Make a Better Game – Limit the Player

Okay, okay, I know what you’re saying.

“Limiting the player makes a better game? Are you crazy? Games should have fewer limits, not more!”

Players should always feel like they have options – but having limitless options is definitely not a good thing. To kick things off, let’s start with a little story completely unrelated to games.

You go to the grocery store because a friend asked you to pick up some flour for a recipe. Now, this happens to be a very unique grocery store with forty varieties of flour to pick from. You find yourself rattled standing before this Great Wall of Ground Wheat Product… What should I pick? Does it even matter? My friend wants to bake some cookies, is there a special kind of flour for that? Before fully succumbing to a panic attack, you race for the nearest emergency exit and make your escape.

Your next destination is a much more typical grocery store, and this time around you find but three different kinds of flour: all-purpose, bread and cake. You think about your options for a few seconds, but it’s pretty clear that cookies are neither bread nor cake, so you quickly settle on the all-purpose flour and return from your shopping trip flush with victory.

While this story is a bit silly, the obvious lesson is that contrary to what you’d expect, presenting someone with a huge number of options does not give them more ‘freedom’ –  in fact all it does is overwhelm them. This has long been a tenet of good interface design. There’s a bit of a ‘rule’ which states that a user’s attention should be split between no more than seven items. The human brain is equipped to weigh only a handful of possibilities simultaneously. I’m sure at some point all of you have opened up some random website that had waaaay too much going on. And you probably weren’t thinking, “oh boy, I can’t wait to dig into all of this, where should I start!” Once someone passes that invisible threshold the end result is nearly always frustration.

That having been said, there are definitely a few individuals who do love being ‘overwhelmed’. The reason why open-world RPGs have become so popular is because they offer players so many things to do. It is possible to provide an immense amount of depth without catering to only the hardcore – the key is proper pacing. Throwing a list of 40 possible quests at a new player within the first minute of gameplay is bad. Starting them with three quests, which then branch into nine, which then branch into 27 and so on is much more inviting.

With regards to the strategy genre in particular, restrictions on unit movement is one of the best examples of how limitations can make a game better. The inability of land units to enter water is why ships are so valuable – and just plain cool. Gaining access to new units with unique ‘powers’ is a major motivation for many players. Just like in economics, scarcity is what drives value – the fact that most units are unable to perform certain actions is what makes the few which can so much fun.

Movement restrictions also show that there’s a place for even permanent limits. An example from the Civ series is how mountains became impassable for the first time in Civ 4. It’s a subtle change that very few players would point to as a major innovation, but even something small like this helps breathe life into the map. Instead of mountain ranges being just another part of the map with a slight movement penalty, they suddenly transformed into true barriers that now require serious consideration.

A dilemma I faced while designing Civ 5 was what to do with strategic resources. I knew that I wanted the game to have a ‘quantified’ resource model where you can have a lot of something or a little bit (in earlier Civ games you either had a resource or you didn’t), but I was unsure exactly how to proceed from there. One idea I played around with was having resources increase the production rate of certain units – say, iron for swordsmen – but still allow players without access to iron the ability to train swordsmen, mainly for balance reasons. After some playtesting I came to the realization that something was off… I eventually figured out that the lack of limits on what you could build made both the units and the resources less interesting.

‘Soft’ limits which hinder the player but don’t completely block him also have a subtle, but still very important role to play. In 4X games a randomized map is nearly always the main source of soft limits. In one game you might start with iron next to your capital, while in another there might be none within 15 tiles. The lack of convenient iron is a type of limitation which helps direct a player towards the best strategies and away from the nonideal ones, but this doesn’t preclude him from committing to get it one way or another.

I’m a big fan of nudging the player towards and away from strategies with the map. When there’s a web of trade-offs to consider, limits of this sort help crystallize what the player’s options are. Let’s say you’re playing a 4X game and want to specialize a city for the production of money. If this can be done equally well in any city then there’s really no special considerations to make – after all, if every city is just as viable you might as well just flip some coins to decide. Which, for the record, isn’t terribly interesting or fun.

