Hi

Im doing a game similar to xcom (both new and old dos versions), fallout tactics, hard west etc. You build bases, train and recruit a squad of soldiers and do missions in turn-based fashion. The setting is present-day civil war in a fictional country in africa.

First i designed it with time units: each soldier have different amount of TUs depending on stats, all weapon reload, and different firing mode cost different amount of TU. Movement cost TU depending on lenght of the move (like old XCOM and fallout tactics).

However, this was mainly for nostalgia reasons (i understand now). So im planning to change it to a more distinct /simplified version with a couple of "actions" per soldier instead (like most games do nowadays). The TU-system is cool and gives a lot of old-school simulation feeling. But it slows down the game alot and forces the player to calculate alot (or be frustrated becouse ending up with 14 TU but needing 15 TU to actually fire/do something else).

1. What are your general input about TUs vs APT (actions per turn)?

2. I will most likely change to APT. Standard these days is 2 APT, which works like this:
a. Move short and shoot

b. Just shoot (typically no benifit so you might as well move as well)

c. Move long (double move) or use special equipment.

What would happen if each soldier has 3 APT instead of 2? I want some more details than most games, such as different firing modes. A draft:

1 AP: move (for max X tiles depending on stats). You can also move double or triple lenght by spending more AP.

1 AP: quick actions like snap shot or reload pistol

2 AP: slow actions like aimed shot, burst-fire, reload main weapon

3 AP: especially slow actions like fire rocket launcher or use some special equipment like radio (can only use without moving)

This would maybe give more flexibility. You can move a short distance and then take 2 snap shots or 1 aimed shot. You can reload and move out of the way etc.

There might be rules that says "max 2 snap shots per turn" or whatever is needed for balance.

Thoughts? Or just makes it more complicated? To go further and give 4 APT seems to much to me.

Thanks

Erik

I'd start from the feel you want and work backwards from there. Is it "shock and awe" troops pouring in with guns blazing? Or is it careful positioning, sneaking along a wall, outflanking your opponent? Is it mopping up the entire enemy force, or surgical strikes on key targets against a much larger force you can't beat in a firefight? Do I keep my squad close together to maximize firepower, or do we spread out to outflank the enemy and cover each other? Is it a heavily armored knock-down drag-out shoot out, or is a misstep out into the fire-zone instantly fatal?

I'd also make the action choice in conjunction with other tactical decisions. Is sight symmetric or asymmetric (can my forces stand safely behind a corner and pop out to take shots, or if I can hit somebody can they hit me?) Is there automatic return fire if I get into an opponents line of sight on my turn? (and if so, is that conditioned on them having saved up actions? Does the attacker or the defender get the first shot?)

Regarding TU calculation complexity, I think a good interface goes a long way to relieving that. When you mouse over a tile, it should give you an indication of whether you'll be able to do a full attack, partial attack or no attack from there. I'd even allow the player to right click to simulate ending the turn there: if I stop there, can I get to the cover next turn and still have a partial attack left?

The other thing that can help TU calculation is to have fewer hard cutoffs. Rather than 10 TU for a snap shot and 20 TU for an aimed shot, have a shot take 5-25 TU and the quality vary continuously. So it's always better to have more TU, but + or - one isn't the end of the world: Getting in position with 9 TU means you're probably missing, but you were probably missing at 10 TU anyways.

But if i want to skip timeunits and go for APT (actions per turn):

How do you guys think 2 vs 3 APT would affect gameplay? Any preferences? Any examples of games that uses 3? (all I know use only 2).

The TU-system is cool and gives a lot of old-school simulation feeling. But it slows down the game alot and forces the player to calculate alot (or be frustrated becouse ending up with 14 TU but needing 15 TU to actually fire/do something else).

This is a problem only with a rigid turn system where everybody gets e.g. 20 TU per turn and that soldier has to delay the 15 TU action by one whole turn instead of the very short time (1/20th of a turn) needed to earn 1 TU.

