Wednesday 5 April 2023

Game Design #92: Complexity v Detail v Realism, Decisions and "Flow"

While sitting in the car pondering my tank rules, I was thinking about where the rules would be allowed to be complicated and detailed, and where they are not. I.e. I didn't mind complex activation rules (see previous Tankhiem post) but I was keep to keep the shooting/damage mechanics pretty basic and minimising recording.

Complexity. This is how elaborate, or convoluted the rules are.

A ruleset is complex if it has many special rules/traits/abilities which 'break the rules' or are an exception to the normal rules. For example, the Song Of... series boasts of its simplicity with only 2-3 unit "stats" but then has hundreds of special rules and traits to differentiate units - not so simple!

A ruleset is needlessly complex if it does not follow normal gaming stereotypes and tropes. This could even be how the rules are laid out - having to find a rule in a random spot. Or the language used. I recently read Killwager and it was needlessly complicated by renaming common wargaming terms "flow" "measures" "performances" "conflict" - having to mentally translate them to "action points" "actions" and "turns" "opposed rolls" is exhausting (and it smacks of pretentiousness).

A ruleset is complex if it does not use uniform, consistent mechanics - i.e. Warmachine uses 2d6 + stat to beat a target number. This method is used pretty consistently.  In contrast, Bag The Hun uses 7 dice methods (and dice types!) including tables, beating a target number, counting doubles, etc.

A set of rules can be complex in many ways. A good sign a rulebook is complex is if you have to regularly consult it, not for tables and modifiers (see detail below) but because it is often unclear how you should do something.

Detail. A detailed ruleset might have modifiers for everything, including if the commander spilled coffee on his lap. It might have individual stats for every tank gun used in WW2. Detail does not automatically = complexity, but usually infers an increase in things to recall. The super detailed tank guns might share simple, easy-to-follow rules: beat a number on d10 to hit, beat a number on d10 to knock out. That's quite a simple method. There just might be 101 different variations of gun. Detailed rules tend to have long lists, tables and inventories which may indirectly be exhausting. A more abstract ruleset might classify those guns into 4 groups, lumping similar-ish guns into Type A, B, C or D. Detailed rules may also be more dense, detailing every possible outcome or permutation of an action. More abstact rules tend to be far more brief (but may be ambiguous or feel too 'vanilla'). Detailed rules are not necessarily tricky to play - the mechanics may be logical - but usually involve checking tables and modifiers as there is too much to easily remember. It's a sign it's detailed if you check the rules - not how to do something, but to check exactly what a particular modifier, range band or piece of kit does.                                                                                                                                                   

Decisions. When discussing my tank activation rules with a friend, we kept using words for 'gives you a chance to decide' or 'gives you a chance to act'. aka Player agency.  

Now your decisions usually always include Where to Move and Who to Shoot.  This is your basic old school IGO-UGO 40K-a-like.

However they can also include: Who Moves First (alternate activation) - which is about the action sequence/move order.

But depending on your rules, it could include:

Who activates first? Who shoots first? Should I make an opponent activate? Should I move OR shoot OR choose between other actions?  Should I 'save' my shot (or move) for later? Should I interrupt my opponent? Should I use area fire or direct aimed fire?

Many decisions lie in the part of the rules which is usually named something like 'initiative' or 'activation' or 'move order' or 'sequence of a turn.' It usually is to do with WHEN and WHO of an action. As you can see, relying on IGOUGO tends to strip away many decisions. Alternate move (you move a unit, I move a unit - like chess - at least adds a "who acts first" element.

Having lots of decisions allows a lot of player influence on the game. However, too many decisions can actually be exhausting and create mental overload. A bit like trying to watch Inception when you are tired. I call this mental load. Sometimes you're just in the mood for Pirates of the Carribean.

Realism. This does not mean rules are complex or detailed; just that they give results that are similar to real historic battles. Players would make decisions similar to those historic commanders would have faced. Historical tactics work in game as well. A historical battle, if replayed using game rules, will yield a similar or logical result. Science fiction and fantasy has 'realism' with regards to aligning with popular media such as movies, books. I.e. star fighters tend to be expected to behave like WW2 aircraft, while the larger ships act like WW1-WW2 naval ships. Let's say a division always inflicted double the casualties than another. Having the first division damage on a 1234 and the second on a 56 is actually realistic - just not detailed. Realism does not automatically equate to complexity.

"Flow" / "Actions per turn"

We are trying to simulate a simultaneous real time event with turns - slices of time that represent actions a real soldier can take within that period. Basically, everyone but the acting forces 'pause' while your unit/s take their turn - frozen in place. Imagine a slide show. A 10 second pause is more obvious and jarring than a 1 second pause.

The slice of time your 'turn' takes can vary; but I'm more interested in in-game effects. The more you can do in your turn, the more obvious the 'pause' and the more disjointed/less 'real time' the gameplay is. I use the term flow to describe how fluid a wargame is. Is it 60fps or 1fps?