If instead an ideal money city is built next to a gold deposit, this provides the player with a basic set of expectations. He knows that if there’s a lot of gold around then a strategy built around generating tons of money is worth considering. He knows that a neighbor with cashflow problems is going to have his eye on that gold right next to his border, so maybe it’s a good idea to build a city there sooner rather than later. But maybe there’s an iron deposit he really needs – our player now has a tough choice to make. It’s obvious that this kind of soft limit makes for a much better game than giving the player complete ‘freedom’ in deciding which city does what.

The last benefit of limits that I’ll talk about is their ability to help ease new players into a game. Developers nearly always get too close to their games and forget how intimidating it is to learn as someone picking it up for the first time. If the player knows his first goal is to find and harvest a particular type of resource, or that he needs to capture a certain part of the map it helps focus his attention and keep him from becoming intimidated by a vast array of options – the alternative is abandoning him in front of the dreaded Wall of Wheat Product, leaving him to sink or swim on his own.

Just don’t go overboard and eliminate all of the player’s control. This is the mistake many tutorials make. You want to teach players, but you also want them engaged and having fun while learning the rules. A bad tutorial is often worse than no tutorial at all, because the wasted development time could have been spent on improving other aspects of the game. If you’re going to bother, do it right! Ultimately, every good game should have multiple ways to complete any one goal. If there’s not, then focusing the player’s attention on a single feature will only bring to light other issues with the design.

Now then, go forth and limit thy players!

– Jon

Categories Design Thoughts

30 thoughts on “Make a Better Game – Limit the Player

  1. Speaking of limiting choices, I went back and replayed some Alpha Centauri (SMAC) recently, and the main thing that struck me was the notion of choices vs. tedium. That is, if there’s a choice that the player should/does always make, automate it. In SMAC, cities should always have a garrison; in Civ V, cities automatically have a defense rating–problem solved. Sure, there may be situations where players don’t want to make the “recommended” choice, but make that the exception and not the rule.

    On the subject of tutorials, scripted tutorials are usually meant to limit players’ decision space so that they can understand the consequences of specific gameplay choices. At the same time, I don’t care for tightly-scripted tutorials that are functionally the equivalent of “hit X to continue”–they’re good for teaching players how to do something, but not necessarily why they should do it.

    In summary, I don’t think it’s so much about “limiting” player choices as it is about getting rid of bad choices. To paraphrase Mark LeBlanc, “all choices should be good ones.”

    1. I disagree that all choices should be good ones. I would change that to “all choices should seem like good choices to the uninitiated.” You don’t want a situation where making choices randomly leads to success. I’ve played a few games like that, where you don’t even have to understand the game to win it. There should always be a path to failure, or else success has no emotional impact.

      1. I think we’re talking about the same thing, even if there’s some confusion about the terms “good” and “bad”; I’m talking about choices in terms of the player experience and not the game outcome. Maybe “meaningful” is a better term for it, and by that I mean choices that are neither equivocal nor obviously lopsided, because neither is interesting for the player. Those are the “bad” choices I think should be eliminated.

  2. That’s a good point Alan. I would argue that limiting options and making sure there are only good, interesting choices are simply different sides of the same coin. The more options you have the more diluted each one becomes. You can spend a ton of time designing 10 expertly-crafted choices but there’s definitely diminishing rewards both for a developer and a designer.

    Of course, you can cut down to two choices that are still unfun, so limiting the player is just one piece of the puzzle!

    With regards to tutorials, the main issue is the level of complexity you find in most strategy games. I don’t think most players would complain too much about needing to read a couple paragraphs in order to feel comfortable with a game – the problem is that strategy games tend to need closer to two HUNDRED paragraphs to provide a decent explanation. To properly teach a strategy game to a new player you really have to go above and beyond what other game genres can get away with.

    This is what we tried to do in Civ 5, really integrating the advisors into the core gameplay. This sort of thing is especially important now that fewer and fewer players bother starting with the tutorial and instead prefer to dive right into the game (for good reason – most tutorials are about as much fun as installing the game). Games have gotten better here but there’s still a ton of work to do. Hmmm… that gives me an idea for a new article…

    – Jon

    1. Not directly related, but I enjoy mechanics that further focus your choices between short-term and long-term gains through use of some imminent threat. The best contrast that comes to mind for me are the key differences between Uwe Rosenberg’s Agricola and Le Havre. In the first you must balance the immediate need of providing food under threat of a severe penalty with actions that will benefit you in the long term, whereas in Le Havre ignoring immediate needs results in almost no penalty. The restraint provided by immanent threat in Agricola makes it a much more interesting and enjoyable game for me. On the other hand, I know players who find this mechanic extremely stressful and prefer Le Havre. So maybe I like having some “bad choices”, at least in board games.

