Monday 14 December 2015

Game Design #59: Unit Count - Is there a "perfect number"?

This was inspired by someone asking how many units my homebrew Aeronef game was intended to handle.  I was thinking "oh, about a dozen ships"  because it was pointless to aim for the twenty-to-thirty per side of the original Aeronef rules as it was "too many."  Yes, too many toys on the table. How could this be bad?

My thesis is that there IS a "correct" or "best" number of units for a wargame, and too little or too few is a bad thing.

First, some home made definitions:

Unit = a miniature or group of miniatures or stands ("elements") that maneuvers together and fires together as a group.  In effect, they are a single entity although they can be made up of many elements.   They tend to either be kept a in "coherency range" of 2-3" or kept on a movement tray to keep them together and distinguish their "togetherness."

Element = this is either a single miniature or a Flames of War/DBA type diorama base with miniatures fixed onto it. 

If the troops move and shoot individually, there are 9 units.  Each mini is an independent unit, like in LOTR:SBG or Infinity.  For those familiar with 40K or Warmachine, you'd see three units - the crossbowmen form a unit, the pike are a unit, and the priest is heroic unit.

My suggestion is, no matter what the "scale" or "level" of conflict the wargame represents, be it 1:1 squad level skirmish, platoon level, company level, battalion level or even division level, there is only so many units a wargamer can comfortably handle. Too few units and the game becomes lame or simplistic. Too many and the game bogs down without any extra tactics or strategy being added.

Imagine a game where each side has a single squad of 10 men. That squad travels in a group, moves and fires as a single entity. The 10 men are a single unit.  The game is essentially a 1:1 duel between two units.   The game will not have many tactical choices (decision points).  There are not enough units to make it interesting. Now imagine the squad of 10 has been broken up. Each miniature can maneuver and attack separately.  There is now ten units per side - as each soldier is his own "unit."  Obviously, there are now a lot more decisions to make.

But there is a point where more units do not add more decisions and the game bogs down.  Let's say the game is a company level one.  You could have a game with 90 troops, each moving and firing independently.  That would be, in effect, 90 units.   A bit crazy.  It makes more sense to collect units into, say, 9 squads of 10 - reducing the number of independent units to 9.  It's also very unrealistic and unhistorical  to expect a commander to individually micromanage the individual actions of every single one of those 90 men.

How many units here? If the space marine squads move independently, there are three.  If they act together by platoon, there is only one unit. 

I don't want to confuse the argument...
Now, the point a game "bogs down" can also vary wildly depending on the complexity of the rules.  A game of Infinity can be slow with only 15 minis if they all have complex special rules.  A game of LOTR can handle more in a fraction of the time.  This is not amount of units vs time.  Because that depends on how the rules were written and is impossible to "prove."

What I'm trying to focus on is the amount of units vs the tactical choices.  With only 1 unit each side (even if they have 10 men in them) if they move and shoot together as "one" you have a 1:1 duel.   The 10-man unit could equally be represented by a single man, with 10 hitpoints.** 

So we understand a "unit" can take a lot of different forms: single guy, a bunch of guys which move in "cohesion", a movement tray with 20 guys... so what IS the "perfect" amount of units?

Well, I'd say you need at least four units, preferably five or more.

Let's look at historical examples of what was "manageable."

If you think of ancient armies, there was often a right flank, left flank, centre, reserve, and maybe a skirmish line and cavalry element (usually on a wing or both).  That's 4-7 independent "units." 

In a WW2+ platoon, there's three squads, divided into fire teams, (6?) with maybe a HQ element and maybe transport vehicles. Depending on how they are divided, that's 4-10 units.

A naval task force might comprise of a amphibious assault ship, a destroyer, a submarine, a few frigates and a supply ship.  That's 5-6 units.

A fighter squadron of 12 might be divided into pairs or finger fours. That's 3-6 units.  

