Putting units or actions in the outer loop ...

Many of the discussions regarding rules design in miniature wargaming involve the turn sequence, and how to structure it in order to achieve some desired effects. Over the years, there have been many different variants of the turn sequence, illustrating various schools of thoughts on how to organize the game. Often, discussions are framed as IGO-UGO vs unit-activation, but the spectrum is more varied than that. See also my earlier post on discussing various ways of looking at the turn sequence.

In this post, I will offer a slightly different way of looking at the turn sequence, and how it can affect rules design.

Action-based turn sequences

In an action-based turn sequence, the turn is defined as a (strict) sequence of actions, and within each action, every units can do exactly that action. This is best illustrated with the classic move-shoot-melee-morale sequence. First, we have the movement phase, in which all units get to move. Then we have the shooting phase, in which all units can shoot. The same goes for the melee phase and the morale phase.

In computer programming terminology, one can see such a turn sequence as 2 nested loops . An outer loop loops over all actions, and an inner loop, for each action, loops over all units doing that action:

• for each action X (in fixed order move, shoot, melee, morale):
• for each eligible unit Y (order usually chosen by player):
• unit Y does action X

Such a turn sequence has a number of advantages - there is a fixed order for all actions, and this implies one can be sure that all movement has happened before shooting and melee, morale checks are performed after all combat etc. This can lead to a design in which those various phases of the turn are strongly coupled. However, it also often leads to rules in which all turn phases use separate mechanisms which have not much in common from a procedural point-of-view. The mechanisms for shooting and melee can be very different, and morale usually is a separate subsystem.

An important and crucial disadvantage is that actions which are not explicitly listed in the turn sequence cannot be added to the games engine in an elegant manner. Such actions are often defined in terms of an existing action. E.g. setting a building on fire might take half a move; or instead of fighting a unit might build (part of) a bridge. An alternative solution is to add such additional actions explicitly to the turn sequence. E.g. one can imagine having a separate "engineering phase", or "spell casting phase". The drawback is that such specialized actions become very visible in the outer loop, while relatively few units might be eligible to act on them. These actions might also be rarely used during the entire game, depending on the specific scenario. In a fantasy ruleset we used during the late 90s, there was a phase in the turn sequence called "Morph Friendly Cyclics". It always gave rise to the same joke in each and every game. "Hold on guys, we forgot the Morph Friendly Cyclics phase! Friendly cyclics anywhere? No friendly cyclics? Are you sure? Ok, now we can move on ... "

Unit-based turn sequences

In computer programming, 2 nested loops can often be interchanged. We can also do this in the turn sequence. Instead of defining the turn as a loop over all actions, we define it as a loop over all units:

• for each unit Y (order usually chosen by player, and/or driven by an activation mechanism):
• for each eligible action X (chosen from a list of possible actions):
• unit Y does action X

Such turn sequences are often called unit-based activation, since the units might be activated by die rolls, cards in hand etc. There might even be a terminating condition (e.g. a failed die roll), indicating the end of the turn. When a unit is activated, there usually is a choice of what actions, or a combination of actions (perhaps expending a budget of 'action points'), can be executed with that unit.

The main disadvantage of such a sequence is that it is difficult to maintain coherency between units. E.g. if one would like to maintain a battle line, and one cannot be sure that all units can move forward during this turn (e.g. if there is a possibility not all units might get activated), this can really frustrate player or result in ahistorical behaviour of the troops on the gaming table. Often, a mechanism for activating a group of units is necessary as part of the activation procedure.

On the other hand, there is a big advantage when one wants to add additional types of actions. Since actions do not drive the turn sequence at the top-level, adding actions does not change the turn structure. Rather, actions are an add-on to an already existing structure, which makes it easier to maintain coherent procedures and game mechanisms.

This does not prevent from actions having their own disparate procedures, but it does create a structure in which a coherent mechanism for resolving actions is more viable. One of the best examples is the Basic Roleplaying System by Chaosium. It exists in many different variations (the Call of Cthulhu rulesystem probably is the best-known). Looping over the characters is the outer loop (as in most roleplaying games), and each action - whether it is searching in a library or swimming across a pond - is resolved by checking a D100 vs a target number. That results in a very elegant and tight design, and adding new types of actions is rather straightforward.

The same structure can be used in miniature wargaming. Whether you want to add engineering actions, special movement or shooting actions, or actions invented for a specific scenario only, they can be fitted much easier in the overall game structure as opposed to an action-based turn sequence.