      In a 4X game, this tension is provided by you opponent’s military actions. Speaking of which, I’d like to compliment you on Civ 5. I am really enjoying it, although I didn’t start playing in earnest until the latest patch that fixed many multiplayer issues. Have you played any games to see how the Civ5 multiplayer community has developed?

    2. Jon,

      I have felt that integrating the tutorial into the beginning of the campaign really helps, since you are progressing at the same time you are learning the game. It makes it more fun than just running through a boring tutorial where you aren’t accomplishing anything tangible with the work you are doing.


  3. There’s a body of research coming out now about the phenomena of decision fatigue. The general idea is that if you ask me to make too many choices in a short period of time the quality of my choices deteriorate. Additionally, I get crankier and more annoyed at the number of choices that I have to make. This holds even if the choices have no practical impact (eg. what color of identical detergent would you like?). The research is focused mostly around consumer choices, but I can see how the idea also applies to designing choices in games.

    Nice blog, I look forward to hearing more from you.

  4. I imagine another tough aspect to game design is to make sure that early choices don’t limit later choices too drastically and start to corral player actions down a particular path. Let the player adapt their strategy, otherwise the mid to late game suffers. I don’t think the main failure of most games is presenting the player with too many choices, but not having a clear indication of the effect of making that choice. Another biggie is not giving the user access to all of the information so they feel like they are making an intelligent choice. I don’t think Civ V suffers from the latter 2, but sometimes what choice to make seems a little automatic to me. Maybe it is because I have played every version for hundreds (if not thousands) of hours. Having quantities of strategic resources is perhaps one of my favorite changes Civ V made to the series.

  5. I like the limits that noone really notices: For example in Canabalt you think you’re in control of the avatar, but you are limited to a single action: make him jump. The running aspect is taken out of your hands. It’s like a split in your personality: One half of you (the insane half) controls the legs and runs no matter what, the sane half (that’s you!) tries to compensate for the running madness by jumping at the right time in order to avoid obstacles.
    One half runs straight into death, the other half wants to stay alive. What an easy but effective way to set up conflict 🙂

  6. Thanks all for the support. Glad you’ve enjoyed the first article. Many more to come!

    I agree, having to balance the present and the future is a lot of fun. A game I’ve been playing a lot of the past few months is Out of the Park Baseball in a multiplayer league, and that is one of the constant trade-offs you make when managing a sports team. The fact that things can suddenly change and a plan can unravel is critical to this being possible. This is tricky to get right in a strategy game though where players expect to have control and don’t WANT something derailing their perfect plans. One of the articles I’ll be writing soon is on the subject of how important player adaptation is to a successful game.

    I have not played any Civ MP lately but I’m glad to hear that it’s received some more love!

    That is definitely a challenge. Pacing is one of the hardest things to get right in a game and it’s a concept intimately tied to when to apply limits to the player (and when to lift them). With regards to information, a key aspect for game developers to keep in mind (courtesy of Sid Meier) is that the real meat of a game takes place in the player’s head, NOT in the game itself. It’s something that Civ 5, along with nearly every game, could improve on.

    – Jon

  7. Jon, great to see you finally getting a blog going. 🙂

    There’s another form of ‘limits’ that I think you’ve not considered, that is the self-imposed limit. Basically, the self-imposed limit is something the players themselves do to limit decisions. This is really good for the player if they want to focus on some decisions but not others.

    A great example of self-imposing limits is in Hearts of Iron 3. The unit hierarchy management model allows for full customisation of the decision tree that the player is exposed to. For instance, in September 1939 the player can create two “fronts”, being east and west. For the entire western front the player can assign a Field Marshall to manage all aspects of the front (playing for the player). For the eastern front the player can further split that into Army Group North and Army Group South and assign another General to Army Group South. This leaves the player in command of Army Group North to manage the invasion of Poland. Meanwhile, the other two areas the player has AI’ed off are managing themselves and make infrequent requests to the player for units/reinforcements, resources etc.

