerto Rico, Race for the Galaxy, and San Juan. In these games, the player who selected the role gets a small bonus when the corresponding phase occurs, but every player who wants to performs the role at the same time. Race for the Galaxy in particular is a relatively fast game with little downtime despite its complexity: although there are five possible phases and countless possible card/phase interactions, only a few of those phases are played on a given turn, and each player simultaneously decides whether or not to participate in a given phase. Robo Rally features a similar mechanic where players “program” their turn simultaneously (and therefore make decisions at the same time), then they execute their programs simultaneously.
Of course, simultaneous actions aren’t perfect. In theory, the draft selections in 7 Wonders or the role selections in Race for the Galaxy all occur at the same time. In practice, it’s not an ideal solution because the lengths of time it takes players to make their picks will follow some distribution rather than all being the same. And simultaneous actions aren’t really feasible for certain styles of games: players need to know exactly what cards are available for purchase in deck builders or exactly what land is ripe for conquest in territory control games, so it’s important that these kinds of games play out sequentially rather than simultaneously.
In situations where simultaneous actions aren’t really appropriate, a good alternative is to use at a mechanic not tied to player turns where everyone is engaged. The auction in Power Grid (among many other games) and the silent reverse auction in Hotel Samoa bring everyone’s focus to one event and one mechanic while at the same time preventing what’s sometimes derisively referred to as “multiplayer solitaire.”
An even more popular mechanic is to use either off-turn interactions or off-turn triggers to keep players interested during what would otherwise be downtime. An off-turn interaction is a means of affecting both your and someone else’s turn. Great examples include the trading mechanics in Bohnanza and Settlers of Catan. Although it’s against the rules to propose a trade during someone else’s turn, the active player can propose trades with anyone else, so it’s in everyone’s best interest to constantly assess the situation in case a trade opportunity arises.
Similarly, off-turn triggers are events that arise during another turn that provide some small benefit or alteration to your game state. The Fate mechanics in Ascension allow you to manipulate your deck when a card appears on the table, regardless of whose turn it is. When another player builds next to you in Terra Mystica, you can choose whether to sacrifice victory points to gain a resource called Power. And in Lords of Vegas, a card is pulled on every player’s turn and payouts occur. Lords of Vegas essentially pays out a fraction of everyone’s income as the turn order moves around the table, while many games simply have everyone take a turn after which everyone takes full income at the same time.
Other ways to reduce downtime are to subtly and situationally reduce complexity. Any hand limit in a card game is a restriction imposed on a player to limit the number of choices that are possible to make and therefore reduce the options the player is forced to consider. Carcassonne in some ways is a card game with a hand size of 1; a version of Carcassonne with more cards in-hand would be arguably strategically deeper but would drastically increase downtime because players would have so many more options to analyze. The somewhat abstract restrictions on placement in Kingdom Builder are the source of the strategic appeal of that game but also serve to limit a player’s possible moves and locations for expansion.
Designer Tom Lehmann discussed this idea in our recent interview:
“You need to eliminate open but mostly unimportant information that some players might stare at forever, halting play.”
One of the most effective solutions to reduce downtime is to simplify the amount of information visible to players, starting with the least important. This is the game design equivalent of “Out of sight, out of mind” and it is a favorable approach to assist players who have a tendency to search for importance in every object on the board.
In addition to making it easier to pursue a coherent strategy, the Worker mechanic in Castles of Burgundy helps to reduce turn length (and therefore downtime for other players) by “funneling” possible die rolls into a smaller set of actions that might benefit a given player. Part of the difficulty in Castles of Burgundy is the inability to plan a turn ahead of time given the dice rolled at the beginning of a turn. To truly cover all of the bases, a player would need to come up with twenty-one different contingencies for what would happen with any given die rolls, which could in turn be altered according to the other players’ actions. Workers, however, allow the die rolls to be modified, so a player might need to consider many fewer options: “if I roll a 4, I take the mine; if I roll a 3 or 5, I burn a worker and take the mine.” In other words, you can plan ahead during other turns because you don’t need to know your precise rolls ahead of time, as long as there’s a feasible path to ending up with the number you really need.
Finally, downtime can be reduced and/or made more tolerable with metagame design, or elements of the game that aren’t coded into the game’s rules. Half the fun of territory control games is persuading other players not to carve up your territory: don’t kill my dudes from my empire in decline in Small World; leech that other guy’s mega-city in Carcassonne; I’m not even the biggest threat so don’t target my last card in Coup. The unwritten social component of territory-control games makes other players’ turns even more tense than your own because you have so much incentive for those other players not to tear down what you’ve worked so hard to build.
It is sometimes possible to maintain players’ engagement or interest outside of their turns by having them continue to focus on their own turn. For example, players may be able to perform clean-up or scoring for their turn during the next player’s turn as long as it doesn’t affect the game state. In Agricola, there’s no reason why a player who just built six fence lengths couldn’t figure out how to configure them during the next player’s turn. And if all else fails, players can be given “bookkeeping” roles: one player distributes income, one totals taxes, one clears antiquated cards, one adds up the points at the end of the round, and so on. (The Cones of Dunshire has a ledgerman who not only does this sort of thing full time but also gets to wear a hat while doing it.)
In this article, we’ve talked about several major sources of downtime. Downtime can be inherent in the design of a game, as it arises from the intersection of game length and player count. Or, it can emerge from the players’ underlying psychological characteristics in the form of the paradox of choice and analysis paralysis. Simply chopping up turns to make them shorter does not necessarily reduce downtime but can make it more palatable in the form of smaller doses.
Making action simultaneous appears to the the most effective way of eliminating downtime, though it only makes sense for games with drafting, role-selection, or similar mechanics, and it may inadvertently lead to “multiplayer solitaire” situations. Auctions, trades, and resource gains are other great ways to maintain player involvement. Finally, don’t underestimate the role of social and metagame solutions: players won’t be complaining about downtime if they’re having fun or keeping busy during it.
One final thought is downtime can’t be eliminated but can be reduced drastically by the order of operations in a turn. Pay particularly close attention that you don’t put too much information in front of players at any given time. This is a hidden value of hand limits where players can’t just keep drawing cards and kick the can down the road until the eventual decision can be overwhelming. Unless you’ve got a great reason, the difference between allowing players to draw a card at the end of their turn versus the beginning of their turn can be entirely the difference in downtime. Even players who take turns quickly need a moment or too to digest new information if received at the beginning of their turn. By changing the order of operations of a turn so that players only accumulate new information at the end of their turn (versus at the beginning or in the middle of taking various actions) you can avoid creating unintended periods of downtime.