Asynchronous Combat Abilities Usage

[IndiraLightfoot]

The problem I find is two-fold.

 

First, each character finishes their "turn" asynchronously.

Second, each character has "abilities" that require player input.

 

What you end up with is a battle that takes 20s in-game but about 20 minutes IRL due to having to pause every 0.8s for each character's next instruction.

 

It's even worse than this. I've been thinking a bit more on this, and then I realized that you have a potential party of six members plus summons and pets, and all of their "turns" are indeed ending asynchronously. So, the ideal pausing for maximum player input would be almost in every instant. Moreover, enemies have similar asynchronous countdowns going on, It's a mess impossible to control, even with extreme amounts of pausing. Without a forced flurry/round/turn rhythm for every combatant's actions, preferably visible and clearly shown during combat for the player to make informed decisions by, there will never be a combat mode with full strategy in PoE. This is bad news for those that pause often or those who'd love to use slomo mode, and it's even worse news for those like me, who prefer a more flowing RTwP combat.

Edited September 11, 2014 by IndiraLightfoot

[nipsen]

*sigh*

 

...so with an artificial turn of, say, 2s - instead of pausing only when you really need to, you would pause, at least, every 2s.

 

Success?

[Sensuki]

Having each actor on their own unique time-sequence with the added potential for each to be altered by interruption and movement, the exact nature of the problem is difficult to discern. The experience is reminiscent of solving multivariable calculus. If I were to guess, I would wager that the problem is with the standardized cool-down and use speed of spells/abilities. I think that they too will need some variation--likely based on spell level and attribute scores rather than weapon type. To reiterate, I'm not sure. I'm just wondering if this is a problem for others, and what their thoughts might be.

This is going to be one of my future topics.

 

I have some ideas about how to make the system feel better. Part of the answer will actually be coming up in our Attribute System paper. But I have more ideas as well.

 

I do not see units as having turns, the system is more akin to something like Warcraft 3 if anything, but the recovery times between actions are much longer.

 

Currently I think recovery times are too long, the amount of actions a unit can perform in a given time is actually slower than all of the Infinity Engine games at their ideal pace of combat (which is between level 7-14 IMO).

 

Everyone needs to get the hell out of a turn-based or round-based mindset though. This system should be a proper real-time system IMO, not something where everyone is thinking in time divisions.

Edited September 11, 2014 by Sensuki

[IndiraLightfoot]

*sigh*

 

...so with an artificial turn of, say, 2s - instead of pausing only when you really need to, you would pause, at least, every 2s.

 

Success?

 

At least it's considerably better.

 

You know that auto-pause function from CRPGs of yore - "Pause at end of turn"?

It would get its comeback with everything in sync.

If you would use a similar function as the system is now, you would get 6+pets and summons, spread out unevenly, plus enemies on different cycles too.

 

Which is easiest on the player, do you think?

[Sensuki]

Warcraft 3:

Each action has an animation, the animation is split into two parts by a 'hit frame' which in WC3/DotA terms is called a damage point, or cast point (for abilities). Both the action before and after the hit frame can be cancelled by the stop button (and for normal attacks, movement also cancels). Following the action, there is a 'base attack time' (exactly the same as recovery time in PE) which is the delay in between actions which has a base of 1.7 seconds modified by IAS (Instant Attack Speed). In Warcraft 3, movement does not pause recovery time. Spells don't use Base Attack Time though as they're balanced by cooldowns.

 

Some units have a different base attack time than 1.7, but not many.

Pillars of Eternity

Each action has an animation, the animation is split into two parts by a 'hit frame'. When a character is performing an action, you can cancel the action before the hit frame by moving that character with RMB (exactly like in the Infinity Engine games). After the hit frame, you cannot cancel the 'backswing' and the full animation plays. Following the action, there is 'recovery time' as you know, is the delay in between actions, which is the same length as the animation performed modified by armor/shields/buffs/debuffs. For Pillars of Eternity movement pauses recovery time.