    Of course none of this would be possible without the design of the game being able to handle it, which brings it back to the designer to foresee these possibilities and cater for them.

    Good luck with the blog mate. 🙂

    1. Thanks Dale!

      That’s an interesting observation. Trying to build in self-imposed limits is very tricky, because you don’t know how a player will react. They might take the ball and run with it, or they might find the system an unnecessary burden and seek out something else to latch onto.

      It also brings up the question of how to handle special game modes like one-city-challenge where the player voluntarily sets arbitrary and harsh restrictions on himself. To be honest, I’m not sure any designer really understands the process by which a community can make a game its own and help it evolve beyond anything someone could have anticipated. Just another reason why fostering a strong community is so important!

      – Jon

      1. Totally agree on the community point and this is one area they really shine as they can help a designer gauge how the player wants to actually play their game. This is one area that Firaxis has really shone by integrating their community into the development process they get that fanatical fine-grain feedback. They also are able to see which decisions are important to players at whatever point in the game.

        I don’t think it’s too difficult to gauge what self-imposed limits a player could use. A lot of games already unknowingly (or possibly knowingly) cater for these. Take for example Colonization. The original had the custom house which the player could use to automate trade to Europe. The new one has trade routes. The player sets a max stockpile, and the custom house/trade route takes care of the rest. Civ’s research goal is another example. The player sets a research goal and then let’s the science adviser pick the techs needed to get to that goal.

        By self-imposed I’m not really talking about house rules like the OCC in Civ. Self-imposed limits are more options that a player can use to help reduce decisions the player doesn’t deem necessary at that point in the game so they can focus on the area of the game they need to. When the focus changes again later on in the game the player can then alter their self-imposed limits to change where the decisions need to be made. By allowing the player to change what they focus on in the game at any particular time, you’ll be able to keep a deeper decision tree as the player can self-impose limits to their desired level.

        I suppose you could say these are “temporary limits” which the player can use to help keep their focus on the particular area of the game that requires it at that point in time. 🙂


  8. Christopher McClatchey April 10, 2012 — 4:42 pm

    I think it’s worth considering that there may be a point at which being overwhelmed with choice ceases to be a detriment and becomes a positive. Think about Minecraft – with no limits whatsoever, it managed to create a pretty darn compelling experience.
    I think a large part of this, moreover, is the presence of the friend – the director telling you to do something. “Find the right bag of flour” is not fun; “find a bag of flower – any bag” can be, as is “I want to buy flour to make cookies; let’s experiment with one of these.”

    1. Minecraft and other similar games ala SimCity are definitely an exception and a fascinating case study. In a way they’re more toys than games, giving the player a sandbox and a large set of tools and setting him loose to make up whatever goals suit him.

      You could apply the same logic to open-world RPGs, but it doesn’t work quite as well. I think they tend to occupy an uncomfortable middle-ground between game and toy – some structure is in place (a main story, quest chains, etc.), but the player is still free to do whatever he wants. On the other hand, games like SimCity provide no real goals beyond what the player decides for himself.

      I wonder if there’s room for a more sandboxy RPG with almost NO structure… just a bunch of realistic AI characters and locations that have rules which make logical sense. Maybe a project I’ll tackle when I’m retired in another 30 years!

      Oh, and don’t get me wrong, I love me some open-world RPG. Logged 95 hours into Skyrim at this point, which I know is pathetic compared to some of the folks out there!

      – Jon

      1. Christopher McClatchey April 11, 2012 — 2:55 pm

        This is something I’m working on in my spare time – especially where artificial societies are concerned. It seems quite possible to create an autonomous agent which presents a reasonable facsimile, not just of human likeness but of human actions. Stealing if hungry and dishonest, taking revenge for the death of a son or brother… All the typical RPG quest types should be natural artifacts of such a system.
        Making said system, on the other hand… Well, Dwarf Fortress proves it’s hypothetically possible.

  9. So this is one of those topics that I can see the intent of where the author is heading, but then I have to say… “except,” “but,” and “however.”
    Now obviously (even in the Matrix), games have rules. Rules are, by definition, limits and getting to evetually break those rules is the when players reach the hallowed ground.
    Movement limits, action points, hand size, weapon or armor restrictions, line of sight, fog of war, research and skill trees, spell levels, jump distance… all limits. All key elements in their respective game types, and rules which are pretty much taken for granted which help define the genre of the game.