Last turn in the game

When playing a scenario, it is often convenient to set a number of turns within which one or both sides need to achieve their objectives. However, how many turns are appropriate for a given scenario? This clearly depends on the ruleset in use, but there are some rules of thumb I have used over the years, and a neat little mechanic to make the number of turns a variable factor in the game.

Number of turns

In my gaming group, we have a running joke that says that a given scenario ends "at the end of the wargaming day". This is a phrase often used by Charles Grant in his famous scenario books, without any further explanation. We always are/were confused whether he meant the end of the wargaming day for the players, or the end of the simulated wargaming day for the armies on the table. Hence, we often refer to it jokingly whenever someone asks about the end of the scenario. It's always "the end of the wargaming day".

However, that does not mean the length of a scenario cannot be set in advance. Clearly, this depends on a number of factors:

1. How far do troops move per turn, versus the projected total movement distance?
   A typical example is an attack/defence scenario. Suppose the attacker starts at his own baseline, and needs to take a number of objectives (hills, towns, ...) roughly situated at the defender's baseline. Let us further assume that the table is 120cm in width, and that an average move is 15cm (6") per turn. This then translates to 8 turns needed to cross the width of the table. Thus, if you put the scenario length to shorter than 8 turns, the defender will have a hard time - or an impossible time - of reaching the objective.
   However, troops will not move at full speed in a straight line. You need to give some allowance for the attacker to manoeuvre across and around the battlefield. Hence, I often multiply the minimum number of turns needed by 1.5, resulting in 12 turns for the scenario above.
2. Can troops move and fire, or only move during the turn?
   The above result of 12 turns assumes that units can move every single turn. But suppose your ruleset only allows movement or fire. Sure we want the attacker to fire at the defender as well, and any turn in which firing happens, movement will not happen for a firing unit. In the case of a static defender, the defender does not need to move and can fire every turn. In a typical attack-defence scenario using a move-or-fire ruleset, we might want to give the attacker at least the same firepower as the defender. The attacker therefore needs twice the amounts of units, half of which will/can  fire during a turn, while the other half is moving. Thus, if we want to give every unit the opportunity to reach the objective, we need to double the number of turns, resulting in 24 turns.
3. Do we have an activation mechanism?
   Some rulesets do not allow every unit to do something. A system using cards or command rolls might restrict the number of units that can be activated during the turn. Suppose any given unit will have a 50% probability of getting activated. That means we have to double the number of turns, in order for our unit to reach the objective. Doubling our 24 turns from above, we now have a 48-turn long scenario.

The calculations above are only here to illustrate the process. If you use different assumptions, or put different restrictions on how many units need to reach the objective or move over a certain total distance, the multipliers will vary, and you will end up with a different total number of turns.

Variable number of turns

Computing how many turns we need is one thing, but should that imply the number of turns is this fixed number?

Most gamers are probably familiar with the last-turn-all-out-attack syndrome. If a player knows it's the last turn, he will risk anything to reach the objectives, because he knows there's no risk for the enemy to take advantage of it if all goes wrong, since there are no more turns left.

Some boardgames therefore use a mechanic to determine the last turn in a random manner. Starting at some point during the game, a random number (die roll, card drawn, ...) determines whether this was the last turn or not. Many different variants exist. There is no reason why something similar cannot be used in a miniature wargame.

E.g. suppose we have set the number fo turns at 20. We want a spread of plus or minus 2 turns for the game to end. The game could end at turn 18, or it could go on till turn 22. We can easily design a mechanic that gives a 20% increment each turn to end the game:

• Turn 18: 20% chance the game ends (e.g. roll a D10, game ends on a roll of 1-2)
• Turn 19: 40% chance the game ends (roll a D10, game ends on a roll of 1-4)
• ...
• Turn 22:: game ends with 100% probability.

Whatever mechanic you use, be sure that the game does end!

Players control the number of turns

In my games, I often use a mechanic(*)  that allows players to take a certain amount of control over the length of the game. Suppose the scenario is set to last for 20 turns. During the game, whenever a players rolls an exceptional result during some procedure (e.g. a double-6 on a 2D6 for a command roll, or a random event, ... ), he has the opportunity to increase or decrease the number of turns by 1.

Thus, if a player feels he still needs time to reach his objectives, he will often add 1 to the number of remaining turns. If he feels he needs to speed things up, he might subtract a turn. Or he might choose to leave things as they are.

We have used this mechanic in various games with different rulesets, and it often provides a healthy dose of suspense. In some games, it doesn't do much, but in other games, the count has gone up or down a few times. And towards the end of the game, the situation can become really tense and provide a lot of amusement and unexpected outcomes. If the event of altering the turn count happens infrequent enough, players do not have the impression it's an uncontrollable random device, but rather a little extra resource they can use to their benefit.