A lot of wargames allow players to micromanage to an unhistorical and unrealistic extent.  A WW2+ company commander might control 3 platoons (9 squads), a HQ element, and maybe 2-3 attached heavy assets. However he likely only manages the platoons or at lowest, the squads - he doesn't interact with each individual fire team or direct the fire of individual soldiers.  So he in effect controls 5-7 units, not the 40 or so sub-units that actually exist in his force.   Allowing players to mico-manage at command levels far below their "level" is a trap a lot of wargames fall into, and creates many more units than are actually needed.

So about this mystical "perfect number" of units in a wargame?  I'd say 5-12 "units".

In a game like Tomorrow's War, this would be six units. The troops are grouped into three units of four, but each vehicle would be a separate unit.  Also, I wanted to show off my $2 sci fi APCs from the junk shop.  Many Hotwheels-scale cars scale well with 15mm.

Your reasoning, sir?
Well, many military organisations tend to historically divide into 3-5 units, be they squads in a platoon or divisions of WW2 warships.  Besides, common sense.

An army made of a single unit isn't very flexible. A two unit army is also very simplistic.  Three is the bare minimum for pinning, flanking etc. Four or five units you're starting to get more options - like even ancient armies had - you can have some sort of HQ/reserve and perhaps a scouting or specialist unit.

I've noticed any wargame which only has 3-4 units or less tends to rely on gimmicky rules or mechanics to add interest, be it resource management or special abilities/attack combos - and not on traditional wargame maneuvers.  

Now, what about the "maximum"?"  Assuming you also can "macro manage" one command level below, traditional armies could potentially divide into ~9-16 sub-units if you include the tier below.

Is there any strategies you could execute with 16 units that you couldn't also perform with 9-10?  I'd suggest anything reliant on the successful actions of more than a dozen units acting independently is overly complex (and likely to fail).  I like 12 as it's divisible in a range of ways.  So there you go.

Furthermore, if you look at commercially successful  rules, I'd say 90% fall into this bracket. 40K? Warmachine? Flames of War? Infinity? W-Wing? Check, check, check.

Conclusion: 5-12 'units' is the best size for a wargame, no matter what the command level it is at. 

Gees, you wasted a lot of ink discussing something that seems rather obvious....
If  I only raise the awareness of this, I've accomplished my objective.  The question:
"How many units do I expect to see on the table?" is an important one when you design your game.

I think it ties in closely with the questions "what command level is this game aimed at" and  "am I choosing to give players the ability to micromanage far beyond the historical norm?" - which is perhaps grist for another post. 

**This is why I have gravitated towards skirmish over "big battle."  Friends used to play Warhammer Fantasy. They would lovingly paint "trays" of say, 20 Skaven each.  Their army might have 8 trays of Skaven - 160 models to paint. In game terms, they merely had 8 units.  The 160 Skaven were simply glorified hit markers.  Whereas in Mordheim, you could paint 8 Skaven and have 8 units.   Skirmish gaming provided twenty times less painting for the same amount of units.


  1. I have to concur with your conclusion, though I'm not sure I could have articulated the reasoning behind it as well!

    I do think your final point of "what command level is this game aimed at" is a key aspect of what KIND of units they are and what their capabilities are. In your Aeronef example, you could have capital ships are individuals, and Cruisers and escorts working in Squadrons (of differing sizes) - that will give a different level of game the if each is treated as an individual ship. Its the difference between replaying Dogger Bank and Jutland.

    I did, however, participate in an awesome 30 player game recently where every person had their own ship and fought it individually according to their Admiral's orders. Wonderful experience. Its the last game detailed in the post here:

  2. "In your Aeronef example, you could have capital ships are individuals, and Cruisers and escorts working in Squadrons (of differing sizes) - that will give a different level of game the if each is treated as an individual ship. Its the difference between replaying Dogger Bank and Jutland."

    Yes, as part of my "escorts matter" campaign I decided not to make escorts as simply hitpoints within a collective "division/squadron" (like GQ does) as I felt it was "demeaning." A design aim is that a 1v1 between two escorts is just as interesting and deep as a 1v1 between battleships.