Do you want or need simultaneous actions or a fancy priority/initiative/surprise system to determine who acts first "at the same time"? If you don't, you can simply have time steps spaced by 1 TU, maintain (and probably display) a queue of all entities who can act on that time step, and process them in order, putting processed entities into the queue for the timestep when their action ends. No need to count available TU points because all entities can begin any action at every turn; long actions just finish later. For example someone beginning a 5 TU action on timestep 52 acts again on timestep 57.

Actions should be reduced to the smallest possible units (e.g. walk one map square, one serious attack in melee combat, one shot or burst with a gun) but timesteps should also be very small to allow nuanced speed differences (e.g. a high quality automatic pistol shoots every 19 timesteps, a clunky revolver every 23 timesteps, plus optional aiming actions before changing targets).

But if i want to skip timeunits and go for APT (actions per turn):

How do you guys think 2 vs 3 APT would affect gameplay? Any preferences? Any examples of games that uses 3? (all I know use only 2).

The big difference there is move-shoot vs. move-shoot-move: Can you get in position for a good shot, and then retreat to cover?

Having multiple shoot options (fire twice or take better aim) also encourages a slower sense of pace, at least to me. You often burn 2/3 of your action points so that if you turn a corner, you've got enough left over to take careful aim.

Personally I hate the reduced number of actions in the new xcom reboot (I'm assuming that's what you're going for?), because if I want to peek a corner, shoot, than take cover it uses almost my entire turn. I liked how in the older xcom's with lategame soldiers you can do a very elaborate movement that changes based on what you find. For example, when breaking into an alien ship, you can peek corners, see enemies, throw down mines, exit the alien ship, and plant some C4 on a wall near them.

@Lorenzo

Isnt that kind of semi-realtime? It seems hard for the player to understand and plan what will happen, and in what order.

@polama

Well you can have a rule that prohibits movement after attacking. So its move-quick attack-quick attack OR move-slow attack.

But 3 instead of 2 of course makes it a LITTLE bit more complicated. But maybe thats good? I think the new XCOMs are a bit too simplified.

@conquestor

Yeah I understand the wish to fine-tune your turns, but couldnt some of this be elliviated by having 3 actions instead of 2? Also I personally dont like that you can retreat after firing; its too easy to exploit the AI and be safe (peak around corners and back, be cowardly). Which is not so fun gameplay (i used it myself alot in old XCOM becouse i felt forced to do so in hard situations).

Large numbers of action points (or time units), feel very old fashioned to me now. It feels clunky, and requires too much extra calculating, like you described. I think the change from the old X-Com AP to the 2 actions in XCOM gives you almost everything good and fun from the original game. Any limitations are greatly outweighed by the benefits of simplifying the actions.

That doesn't mean you need to do 2 actions though. 3 or so actions will probably work well. I think the important thing is that each action point is used deliberately, and feels like a useful thing on it's own. I enjoy the 8 or so action points in invisible inc, where each action is one tile of movement. But having 8 to 10 action points in the original fallout games feels a bit pointless, with guns that often use 5 action points. With this system, points often felt wasted. Sure, you could get more agility and look for a weapon that uses a factor of your total action points to allow you an extra attack, but it wasn't really a fun system.

For my turn based tactical game, I plan on using 2 actions, with a subtle twist so it feels unique. Like how XCOM ends your turn when you attack, so you can't attack and then move, my game will have a limitation to how actions are combined, so you might attack and move, aim and attack, move and defend, but not do two actions of the same type.

A big question is: How many entities do you want active at a given time, and what is the ratio of friendlies to enemies?

If you have maybe 5 allies against 5 enemies (at a given time, possibly more than that on a map/mission, but positioned in bite sized chunks) then it may be practical to run 'Tactical Time', a sort of Realtime-auto-pause where each turn is a 'tick', and different actions require a different number of ticks to complete, but gameplay is in a "We-Plan-We-Move" manner.

So Tick 0: Your crew are all spawned beside your transport unit, and you set orders that they all run for cover a ways away, which generates an order queue for all of them. The enemies are all in 'sentry mode' and aren't making active plans, but may be moving through patrol routes.