I think movement should not pause recovery time in Pillars of Eternity, as that's stupid although I understand why it was done - to balance ranged weapons. Movement should slow recovery for bows and crossbows and be treated as a reload so the recovery is not cancellable by switching weapons, and arbalests and guns should be as normal.

Edited September 11, 2014 by Sensuki

[IndiraLightfoot]

Sensuki: Some of that I already knew, but not all of it. Cool stuff!

 

Perhaps this is a blunt question: But why did Josh & Co opt for a system fit for a Moba game, where you control one character in RT only, and then apply this on six party members in a RTwP? Are there any benefits from this at all?

Edited September 11, 2014 by IndiraLightfoot

[Sensuki]

Probably because most real time games either use a system where every action either does or does not have a pause between animations - that's the only difference here. It's not something specific to MOBAs (which are all based off Warcraft 3 btw). Fighting games was actually where this originated. Stuff like Street Fighter.

 

edit: and in Warcraft 3 you control like 12 units at a time or something (you can only have that many selected at once). Many of them have up to four abilities.

 

Not many RTS players here?

 

edit2: Actually there hasn't been a good RTS game since Company of Heroes (I don't count Starcraft 2), so that wouldn't be surprising.

Edited September 11, 2014 by Sensuki

[IndiraLightfoot]

But the big difference is no-pause, no? If W3 had 12 units with 4 abilities, that would run something like COH2, which I know well? You're fully dependent on visual input, and the game's not pausing, so you have to keep track of their animations, literally, even looking for windows of opportunity for attack and defence.

 

With pause, it's a different beast altogether. You basically freeze the individual, out-of-sync turns of like a dozen units at one moment in time, so regardless of when you press pause, they will be on different places in their turns - some early, some mid, some at the end. This really makes my decision as a player much more arbitrary, unless there's auto-pause functions drowning me all the time.

Edited September 11, 2014 by IndiraLightfoot

[Sensuki]

You're fully dependent on visual input, and the game's not pausing, so you have to keep track of their animations, literally, even looking for windows of opportunity for attack and defence.

 

DING, there you go - you've identified one of the problems Pillars of Eternity has

 

THIS is the main reason people are finding it confusing, recovery time needs a different stance/animation to the combat idle, so that people can VISUALLY see when their character's recovery time ends, due to the longer waits in between actions.

 

In RTS games units act in concession pretty quickly.

 

If anything pause should make it easier, but the HIGH WTF DAMAGE combined with lots of active/ability use required by classes makes it into a pause fest. The incoming damage is why people pause so often. You can't play as relaxed because if you just select all, left click the beetles - what happens?

 

Your characters die very quickly.

 

Damage input in Pillars of Eternity is very normalized, and Accuracy ratings for everyone is generally higher than Defense scores, so the ranges are usually hits or grazes every action. The IE games had a high-ish chance to miss (particularly enemies) due to your characters high AC, so incoming damage was infrequent at best.

 

There's various ways they can go here, and this will be a topic of one of my super threads eventually. I would like to examine the various avenues to make PE play more like the IE games.

Edited September 11, 2014 by Sensuki

[IndiraLightfoot]

Thanks. I get it. PoE is thus a RT computer game with pause, more than a turn-based PnP RPG turned into RTwP.

 

EDIT: This is no sucking up to you, but I repeat what another poster said. I'm looking forward to those more than the next OE update. Can't wait!

Edited September 11, 2014 by IndiraLightfoot

[nipsen]

 

*sigh*

 

...so with an artificial turn of, say, 2s - instead of pausing only when you really need to, you would pause, at least, every 2s.

 

Success?

 

At least it's considerably better.

 

You know that auto-pause function from CRPGs of yore - "Pause at end of turn"?

It would get its comeback with everything in sync.

If you would use a similar function as the system is now, you would get 6+pets and summons, spread out unevenly, plus enemies on different cycles too.

 

Which is easiest on the player, do you think?