    (Also, back in the era of dodgy comms there's lots of examples of destroyers/cruisers getting separated during the battle and 'doing their own thing' or failing to report contacts/skirmishes to base)

    1. I dont disagree, but the mechanisms which give a 1 v 1 duel the detailed fun are usually different to those for fleet actions. Hard to play the interesting battle squadron actions without some sort of abstraction for the escorts, otherwise you overload the player

    2. Don't disagree with the 1:1 vs fleet differentiation (i.e. Fighting Sail is a good example of this). I'm aiming for "Battlefleet Gothic" territory (multiple squadrons - say 2-4 battleships, 2-4 cruisers, 4-8 escorts)

      I have considered a simpler set of rules using 4 hitboxes and a single salvo roll (a la GQ1-2), but I'm more interested in the middle ground.

      Though it depends on what you view as a "fleet." I'd say in modern naval terms, 12 combat-capable warships would be a goodly "fleet" (it's probably the entirety of the Australian Navy!).

      I'd presume aeronefs would be much rarer than conventional warships due to the unobtanium needed to manufacture and run them. So (in my thinking) a country's "battle fleet" might be 20+ battleships, but only a few squadrons of big battleship nefs.

  3. There's a reason the bigger FiveCore games assume 10-12 units ;-)

    Enough to have some spares and to make actual deployments and movements, but small enough that you can handle all of them easily (and be cheap on the wallet).

  4. I think it depends on the cognitive load that you're planning on giving to the players. I think it's a standard technical writing saw that most people can only handle about 7-9 things at once, so if you're going to go over that, then you either need to 'chunk' it down to manageable pieces, either in terms of game elements or decisions-points.

  5. An interesting discussion as always!

    Oh, and I have some of those APCs too...

  6. Your timing is excellent. Your more in-depth analysis of the genre-blind aspects of command and control as a player are spot-on.

    I touched on just this sort of idea with my full summary review of Neil Thomas' book, One Hour Wargames. ( At first blush it looks a lit less complicated than similar games, such as the DBA/HotT rules. Look closer and you see that most DBA games combine stand-types into maneuver groups where you're really only making decisions on 3-8 units. In OHW that's all handwaved away by making the implicit 3-6 units an explicit part of the rules.

    One thing I would add - actually, rereading this post prior to hitting "Publish", I realized this is an extension of Nurglitch's comment - is that the number of units isn't so much a limit to what a player can handle as it is one of the complexity variables to consider. I think the other two are time available and rule complexity. Given two rulesets with the same number of units you can play the simpler set much faster than the more complex one. ALternatively, given two games with the same complexity, the one with fewer units will play faster.

  7. "Given two rulesets with the same number of units you can play the simpler set much faster than the more complex one. ALternatively, given two games with the same complexity, the one with fewer units will play faster."

    This is a given.

    However as I noted, I'm not focussing on "amount of units vs time." Because that depends on how the rules were written and is impossible to "prove." Also the amount of time available (and what is acceptable) can vary from person to person.

    What I'm trying to focus on is the "amount of units vs the tactical choices."

    Too few units = too few choices, whereas there is diminishing returns when you have lots - i.e. 16-20 units does not provide more choices than 8-12.

  8. So what happens when the unit count is way over 10-12 (for example LOTR) but player usually keep most models together in one tight formation, rendering most decision making obsolete for most of the game. I mean, of course, once combat starts happening you regain the choice, but usually players have like 1 or 2 big formations moving around until they meet the enemy.

    In skirmish fantasy/ancient/medieval games I find it's pretty hard to find a balance between battle formations, choices, separate units and all that. Maybe it's time for some kind of hybrid games, with some troops ordered in small units (like link-teams in Infinity for example) and some others acting independently. Skirmish games are usually the most interesting, but the balance between having too much units and too little to make it tactical is sometimes really hard, at least for me.

    On a side note, I've been noticing many Warhammer (either 40K or Fantasy) players having less and less units in their games. When I played (years ago, during 6th edition), we usually had 7-9 units in a 1500 point game. Now everyone has 4-6 in a 2000 points game, it's ridiculous.

    Good points through the whole article, I really liked it. I missed your posts, they're food for thought, specially for people wanting to design mechanics and games. The Game Design series are the ones I enjoy the most.

