No announcement yet.

Abstracting the Squad as a Unit

  • Filter
  • Time
  • Show
Clear All
new posts

  • Abstracting the Squad as a Unit

    I recently got to thinking "if this were an RTS game, how would I design a 'squad' unit such that it behaved roughly like the squads in PS2?" This sounds like a useful abstraction for understanding squad dynamics and the flow of battle from an SL or PL position, and to help analyze strengths and weaknesses of my own squads and those around me. I've puzzled out what I think is a workable abstraction and I wanted feedback from you guys!

    Squad Design

    To start with, our squad unit is a blob instead of collection of units. The squad has a number of attributes:

    S - Unit Strength:

    Unit strength is a measure of how many individuals within the squad are currently combat effective. It's not merely how many troops you've got in the area, but whether they're able to unify their effort to completing a task. A scattered or uncooperative squad has low unit strength. As a result, this is strongly influenced by squad cohesion, individual teamwork, and effective leadership.

    D - Squad Defense:

    This is a measure of how difficult it is to render a single individual within the squad ineffective (that is, reduce the squad's unit strength), such as by killing, wounding, or otherwise neutralizing their ability to do their jobs. Having good cover, more durable units like MAXes or Tanks, and player skill at minimizing incoming damage contribute to this. This will vary substantially as the tactical situation changes.

    R - Squad Recovery:

    This is a measure of how quickly the squad returns to full strength after suffering attrition. This includes revives, respawns, healing, repairs, pulling fresh vehicles, and so on. This is not a measure of how quickly the squad reforms off the front lines after getting wiped, but how quickly they can recover in their current position. If a soldier can spawn at a nearby Sundy and walk over in a short time, it counts, but if they can't get back to the original fight in a timely fashion it does not.

    This scales with unit strength. A unit that's suffered a lot of attrition (or was never cohesive to begin with) will recover slower than a close knit, mutually supportive team. Player skill (particularily medic skill), squad composition, and proximity to respawn options are also important factors.

    A - Squad Attack:

    This is a measure of how quickly you can knock down the unit strength of an opposing squad. Basically, how many guns you can point at them simultaneously multiplied by how big those guns are. It's highly variable by weapon type, target type, range, and individual skill, but scales strongly with overall unit strength.

    You can make analogies to RTS units in a pretty straightforward way with these properties: unit strength is HP, defense is armour value, recovery is passive health regeneration, and attack is... well, attack. They interact a bit differently than your typical RTS, but it's a decent shorthand.

    There are other factors to take into consideration as well, like overall mobility, but I'm glossing over them for now.

    Squad Interaction

    My conceptualization of two opposing squads interacting with one another is the idea that they start to inflict "pressure" on one another as they come into contact. This is basically the following formula:

    P1 = A1 * S1 / D2 - R2 * S2

    Or, to put it in plain English: the pressure delivered by squad 1 is how quickly they can do damage minus how quickly the enemy can recover from that damage. If pressure is positive, the enemy will begin to lose unit strength. If negative, they’ll be able to start recovering unit strength.

    This has a few implications:
    • If your unit strength is lower, your recovery and attack capacity suffer
    • If one squad has full unit strength against a squad with low unit strength and the pressure is enough to start damaging unit strength, over time the disadvantaged squad will be put at a greater and greater disadvantage.
    • Unit strength is the most important factor in an engagement. Once lost it is difficult to regain without breaking contact.
    • Any opportunity to attack the enemy’s unit strength is valuable. Either watch for opportunities to take advantage of low defense (squad on the move, caught in the open, etc) or endeavour to create them. Likewise, anytime an opponent has poor recovery options or you have an opportunity to remove some of those options, you can make them more vulnerable to attrition.
    • A combination of high defense and high recovery will allow a squad to stay at full unit strength against bad odds.
    • A high offense will allow a squad to beat opposing squads into ineffectiveness quickly and nullify the threat they pose.
    • A squad with high defense but low recovery (like a tank squadron) is vulnerable to attrition over an extended engagement, and so either needs to keep fights short or keep enemy pressure low.
    • A squad with low unit strength will want to break contact to recover.
    • Breaking contact with a squad that has low unit strength will relieve the pressure on them and allow them to recover unit strength.
    • If two squads are not applying positive pressure to one another, the fight is completely indecisive and both will remain at full strength.
    • Since offense increases as range decreases, squads that approach one another will create greater pressures on one another, increasing the degree of pressure and potentially creating a decisive result.
    • If opposing squads manage to deplete one another without achieving a decisive advantage, their offense will have dropped and the battle will be likely to draw into a stalemate.