 

 

Like I've been harping on... I don't know everyone else thinks.

 

What I do know is that I pause the action less in PoE than I would in BG, NWN and IWD2. ..The weird thing is that I pause less than I did in Kotor and ME1 as well.. 

 

And I think it's because I only pause when I think I should counter something unexpected, or when I want to place things strategically. So I pause at the start of a battle, assign an action to each character.

 

And then I only change or add extra actions if I think I need to. There's no "hmm, what should I do this turn". I just set up a formation, put the characters where I want them to be. Add some spells and change the lineup to counter something the enemy did, which monster charged what player. And that's that.

 

So what happens in PoE is I see some action or other take place, and I respond to that. I'm frozen and webbed - so I need to up the defense, perhaps. Or choose to take the initiative. These guys are weak against fire, so fire it is. Ooh, here's an opportunity. Three pauses. Instead of constantly fighting to cram in another ability or action every turn.

 

I don't know. To me, it seems that to people who have to constantly feed another action in at the instant the previous one times out -- you would have rounds in BG and IWD that would last yonks anyway. For even just a two-member party, it'd take forever at level 10.

 

So I really don't understand the argument you're making. For example, on how a structured turn of 2-3 seconds could be easier to deal with. Since from my perspective, what you're proposing is another type of micromanagement - that punishes reactivity by setting you back 2-3 seconds, if you don't pre-empt the enemy every turn. I guess.. with overwhelming force and overkill spelldamage.

[Sensuki]

Thanks. I get it. PoE is thus a RT computer game with pause, more than a turn-based PnP RPG turned into RTwP.

 

EDIT: This is no sucking up to you, but I repeat what another poster said. I'm looking forward to those more than the next OE update. Can't wait!

You might find a nice treat in your PM box.

Edited September 11, 2014 by Sensuki

[Namutree]

 

 

 

Not many RTS players here?

 

I'm an RTS player, but I just started checking out this thread now.

 

 

 

edit2: Actually there hasn't been a good RTS game since Company of Heroes (I don't count Starcraft 2), so that wouldn't be surprising.

I liked Halo Wars; it's kinda like the Mario Kart the RTS genre. Like Mario Kart it's simple and fun, but doesn't have nearly enough depth for serious players. It's fun in small doses with friends who are only casual gamers though.

 

Anyway, I think the 6 second round system of the IE games was underrated. Sure it seemed kinda arbitrary, but it made things easier to keep track of.

[Sensuki]

For me, Company of Heroes (2006) was the last good RTS game. Starcraft 2 could be called a good RTS game, but it's an embarrassment IMO compared to Starcraft: Brood War.

[Namutree]

For me, Company of Heroes (2006) was the last good RTS game. Starcraft 2 could be called a good RTS game, but it's an embarrassment IMO compared to Starcraft: Brood War.

I agree. There was no need to defend your not counting Starcraft 2 position. I understood immediately.

 

You know what? I'm gonna start an RTS thread in the Way-off topic section.

[IndiraLightfoot]

Monte Carlo on these very forums was the one who lured me into COH2 (Thank you, MC).

 

He's a COH veteran. So, if I had been around when it was released, I'd surely have loved that version to bits too.

 

Sensuki: Thank you.

 

I have read a section of Sensuki and Matt's hard graft, and I can tell you all - there will be scientific findings backing up suggestions of some very welcome changes to PoE's combat system and its attribute mechanics.

Edited September 11, 2014 by IndiraLightfoot

[Mr. Magniloquent]

I made a quick attempt at organizing spells & abilities by rank. Essentially what you're looking at here is:

 

(1 + Ability Level) * 0.5 * ( Cast Speed + Recovery Speed) + Armor Penalty

 

I used a base casting speed of 0.66 and base recovery speed of 0.66. This is almost arbitrary, but I began on the assumption that casting the lowliest of spells/abilities would be equivalent if not on par with the speed and recovery of the fastest weapon. I then multiplied this value by the spell/ability level and added the armor penalty (if any). I made a linear assumption about the degree of time to activate and recover from increasingly powerful spells/abilities. This linear scale was also used for simplicity of illustration--particularly in observing the effect of armor. Since spells are defined by levels one through ten, I used this as my frame-work. I realize that abilities would have to be categorized to their equivalent level. Here is the chart. I included a table, just so the values would be clear.
 

