Three dimensions of game design: Simulation, Playability, Abstraction

My mate Chris and I often debate game design, and specifically simulation versus playability. Wargamers typically think of these as opposites, with a set of rules being either realistic or playable. Chris is, for example, an advocate of simulation and is willing to sacrifice playability to get something he considers more realistic. But I don’t think these things – simulation and playability – are opposite ends of a single dimension. I’m an advocate for playability and simulation is also important to me. I want both. Abstraction is the key to unlocking this combination and is an important third dimension.

Simulation VERSUS Playability

What got me thinking about simulation and playability was Review of the wargaming rules WGR 1920-1950. The summary was “the game is an incredible well-detailed combat simulator, but isn’t that much fun to play.” And “Great Simulation… Poor Game Playability!”

The simulation versus playability debate has been going on a long time.

What semblance of realism we were led to expect is sacrificed on the altar of playability

The quote is by Victor Madeja back in 1973 (cited in Geerkatimedia) and he clearly had a strong preference for simulation over playability.

And that is the often the level of the argument. Either realistic (good simulation) or playable (good game). A linear axis from one to the other.

One Dimensional Game Design
One Dimensional Game Design

Simulation AND Playability

But there have been contrary voices around for about as long. John Hill, who later designed Squad Leader, said (cited in Geerkatimedia) :

One of the hardest problems facing any war game designer is the careful balancing between playability and realism. Actually, any reasonably competent wargamer could probably design a realistic ‘simulation,’ but to design a good game is something else. As an example, 1914 was an excellent simulation of corps level fighting of that era, but as a game it was worthless — it couldn’t be played.

In other words, simulation and playability are different dimensions. The example is a good simulation that is a unplayable but this view also allows for a good simulation that is also be a good game.

Simulation AND Playability … AND Abstraction

Apparently John Hill later became an advocate on “abstraction” (Geerkatimedia). A game design philosophy I’m very keen on.

Dave Eng, describes abstraction as “just another method of ‘leaving it up to the imagination’ for players” and comments that “abstraction takes a complex concept and makes it easier for the player to grasp” (University XP: Abstraction in Games). That is true but I think misses the key difference about attraction in game design.

Christian Lindke, when discussing Simulation vs. Playability, explains that a designer using abstraction is less interested in “what actually happened” and more interested in the “effects” of what happened. In other words, the designer models the effects and not the action itself. Crossfire, for example, abstracts shooting to a single die roll that glosses over how alert the shooters are, whether they have ammunition, whether their weapons are in a good state of repair, whether they spot the target, how many bullets were fired, whether the bullets hit, where each bullet hit, and the target’s response to the shooting, both individually and collectively. In Crossfire the player just determines the effect of the machine gun shooting, i.e. whether the target is PINNED, SUPPRESSED or KILLED.

Christian Lindke observes that “most gamers or designers are a combination of abstractionist/simulationist”. I agree in general, but I disagree with the implication that abstraction and simulation are the opposite ends of a single dimension.

I think simulation (realism), playability and abstraction are interrelated but separate variables. Games range from realistic to unrealistic, from playable to unplayable, and from abstract to concrete (detailed).

Simulation, Playability, Abstraction - Three Dimensions of Game Design
Simulation, Playability, Abstraction – Three Dimensions of Game Design

Steven’s positions on simulation, playability, and abstraction

Of course I wrote this post because I have a firm view on game design.

It seems to me that some game designers strive for realism via concrete mechanisms. They write detailed rules with special sub-rules for every contingency. From my experience these game designs are unlikely to achieve my playability goals. They might work for other folk but not for me. My main gripe with this type of game is that they take too long to play, often because the players have to play with the rule book in their hands.

As an example of an concrete unplayable game, I recall my father and I tried to play a board game of the War in the Pacific, I can’t recall the name now. It took two days to set the game up. After two days of effort, and no gaming, we just packed it up again. It might have been realistic, although we couldn’t confirm because it was too detailed to play. Too concrete. Unplayable.

Detail in rules (concrete mechanisms) does not guarantee realism. My impression of Live Free or Die is that it is playable, concrete, and not very realistic. Massive caveat there because we only tried it once.

Personally I like game systems that are both realistic and playable. From my experience these game systems tend to be more abstract and less concrete/detailed. I’m not so interested in what actually happened, I can leave that to player imagination. I’m more interested in the effect of what happened. Crossfire fits that design criteria, for me anyway. A machine gun stand fires and there is a rule for whether the stand hits or misses. It is up to the player to explain why the stand missed (distracted, out of ammo, target was veteran and just shrugged the casualties off, whatever).

24 thoughts on “Three dimensions of game design: Simulation, Playability, Abstraction”

  1. Thoroughly agree Steven. I believe that DBA does an excellent job of this as well. As Barker points out; a commander would not know how many men he has lost in an action, only that they are pressing forward, recoiling from the enemy or in flight.

    • DBA always does my head in. I don’t think it is actually a terribly good simulation for certain conflicts I’m interested in (Punic Wars, 100 Years War), but it is has some simple abstractions that make it very playable. So I keep playing it.

    • In my opinion DBA does a good job for abstraction, a sort of ok job for playability (basically it is fast) and a poor simulation. I have never seen a DBA battle that ressembles anything to an actual battle. Games are fun, there are tons of tournaments and it is played a lot, so as a game, it works, but I can’t imagine Scipio wondering about the geometrics of a recoil to kill a Carthaginian base.