    Squad Archetypes

    To help illustrate this, I've got a few common squad archetypes I'd try to slot into this model.
    • The Infantry Squad - Your bog standard, balanced infantry squad. Its defense is highly dependent on positioning, recovery is very high, but attack capability is low. An infantry squad needs to get close to apply decisive damage, but in the right position is difficult to unroot. Generally speaking have low defense and offense while on the move (corollary: applying pressure is a strong incentive for the enemy squad to stay put).
    • The MAX Squad - A squad centered around MAX units, with heavier engineer support. It has high defense-MAXes are tanky-but medium recovery due to the additional effort required to keep them topped up and on their feet. Their offense is a bit higher than an infantry squad, but the MAX squad is largely depending on its ability to tank damage to win fights quickly in close quarters where they can be dealt with decisively before attrition becomes a major factor. (corollary: weather the storm and wear them down)
    • The Tank Squadron - A squad centered around armour units. The tank squad has very high defense and offense, but absolutely rotten recovery due to the added overhead to repair a damaged tank or the timers, travel time, and resource cost of replacing lost vehicles. As a result, they excel best when engaged in low pressure fights (long range support) where attrition never exceeds their low recovery, or in decisive blitz attacks that use their high offense and defense to end the fight before attrition becomes a factor. (corollary: apply enough pressure to exceed their recovery abilities and make yourself difficult enough to attack that their offensive power doesn't come into play)
    • The Air Squadron - A squad centered around air support. They have poor defense, poor recovery (for the same reason as tanks), and high offense. They largely rely on their mobility to avoid being pressured by the enemy and to deliver burst fire on exposed targets. Their ability to keep fights short and still have an effect on the fight is their primary strength (corollary: if you apply pressure to them, they can't stick around long enough to apply their offensive abilities)

    So, that's basically where I'm at right now. It lines up roughly with my own intuitions about how various unit types work, and I'm interested if it lines up with anyone else's. I get the feeling that this can be extended further to be more useful/accurate. I also think you could do a very similar level of abstraction for regional level forces to help analyze strategic concerns.


  • #2
    Re: Abstracting the Squad as a Unit

    I would note for MAX squad both defense and offense are highly position dependent, like infantry, because of MAX's short range weapons generally.

    I think this is an interesting concept. It requires serious leadership to mold a group of players into one of these abstract units, but it is an interesting conceptualization for certain. :)
    The question foremost in my mind is "what will bring the most tactical fun to the server?"


    • #3
      Re: Abstracting the Squad as a Unit

      I'd argue that, as defined here, MAXes aren't as position dependent for defense as infantry. Obviously, good cover helps them, but their speed, bulk and usage pattern usually has them tanking fire rather than avoiding it. As for offense, it's ALWAYS positionally dependent and varies by a huge number of factors by definition. You could break offense down into a squads effectiveness versus different types of units at different ranges (basically a custom function for each based on squad loadout).

      Turning this into something functional instead of just a thought exercise will take some work. Without a doubt trying to actually put hard numbers into this equation would be foolhardy. However, my hope is that you could use intuitive values for the different parameters ("this squad is in a good position so their defense is high, but they're loaded out for long range so their offense is weak against this close enemy") and then the relation between those values in the formula could inform decisions about whether the position is tenable, how to make it tenable, etc. It's painfully inexact and devoid of nuance, but something like this helps clarify my own thoughts, at least.


      • #4
        Re: Abstracting the Squad as a Unit

        This is incredibly cool and abstractions like these are fun mental exercises. I think your initial variables and formula are valid but I'd like to spend some time myself approaching this concept.

        I'd probably ask for you to define the parameters of each variable. Are these integers, functions themselves, or are the fractions? I'll be back with a more I depth analysis tomorrow.