I even have a special large D20 specifically for this gaming mechanic. The D20 indicates the remaining number of turns (thus, every turn, the die is reduced by 1), but when the special roll or special event occurs, the player can dramatically and with the right amount of pathos turn the die up or down by 1.

My oversized D20, next to a regular D20. Some 42mm toy soldiers for size comparison as well. Winston Churchill noddingly approves.

(*) I encountered this particular mechanic in the first issue of Battlegames magazine (2006), where it was attributed to the ruleset Pieces of Eight.

Some thoughts on the turn sequence

When discussing wargame mechanics, the turn sequence is often a subject of hot debate. In a sense, the turn sequence is the engine that drives the game forwards. Many other mechanics that deal with combat resolution or morale, are often embedded in the overarching turn sequence. The turn sequence regulates the alternating role of players, but also regulates the order in which units can act and how.

The classic IGO-UGO turn sequence

During the early days of wargaming, the turn sequence was rather simple, and is now often referred to as an “I Go, You Go” sequence (IGO-UGO). Both players take alternating turns, and within a player’s turn, the order of actions (or phases) is fixed as well. Typically, I move all my units, then I shoot with all my units, then I resolve close combat with all my units, then I resolve morale for all my units, and then you take your turn.

Let’s try to put this in a diagram. Let’s assume we have 2 sides (Red and Blue), and 3 units per side (A, B and C). A complete cycle starting with Red would mean that Red moves all of his units first, then shoots with all of his units, and so on. The fixed order in which all activities take place can then be schematically represented as follows:

	Movement	Firing	Melee	Morale
Red, unit A	1	2	3	4
Red, unit B
Red, unit C
Blue, unit A	5	6	7	8
Blue, unit B
Blue, Unit C

 
The matrix above lists all possible phases for all possible units, and the numbers indicate in which order they are executed. It is a very straightforward turn sequence, and one that is still present in many wargaming designs. It is also a traditional way to play classic boardgames, such as Monopoly, Snakes and Ladders, even Chess. They all follow a similar format:  move a piece, then do something with that piece. If you follow that framework, it is quite natural that you end up with this particular turn sequence. BTW, you also often see this same turn sequence appearing in board wargames. The Avalon Hill classic hex-and-counter wargames also use the framework “move all of your pieces, then do something with them.”

The main problem with this turn sequence is there is no possibility for the opponent to react, or to do something with his troops. The classic example is so-called overwatch fire. When I move my units from one covered position to the next, you cannot shoot at my troops, although they might be vulnerable during some  part of their movement path. Another often-cited problem is that I can have a unit outside of our mutual firing ranges, I move them within range, and I blast your unit to smithereens before you have the chance to do anything.

Solutions often come in two forms: re-order the various phases of the turn, and/or introduce more sub-phases. Re-ordering is the more elegant solution. The sequence might then involve that Red moves first, then Blue shoots, then Red shoots, then a joint melee phase occurs, etc. Another often-used solution is to put the firing phase before the movement phase, such that “move and shoot to smithereens before you can do something”-tactics are not possible.

Adding more sub-phases, sometimes depending on troop-types, becomes complicated very rapidly. It results in games in which you have various movement and firing phases for each side, sometimes interspersed. In my experience, these do not play fluently, since players often not capable remembering what troop type can do something in what phase.

Any turn sequence also has implications for the underlying mechanics of the various phases, and how these phases influence each other. E.g. the morale phase might specify conditions and events that happened in the previous combat phase. A strict ordering of phases makes such interactions more easy to deal with, but also has to allow for “out of order” actions that otherwise do not fit the overall structure. E.g. when charged, troops might be allowed to counter-charge or run away, although it is not strictly their movement phase. Similarly, movement reactions often are part of the morale phase as well.

Alternating unit activation

In this turn sequence, players alternatingly activate units. Red activates a unit, then Blue activates a unit, then Red activates another unit, etc., until all units have been activated. Each player chooses what unit to activate next, and some sort of bookkeeping is needed to remember what units already have been activated. During a unit’s activation, the unit can move, shoot, fight, etc, but not necessarily in a fixed order.

We can schematically represent this turn sequence as shown below, with Red activating unit B first, then Blue activating unit A, followed by Red activating unit A and so on.