 
To me, I feel that it all is very tidy. Simple spells cast quickly--particularly when wearing clothing, whilst other armors add proportionate encumbrance. This is particularly evident in the higher tiers of spells and abilities. What is significant about this data set, is that it would significantly improve the incongruity and discord between timing spells. This way, casting time is proportionate the power of the ability being used. Weak, frequently used abilities--and even intermediate abilities have more permissive activation and recovery speeds to be utilized more tactically. Likewise, earth-shattering spells and profound abilities would take time and require advanced planned and defense to execute.
 
Looking at the impact of armor, the heavier armors significantly impact the recovery over clothing (naked), effectively trading-off flexibility, readiness, and output in favor of defensive bonuses. To me, the penalty feels worthy without precluding the use of armor. Being linear, the amount of exertion (recovery) scales with the degree/potency of the ability. I am strongly in favor of allowing simultaneous movement and recovery, but I feel that this could work even if not.
 
The addition of talents to improve the efficiency/recovery time with armor or specific abilities could also allow a degree of nuance and customization. Example: Knockdown. Currently, for a fighter in plate armor, it takes 3 seconds to activate and 6 seconds to recover. Using this chart, that would place it between a tier 6 and 7 ability. Selecting the talent "Improved Knockdown" could lower its equivalent Tier by 3 steps (arbitrary). This would lower the activation and recovery to between tiers 3 and 4--a significant improvement in efficiency! This would improve its tactical timing use, as well as rewarding specialization with improved recovery times.
 
Obviously, this would have to be balanced out. Since everything requires significant balancing already, I don't see that as a detriment. The argument for scaling the activation and recovery speed of spells and abilities is identical to why utilizing a dagger should be different than utilizing a great sword. This chart is just a stepping stone to illustrate the strength of this model and how it could drastically improve the egregious problem of tactically utilizing spells and abilities.

 

*Edit: To Clarify this chart, Brigandine & Plate Armor currently have the same penalty, so their plot overlaps. This can be seen in the data set below the cart.

Edited September 11, 2014 by Mr. Magniloquent

[Sensuki]

Cool idea, although the total 'round time' for earlier levels is definitely too short, whereas the higher ones is definitely too long. You also need to take into account what type of action the ability or spell is, and what the base cast time is [PE has 3 - short (1s?), medium (3s), long(6s?)]

I'm not 100% sure that splitting up how the system works for weapons and spells/abilities is a good idea .... buuuuuut .... that is how most other games handle it.

Still it's an idea anyway, not really one I'd get behind immediately, but it's an avenue to think about.

[prodigydancer]

So, the ideal pausing for maximum player input would be almost in every instant.

 

That's how you play RTwP.

[Mr. Magniloquent]

Cool idea, although the total 'round time' for earlier levels is definitely too short, whereas the higher ones is definitely too long. You also need to take into account what type of action the ability or spell is, and what the base cast time is [PE has 3 - short (1s?), medium (3s), long(6s?)]

 

I'm not 100% sure that splitting up how the system works for weapons and spells/abilities is a good idea .... buuuuuut .... that is how most other games handle it.

 

Still it's an idea anyway, not really one I'd get behind immediately, but it's an avenue to think about.

 

The values are not important. The key is that they synergize and interact with the base mechanics of attacking/recovering better. Having all weapons swing and recover at the same speed is a poor idea. Having all abilities and spells, some of which (conceptually) should be very divergent (lvl 1 vs lvl 10 spells), is both horrible in concept and practice. Any abilities/spells at either end of the spectrum will be gimped with a one-size-fits-all.