  2. Spot on! Who really cares in a game whether the gun jammed or whether the firer lost vis for a few seconds due to smoke, what matters is the final affect. Like you said with CF. Good abstraction makes a good playable game, which is STILL a good simulation of WHAT MATTERS. Rivets and most details simply do not matter but rules-writers keep adding unneccesary detail.

    • Yes. I suspect some wargamers confuse detail (“concrete” on my abstraction scale) with realism. They are different.

    • Wow. Found it: As it was in the beginning. Over an hour to simulate 20 seconds of combat!

      Personally I prefer simulations that are are at least as fast as the actual event they are simulating. I don’t have a 100% success rate on that, but I drift towards rules and scenarios that generally achieve it.

  3. Interesting article. There’s certainly a tension between detail and playability, and in general I think newer rule sets have got better at avoiding the rivet counting trap.

    • There does appear to be a trend towards abstraction. But some model rules are still very detailed / concrete … and don’t appeal to me much.

  4. Abstraction should be appropriate to the command level of the game, shouldn’t it?.

    Somebody (I want to say Patton but it was probably somebody else, and somebody on this thread probably knows who… 😉 said one should be concerned with one level of command up and two down. GEN Barry McCaffrey said this too, and noted further you should know what’s going on two layers down but should only command one layer down. E.g. in Crossfire, as company commander, you have an idea (generally via the scenario OOB and rules) what higher level assets you have, you command your platoons, and you know approximately where your squads are.

    Greater realism – read: simulation value – means, I think, more abstraction the further away from your command level you get. And this isn’t in contrast to, but rather, in support of increased realism on the game table. Your observation about WHY the MG missed and Julian’s note about DBA armies describe this perfectly.

    • All good points, but for the most part, the position of a set of rules within the 3-dimensional space is a matter of opinion. Most likely only the most extreme examples for each dimension can be located with any accuracy and still, a matter of opinion!

      It might be interesting to have several experienced gamers allocate 4 or 5 popular games on the 3 dimensional spectrum to see how it comes out.

    • Agree 100% that abstraction should be at the appropriate command level. And that is realistic.

      A general doesn’t have to worry about how many bullets a particular bloke shoots, and how many of those hit. A general wants to know if his subordinate divisions are in supply and achieving their objectives.

  5. This was spot on! If it’s not playable, it really doesn’t matter how historically accurate the rules are.

  6. Another way to look at the simulation/playability issue is bottom up or top down. Many rules start at the bottom and calculate weapon effects in detail. On their way up, strange things often happen. Such rules can have you wondering how the hell that happened. It can also see a lot of ink spilled trying to account for highly unusual events like Ramsay’s horse artillery troops riding through French cavalry at Fuentes D’Onoro. Just write a scenario-specific rules if you must.

    The other way is to start at the top and try to design the overall experience, then decide what combat effects to model. DBA, Crossfire are that sort.

    Too many hours were wasted in my youth playing games that took an hour to resolve 10-15 minutes of combat. At 1 in the morning, we packed it in without a resolution, bitching “in another couple turns I’d have captured that chicken coop”.

    • Vincent, Yes! That goes to the heart of what I was trying, in a muddled sort of way, to say…

      An old friend is an avid Air War gamer. Remember the old SPI board game? He once said it was the best four hours one could spend simulating 30 seconds of combat.

      NOT for me!

    • Exactly Vincent. In my terms the bottom up approach is “Concrete”. And the top down is “Abstract”.

      I’m very interested in getting to a conclusion within the time available. A top down, abstract, view helps achieve that.

  7. I tend to work on the basis of the 4″A”s – Abstraction, Aesthetics, Authenticity and Ambition, where ambition is what we – as the player – is trying to get from our gaming.

  8. Very well said, Steven. Also, your preferences re: abstraction/concrete resonate a lot with mine. One thing I’ve noticed is that, as a broad generalization, miniature rulesets tend to execute the ‘abstraction’ part a bit better than hex & counter ones. There are notable exceptions of course!

  9. Steven, enjoyed this post immensely. Being in the middle of multiple game developments, that consider different eras and environments, I have concluded it is abstraction that leads to the top down effect, or desired conclusion, and the choice and methods of abstraction are essential to playable games that don’t simly run amok. Thanks for you insightful posts.

  10. I should have mentioned it in the post, that my Twilight of the Sun King rules takes attraction further than most game systems. I think they are a realistic simulation of battles of the War of Spanish Succession. However, Twilight does not model firing and melee explicitly … these activities are bundled inside morale. The rules focus on the effect (morale) not what happened (firing, melee). The expanded version of Twilight is quite popular, but that level of abstraction might not appeal to a wider wargaming audience.

  11. I enjoyed a lot you post and I think you got it totally right. Many gamers confuse concretism with realism and that is why we have a bad assessment on many games.

    On the other hand, I am not totally persuaded that the three dimensions are the only issue, specially concerning playability. In fact that is a term that requires a cautious definition as I think there is not a universal understanding of it. On the one hand playable can be easy to play, but I have realized that the key issue with playability is actually what people find fun to play. Rock-scissors-paper is a very playable game (easy to settle and fast) but a quite boring one.

    In wargames I have remarked that moving miniatures and having a total control are key factors considered by many wargamers for a game to be playable, but in reality that is what they find fun to play. Thus, there is an overemphasis on moving over other types of decisions, usually having an excess of control over the situations. Taking into account that generals many times had too little to no control of micro management of units you have a clash that is reported to be between realism and playability because actually people find it fun to take decisions that actual generals could not undertake or take little to no atention at all.


Leave a Reply