	Movement	Firing	Melee	Morale
Red, unit A	3
Red, unit B	1
Red, unit C	5
Blue, unit A	2
Blue, unit B	6
Blue, unit C	4

As you can see, during a full cycle, all units still get to do all possible actions, but we have organized them in a different manner. If you are familiar with computer programming, you could see this as having two loops turned inside out. The classic sequence looped over all phases, and each unit got an action during each phase. This sequence loops over all units, and a unit gets to do all phases when it is selected. Another way to look at it is that we subdivided the matrix listing all possible actions for all possible units by rows instead of by columns.

This sequence has some consequences on the mechanics of the various phases. The mechanics of a phase cannot be strongly dependent on previous phases or the actions of other units. E.g. it becomes more difficult to have a morale phase that would take into account the actions or behaviour of nearby units.

This turn sequence is often more flexible in design than the classic sequence, since the phases themselves do not form the overarching structure of the game. Hence, it is easier to add new types of actions or phases that a unit can do. Suppose you would like to include an engineering activity in your game. The classic sequence might make this part of the movement phase – or should introduce a new "engineering phase" in the turn sequence. The alternating unit activation sequence can simply add a new type of activity that a unit might or might not do during its activation. It might seem like a subtle difference, but it works very well in e.g. roleplaying games in which each character takes a turn, and can then do a multitude of different actions available to the player.

Variants on alternating unit activation

To ensure an equal pacing of units being activated on both sides, variants often include that whatever sides has the most units left to activate, must activate the next unit. Group activations are another variant that can guarantee multi-unit coherency.

Another, more extreme variant stipulates that players cannot choose what units to activate. Often, this is implemented as some sort of draw (cards, chits, …) with each card specifying what specific unit can activate. This is not a very attractive mechanic, because it takes away important decisions that the player wishes to make. Moreover, in some scenario setups, it can clearly create bottlenecks, when e.g. a column of troops has to cross a bridge, and units simply refuse to be drawn in the correct order.

A hybrid is possible, by allowing a randomization to check what side can activate, but then leave the decision up to the player to determine what unit will activate. Such mechanisms also often include the early abortion of the entire cycle, such that neither player is certain that all his units get to be activated, and is forced to activate those units first he thinks are most important. Underlying mechanics to accomplish this often include special cards in a deck that drives the alternating activation sequence.

Unit-driven IGO-UGO sequence

This turn sequence tries to combine the best of both worlds, by using the alternating player structure of the  IGO-UGO sequence, but within each player’s turn using a unit activation mechanism. This is a turn structure that has become very popular in modern designs. During a player’s turn, the player can choose which units to activate and in what order. Often, dice control the activation sequence (command rolls), or a hand of cards might drive the player’s choices. An early abort mechanism often is included as well.

The following diagram illustrates this sequence, with units in grey not having activated due to an early abort, e.g. a failed die roll. First Red activates unit A, followed by B, and then fails to activate further units. Then it's Blue's turn, activating unit B.

	Movement	Firing	Melee	Morale
Red, unit A	2
Red, unit B	1
Red, unit C
Blue, unit A
Blue, unit B	3
Blue, unit C

The advantage is that the mechanics for the different phases (movement, shooting …) can still be interlinked, since we can assume that a large number of units will get activated within the player’s turn. Hence, it might be possible to sequence the movement unit by unit, but still keep an overall morale phase at the end or the beginning of the player's turn.

Random phase sequence

The advantage of using the matrix representation to illustrate the turn sequence is that you can subdivide the matrix in different ways (i.e. organizing the turn in rows or columns as shown before), but one can also re-arrange the columns in random order.

An unusual turn structure might therefore randomize the different phases over all players. Thus, we could first have Blue firing, then Red movement, followed by Red morale etc. As a mechanic, this can be easily achieved by making a custom card deck and drawing cards to see what next phase comes up. The diagram below illustrates a possible random sequence.

	Movement	Firing	Melee	Morale
Red, unit A	2	8	6	3
Red, unit B
Red, unit C
Blue, unit A	4	1	5	7
Blue, unit B
Blue, unit C

Such a sequence requires that all phases can be resolved independently, as explained before. I have only tried it once before, in a skirmish game in which many unexpected events might take place, and an emphasis is put on heroics rather than a coherent well-orchestrated battle plan.

Conclusion

There are of course many more turn sequences possible. All sorts of hybrid formats can be imagined. In the end, the turn sequence is interlinked with the underlying mechanics for the different phases and activities, since the entire gaming engine has to form a coherent whole.

As a games designer, it’s always useful to tinker with various ways in which the turn can be organized. Even if you end up with a classic IGO-UGO sequence, at least you thought about it and can defend your particular choice much better.