        • #5
          Re: Abstracting the Squad as a Unit

          My initial conception of those variables is something like the following. All values are Real/floating point functions and they'll vary over the course of the engagement.
          • Unit Strength roughly corresponds to number of functional, responsive units in the squad. A suppressed but unharmed member represents a lower strength than an unsuppressed member (1.0), but higher than a dead member (0.0)
          • Defense would be best represented as a value from 0.0-1.0, though in my initial formula I used division because I felt like that would communicate the intent clearer (so if you switched to a 0.0 to 1.0 setup, you'd flip that around to multiplication). Basically, 0 represents the enemy weaponry having full effect on you with no mitigating factors, while 1 would represent something on your end (like a spawn shield) making you completely invulnerable. Intermediate results represent factors like having a larger HP pool (having twice the HP of a normal infantryman would be something like a .5), cover, dodging shots as they come in, etc.
          • Regeneration would be framed as "number of people you can get to full fighting strength in X seconds", or some other time step that makes sense. Not just revive time, but the time for you to get to them, them to get reoriented and then finally start shooting again, which is a bit higher. This is framed as regeneration at full strength, since as unit strength drops the value will also drop.
          • Attack is the reverse of regeneration: "at this range, against this target with your current kit, how many units can you knock out of the fight per second, assuming they don't have mitigating factors (defense)?" It should largely be agnostic relative to the target save for the basic "is this the right weapon for this target", since the target's qualities are mostly defined under their defense. Instead, it should be the elements of the attack that correspond only with the squad itself. This is framed as attack at full strength, since it will scale with unit strength later.


          • #6
            Re: Abstracting the Squad as a Unit

            Strength could probably be best explained as a set [sm1, sm2, etc.]

            Doing it as such would actually allow you to grant specialization modifiers to each squad member.


            • #7
              Re: Abstracting the Squad as a Unit

              I think that complicates it too much, given that you're never going to be able to calculate every squad member's setup in real time. Instead I favour averaging out the specialized bonuses of individual squad members across the entire squad. That's why my summary of a MAX squad has higher defense and low regeneration, even though non-MAX members of the squad don't have that durability bonus and regenerate just as fast as normal.

              My goal here is a simplified mental model that you can reason about in the middle of battle with incomplete and fuzzy information, and so so in a timely fashion. The simpler the model can be made while still saying useful things about the battle in progress, the more valuable the theory is, IMO.


              • #8
                Re: Abstracting the Squad as a Unit

                I was thinking about this thread a little bit, trying to add something productive, and came up with the conclusion that this would be best suited for a Top Down observation on a squad from some higher element. In this case I was thinking this mental exercise could help streamline the decision making process a Platoon Leader exercises.

                He checks the squad's relative strength, he checks their relative cohesion, their location, and situation. This then allows him to quickly project their capabilities now and in a few times passing. Considering that time tables are important this streamlining process could be critical in a Platoon leader's attempt to either advance against a 'retreating' enemy or tell himself to shore up his defenses and bide some time.

                If we are to keep this as simple as possible then I think any form of algebraic representation is too much. Rather just simply reduce it to a game of arithmetic (like RISK is). Instead of being worried about equations lets focus on 'relative power' and inequalities.

                For example

                For example your Squad 1 has an assumed strength of 4 (I'm picking arbitrary numbers) versus infantry. It is meeting resistance of a n enemy with an assumed strength of 12 (thanks to a Rebirth Room). 4<12 therefor 12 wins. Knowing this the Platoon leader then allows Squad 1 to re-cooperate to an effective strength of 6 while also bringing Squad 2 and 3 with effective strengths of 8 each. Individually they are incapable of taking on the enemy but when assigned certain important positions by the PL they serve to break up the strength of the defenders.

                This leads to the favorable inequalities of:

                6 > 4, 8 >> 4, 8 >> 4

                where >> is a much greater than sign.

                This leads me to think that this exercise would be good on paper as a way too formulate potential tactics and implementations of specific squad groupings.


                • #9
                  Re: Abstracting the Squad as a Unit

                  Yeah, I was largely thinking about this from a platoon leader's perspective, though I think squad leaders can benefit from thinking about their squad in general terms like this as well. Consciously thinking about how strong your squad is now and how strong it will be in the immediate future should lead the squad lead to take steps to optimize those values. Maybe something like "the squad leader tries to make these values as nice as possible, but the platoon leader is the one reasoning about them to determine overall strategy."

                  I don't think the algebraic representation is a problem because you should NEVER be actually calculating anything. Instead, you should be internalizing the relationship between the factors (which are themselves very fuzzy) so that you can make a gut judgement about pressures and unit strength over time. It's a heuristic for aiding judgement, ideally.

                  The equation itself is irrelevant and useless, but an algebraic equation is a succinct way to describe the relationships between different factors. I suspect that this is something that is unintuitive for people without mathematical or computational training... I know that I was thrown for a loop when my third-year class on AI described certain algorithms as a single equation instead of a sequence of steps. If we can find a better, easier to communicate formulation for the same relationships that the equation describes, the concept will likely be more useful for more people.




                  TeamSpeak 3 Server


                  Twitter Feed