 

Here is a simplified chart more akin to Mr. Sawyer's Attack & Recovery illustration. It shows activation time as well as activation plus recovery times in naked, lightly armored, and heavily armored states. The general framework is kept more closely to the current weapons pictorial so that equivalencies can be better understood. I used the current cast time to approximate the upper limit. The using a casting speed of 0.2 and recovery rate equivalent to the cast speed, this general formula is:

 

(1 + (Cast Speed * Spell Level) * 2) + Armor Penalty

 

 

This leaves abilities and spells to be comparable to being naked with two light weapons when naked and using low level abilities, while using the most powerful abilities and spells (lvl 10 equivalent) would operate identically to the current cast speed and recovery. Again, the point is the convey that spells and abilities need scaling activation and recovery times--not get caught up in minutiae that are subject to change on their own right.

[Sensuki]

I don't know if they do need spell and recovery times that scale with level. In general, Pillars of Eternity doesn't have that much stuff that scales with level. 

There's Accuracy, Defenses, HP, Constitution bonus ... that's about it really?

 

Otherwise it's fairly static.

 

I'm not saying it's a bad idea, I'm just saying that I don't think it would be something they're interested in.

[Mr. Magniloquent]

I'm very confident that the one-size-fits-all approach is why spells and abilities;ergo, combat is/are so unwieldy. It's confusing as to distinguish between a Hatchet and a Battle Axe whilst considering every single ability and spell in the game equivalent. No person would genuinely contend that a stilleto and a arquebus should have the same attack speeds and recovery, so why abilities?

 

Something as mundane and classic as Knockdown cannot even function as an effective tool to interrupt or control movement. How on earth are you supposed to sabotage ANY action when it is going to take 3 seconds (likely several seconds greater) just to invoke it? Also, with a base duration of 5 seconds, good luck having your heavily armored tank actually capitalized on the prone opponent.

 

Consider the spell Thrust of Tattered Veils. It's a "quick spell" meant to interrupt something. What that something is, I have no idea, because it's the slowest activating ability in the game. It cannot even plausibly interrupt another ability, because you would first have to recognize an ability is being used, hope that your wizard was standing completely idle, and then hope that your spell finishes quickly enough before your target to reach them before they execute their spell. That literally cannot happen. Even if every contrived and improbable setup existed, you would still have to cast your "interruption spell" before they attempted to use their ability because they have identical activation times plus travel time.

 

How successful has anyone actually been with negating wood-beetle poison? Or Deep Wounds for that matter? If you manage to prevent a K.O., what percentage of their character portrait isn't red? Etcetera, etcetera. This is a major problem. Treating all spells and abilities as though they identical in form and function with respect to usage is ridiculous. Given the speed of combat and that all classes have abilities--many of which are highly situational and time-sensitive, I would consider this one of the most paramount things to address given the universally perceived chaos of combat.

[Matt516]

I'm pretty sure that at least some combat abilities are instant or close to it.

 

I could be wrong though.

[Mr. Magniloquent]

I'm pretty sure that at least some combat abilities are instant or close to it.

 

I could be wrong though.

 

That gave me pause, and I took some time to more diligently test many abilities. It's difficult to determine precisely, but certain abilities do appear to have different activation times. The Priest per encounter abilities appear to be the quickest; casting very close to or at an instantaneous speed. Knockdown is also quicker than previously thought, but again it's difficult to time the exact activation. I have had less success timing the druid shape-shift to say one way or the other. The rogue ability, Crippling Strike, appears to go off instantaneously if invoked as the first action of combat, but has otherwise taken time once combat is engaged (even if idle). I'll have to test these out some more. Despite that, a spell, is a spell, is a spell--and I neither agree with nor enjoy that convention. I feel like my observations and criticisms still stand for those.

[Sensuki]

pretty sure the cast times are 1s, 3s and 6s (we haven't seen any 6s spells yet).