You click 'Go', and your troops start running through their order queue. They move 'on tick' at a time, and at the end of each tick the game auto advances because nothing has changed in your state. Before everyone gets to the end of their order queue someone spots an enemy, and the auto turns are halted. You are given the choice to change orders, or hit go again and the turns keep ticking by till something changes (Such as another opponent spotted, or someone opens fire, steps on a mine, etc. Some trigger event that you wouldn't have expected to happen based on the info state when you last hit 'go'.)

Systems like that can produce a very smooth flowing and intense gameplay, and gives the player the choice between short term reactionary planning "What is around this corner and how am I going to deal with it?" and longer term strategy "I'm going to put the sniper here, covering this zone, then advance the front liners here and cover the machine gunner till he gets there..." as you use 'As Much or As Little' of a turn as you go, but you're not having to sit there hitting 'next turn... next turn... next turn' after you've planned something out, or having to move things an inch manually every time and try to remember what your grand plan was.

Also personally I would suggest, from a programming standpoint, to build your code in a flexible manner, such that you can easily switch between the different turn and point styles, and conduct actual play testing of the different methods to ensure you have a feel and flow that you're looking for.

Honestly, I've always thought turn based isometrics and the like are way too fucking slow, especially Xcom though I did enjoy it quite a bit with that one overhaul mod.

It's like... I play something like ARMA and you suddenly realize how quick and fast paced war can be. You go enter 2 gun or 3 gun competitions and come to find that shooting on the move is nowhere near as inaccurate nor as slow of a matter its made out to be like in CoD where you're suddenly a fucking tortious if you check your irons.

I never liked the serious lack of tactical abilities either. Moving and shooting, having units immediately switching to a pistol if they run out of ammo or their guns jam, the fact that some guns don't jam, the fact that Snipers have to spend 5 rounds to get set up and then never do anything because they have an arbitrarily short range or for some reason don't do any damage at all, the fundamental lack of reactionary moves like moving into cover and what not, the fact that suppressing fire does almost nothing and good god the unrealistic guns that made shotguns into squirt guns, Assault rifles into pea shooters, heavy mechs or weaponry capable of being shrugged off by high level enemy units and all the other disparity. Same reason why I couldn't really get into that many RTS games, because the group based and solitary unit based AI and tactics were so god damned generic with zero complexity, while the greater game is just fucking rock paper scissors which is why it baffles me when people tote Starcraft 2 as being this huge, complex, tacticool spass muhreen simulator, when it really just is a race to see who reaches their end game meta first.

If I were to do a spin on time units, I'd change it to where every unit has a bar, each action takes a certain % from it but if you compound certain actions, they essentially become free, IE if you run and gun, run and reload, shoot and reload immediately after, et cetera, with the player's basically setting up their units in rounds, IE you have unit A go here and plan for him to shoot Target 1, while unit B applies suppressive fire on Target 1; in turn, your opponent anticipates your moves and orders their unit's actions with the players sharing the same planning phase, after which when both sides lock in the actions of their units who mind you, get their own AP bars, the round plays out based on the AI priorities you write and pathing you commanded, all playing out in real time because each round that plays out is 1 minute of actual, real time combat with the % bar being the amount of time the actual units have to execute commands.

So it balances itself out pretty damn well. Small units, who will have reactionary abilities will be able to create an organic game field and compensate for missteps, small units having the ability to crunch time will force players to think in a more tactical sense with unit arming, squad composition, unit placement, cover and the like, heavy units are indeed heavy, tanks can't just roll in and blast people to shit because they're standing in the middle of a field like a dumbass, but they are more limited despite them being a force multiplier, IE it takes some time to position them, move them around and in certain areas they're far from being less than ideal making them somewhat situational, and it can empower the player by giving them a visceral feeling of having significant control over the flow of battle like RTS games have that TBS games don't since it is somewhat real time, it ends up playing in more of a real time.

The only two issues I can see with this is that firstly, It'l be complex as all hell. It's a mindset most people don't work or use in games but then again, same can be said about Red Orchestra 2's and ARMA's realism. Second is that coding the unit and opponent's AI will be a fucking clusterfuck of a hell's worth of coding, planning and nerfing to prevent the computer from starting on fire and/or going full statistics and either pulling brilliant moves or being dumber than a sack of sand potatoes.