    But limits are different than options.
    So, the flour store analogy seems like a good place to start, since it crystallizes a pretty clear image of a wall of headache-inducing confusion. But in the analogy, you are purchasing flour for your friend, who was making a cookie recipe. If your friend, who obviously bakes, was standing in front of the flour wall, would they be overwhelmed?
    So, while a myriad of possibilities can be overwhelming at first, later they might become important in order to distinguish the flavor of cookie a, from cookie b. How does a developer help a player get to that level of discernment? Or should they not bother?

    An anecdote I would offer up is the success of both Starcraft and Total Annihilation, RTS games which were released at about the same time. Total Annihilation focused on a vast array of unit types to fill various combat niches whose strength was balanced strictly on resource cost, while Starcraft favored a diminished number of unit types, but offered up a streamlined, more rock, paper, scissors system of balance. Which was the better game? Depends on who you ask.

    I think the important thing is not how many choices a player has in front of them when it comes time to make a decision. It’s that whatever decision the player does make, whether it be flour or armor, has weight and importance. If scale mail, ring mail and chain mail all do the same thing, the choosing means little. If I can only swim in ring mail, stop arrows with scale mail and enter the royal croquet club in chain mail, then the choice matters and over time players may appreciate the differences.

    That’s not easy to see at first, when someone is new to a game, or simply staring at the box art, but if asked how many avatar portrait choices are enough, or character hairstyles, or available spells, or wandering monster types, or guns, or factions to play? Even those players would likely say “more is always better.”

    1. Great post wmackie. It’s a mistake to assume that one size fits all and that a single set of limitations will appeal to everyone. At the end of the day it comes down to what your target audience is and the preferences of the designer. Which is, of course, why game design is more art than science!

      Your point about decision importance versus simply limiting options is one several people have brought up in response to this article, and I agree that’s the ultimate endgoal. But an article stating “decisions should be interesting” probably wouldn’t have had much to it, nor would anyone disagree!

      In reality, the article is talking about two different things: not overwhelming players (which requires adapting for taste, as noted above) and using limits to increase strategic depth for ALL players (ala impassable terrain). If I were to write it again I think one thing I’d do is better clarify that distinction.

      – Jon

  10. Jon,

    I definitely agree that soft limits can make a game more interesting. For example, think of the difference between building up a city in Civ 4 versus Gal Civ 2. In Civ 4, the only specializations in each city were any wonders you built there, since all cities coukd build as much as they want and could all build the same things. It was boring and most people just had a default build for every new city they founded. Now look at how fun it was to build up planets in the second Gal Civ. You could build whatever you wanted, but there were limits on how much you built so you really had to make choices on what was important. If a planet had bonuses to production, then you would likely make it focused on ship building; if it had bonuses to morale or economics, you would likely make it a banking planet; if it had bonuses to research, you would likely make it a research planet. The point is it was fun to have specialized cities that were good at one thing at the expense of others, versus the boring one-size-fits-all cities prevalent in the Civilization games.

    That said, these limits have to make sense in the game. Limits based entirely on he gameplay that don’t jive with the story will ruin the immersion of the game, and make it seem more like a job than like fun.


  11. I think the player should be able to CREATE their own choices. To take the flour analogy one step further, imagine standing in front of a row of bins of different grains, and choosing a little bit of this and a little bit of that to make your own excellent wheat-soy-white flour mix for the bread YOU want to bake. Not the recipe on the back of the bag of flour!

    The huge disappointment, of course, is that custom unit design never continued post SMAC…

    1. This works for some people, but not everyone. There are definitely those who loved SMAC’s unit workshop, but also quite a few that feel it was the worst part of the game. The big problem is that there were really only a few ‘best’ combinations that would get used every time. The lack of balance in the system meant that there was little to no reason to adapt, and you could get away with using the same models in every game. I believe a unit customization system CAN work, but the one in SMAC was too crude and not integrated into the rest of the game well enough. For those able to look past that and simply enjoy creating cool units it wasn’t a problem, but all of the other players (like myself) were left wanting more.

      – Jon

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

search previous next tag category expand menu location phone mail time cart zoom edit close