    1. LOTR is in a bit of a weird place as it potentially can have 30 individual units or a single 30-man blob.

      It's actually succeeded in something many games struggle with i.e. the ability to shift between independent and small group movement. Ivan Sorenson has been deliberately experimenting with it in his later games i.e. Clash on the Fringe. I.e. making working in a unit desirable in MOST situations without artificially "forcing" everyone into a 2" coherency range. (I think there's a blog post on that particular mechanic somewhere here)

    2. I recall reading something on the topic, something related to Lords & Servants. I don't have that rulebook though, so I wouldn't know for sure.

      And yes, that usually happens in LOTR, and the system actually handles it pretty well. But sometimes the game just kind of forces you into blobbing, because of the benefits of pikes/spears, and the risks of getting your archers overrun by cavalry, making the game blob too much, which takes away much of the choices you're talking about in the article.

      I guess what I would love is a skirmish game that plays with around 30 models, but one in which an objective based map like 'treasure hunt' or whatever doesn't turn into "I'll send 28 guys to the objective in the middle, 2 to the one on my side of the board and just disregard the one on his side".

      Maybe having models acting in units, but with independent targeting and actions, they could just share morale and activation, and remain somewhat close to each other (8 in/20 cm)

    3. I think you are adding in a second factor i.e. "missions vs deathmatch" - a perennial problem in PC gaming as well as wargaming - as it's always easier to accomplish your objectives when all your opponents are dead.

      I'm not sure if this has been explored properly on this blog - I think it was touched upon


      but maybe not properly explored? It's on my holiday 'to do' list anyways...

      "Maybe having models acting in units, but with independent targeting and actions, they could just share morale and activation, and remain somewhat close to each other (8 in/20 cm)"

      Clash on The Fringe does something like this.

    4. Yeah, I'd love to find a solution for that. Maybe limiting the turns and having low lethality for ranged and melee combat, but that could probably make the game too slow or sluggish for a lot of people. Having an objective be 3x or 5x more valuable than killing a target could be another option. But the best solution I see is to carefully craft each mission so that players HAVE to actually move tactically around the map to have a chance.

      I'll check out those articles and I'll see if I can have a look at Clash on the Fringe, but I'm low on cash right now so I'll probably have to wait some time to get it


      More discussion on the topic, not neccessarily concrete answers

  9. Interesting article. Your conclusions are similar to mine, through from different initial assumptions.
    While learning programming, I was told of a "magic number" - the most items a typical person can hold in short-term memory is between 5 and 7.
    More than that, we either lost track, or group some of them (in programming we use a data structure).

    Modern Armies (about the time of the High explosive shell and magazine rifle) found the deadly battlefield required dispersion, and the battalions learned to manoeuvre by increasingly smaller subdivisions.

    The old commanders adage, "command one level down, know your dispositions 2 levels down" remained as true as ever, but the poor old Colonel found himself overworked.
    Instead of ensuring his battalion was in line of column, he had to monitor 6 or 8 companies and 18 to 32 platoons.

    Some reorganisation saw most armies adopt a rule of 4 (4 companies of 4 platoons), but after practical experience, most dropped to a heirachy of threes (with the odd 4 thrown in at different levels to provide a reserve).

    So the battalion has 3 companies of 3 platoons each, and a weapons company for 10.
    The brigade commander commands 3 battalions and monitors 9 companies...

    12 is my top number, as it enables a few reinforcements (Heavy Cavalry, Tanks, your choice...).

    It's little coincidence that the most common on-field team size for team sports is 11.

    1. Regiment: Colonel, 2 Battalions
      Battalion: Lieutenant Colonel, 2 Grand Divisions
      Grand Division (Equalized Infantry Squadron): Major, 2 Divisions
      Division (Equalized Infantry Troop): Captain, 2 Subdivisions
      Subdivision (Equalized Infantry Platoon): Lieutenant, 2 Sections
      Section: Sergeant, 2 Subsections
      Subsection (Equalized Infantry Squad): Corporal, Chosen Man, 2 Teams
      Team: 2 Buddies
      Buddy: 2 Privates