APT 2.0

I'll call the amount a unit can do as 'actions per turn' - and the average is 2 actions. Typically, move+shoot, move+melee, move+move (sprint) or even shoot+shoot are common. In effect, everyone pauses while the unit does its two actions. It's pretty standard in wargaming if you think about it.

So - the less actions, the shorter the pause - the greater the flow. I've been using the "APT 1.5 method" in many of my homebrew games - one-and-a-half actions per turn. Basically, a unit gets 1 guaranteed actions, and rolls against a stat to possibly gain a second action. I reduce the average actions by .5 (better flow) and add more uncertainty (friction) into the game. Obviously only 1 action or less per turn would give even more 'flow' but I have found it can be too restrictive. 

Less actions = better flow; no actions = boring

If a unit could have 4 actions - 4 x 6" moves, or 4 x shots - the game is far more disjointed than if it could make a single 6" move or a single shot. The more the activated unit rambos around; the more obvious that everyone else is a static NPC. The stronger the active turn, the more you do in the active turn - the more obvious it is that everyone else is standing around like dummies.

Reactions, Overwatch & Flow

Reactions and overwatch (hold fire/move) increase flow, adding a feel of 'simultaneity' as you can interrupt an enemies turn. However they do add drawbacks in terms of 'exceptions' to the gameplay (complexity); i.e trying to determine who reacts or when units can react, what sort of reaction is allowed etc - you could end up reacting to reactions!

Some activation methods include an element of interaction. Alternate activation (chess style you-move-a-unit-I-move-a-unit) allows some measure of natural reactions and flow - you can't always react with a specific unit (micro reaction) but you can often respond with a different friendly unit (macro reaction) - there is some degree of response to your opponents action/s. IGOUGO has no natural reactive flow - you move every unit, unopposed, while my whole army stands around like dummies. You can respond only once - after every unit has acted. There is no reactions or give-and-take within the turn.

It's important to be allowed to react at some level to enemy actions - feeling helpless is not fun. For example, one activation method (B) discussed (see previous Tankhiem post) could conceivably have a tank fire 3x and knock out 3 tanks if they got the initiative first - not much fun to be on the receiving end - and also not 'realistic' in the sense a single tank could dump 3 rounds before any other tank could even move.

Okay, this post has gone on for a bit and I feel that reactions could probably be explored more thoroughly a separate post - as I do not enthusiastically espouse reaction mechanics like I would have 5 years ago (Infinity and Ambush Alley games were my jam...)


  1. Great, thoughtful post. I'd love to read a post about your views on reaction mechanics, how and why your perspective has shifted over the years. Your older posts made me really think about activation and player agency in games, it would be fun to read how your views have evolved since writing the Genesis for the evolution of my views on game design.

  2. I realise I've gone over reactions a lot already...

    ^^^Mostly defining reactions

    ^^^Criticism of reactions, and use of phases (or shorter segments) as solution. It ties in with "flow." A new post would be expanding on this

    ....Also depends on genres; firing a musket is not as needful of reactions as exchanges of automatic gunfire

  3. Most of this is motherhood and apple pie for those who've read the design series, and gets to the crux of a game designer having to make clear choices over how the player should be spending their time, and how much of it, what their idealized game flow should be like. If the intent is for something more like a RPG, with strictly limited unit count per side and high flexibility, then something like APT 1.5 plus Reactions is perfectly reasonable. If the intent is for a GW 40k-scale mass battle, then APT 1.5+R is a terrible system that will bog gameplay.

    Igo-Ugo is consistently unfairly maligned when it's actually among the more efficient ways to get a total number of Actions over a given number of Real World Hours, which matters a lot when you're playing with any sort of real world time limitation. The ability to coordinate multiple units is also a pretty big deal - it's a major feature not a drawback. Finally, most players like that they have a very high degree of control over their units during their turn. That is, there are very good reasons to choose Igo-Ugo.

    Yes, interactive gameplay is nice, but there's also a limit to how far one can take it, just the same as with an overload of decision-making or an overload of unpredictability. It's the designer's job to understand how much decision making, interactivity and unpredictabilty a player should deal with, so that the player can focus on meaningful decisions, relevant interactions, and a sense of both strategic and tactical planning. If it's layered decisions over every little thing, constant opponent interruptions, and shifting randomness, then the game becomes unsatisfying because the player cannot plan nor meaningfully control their forces to execute their plans.

    I like 2 APT+R, and am of a sense that that Reactions ought to be formally integrated into a concept of time and quality, where there are costs and tradeoffs for reactions, whatever form they take. I don't yet have an answer of how this should work, but my sense is we ought to distinguish between reactive 'unplanned' reactions like snap fire vs. proactive 'planned' reactions like area overwatch.