Jump to content

My Pillars Combat Analysis-Review

Recommended Posts

I'm cross posting this from the codex... just for fun :)
Having near finished Pillars for the second time, I wanted to stoke the fire a bit about what Pillars means in regards to the evolution of rtwp. As far as I know, Pillars is the only serious attempt at tactical party based RTWP combat since the old IE games, though I'm sure somebody will disabuse me of that fact. (Actually, I'd be interested in knowing what's out there that's good that I've missed) 
I don't see myself ever belonging to the cult of perennial BG2 pilgrims... so excuse me for being a bit irreverent, but I'd like to compare a bit of what Pillars brings to rtwp in terms of the evolution of the rtwp combat. I'm also a fan of 4e dnd combat as a turn based tabletop system. IMHO, Sawyerism could be pretty much defined as a conversion of 4e to crpg, adding funky health mechanics and equipment using DR. I wholeheartedly support this effort. I play 5e now for the greater roleplay experience, but it really has nothing on the concise tactical combat of 4e. Oh, and for all those complaints that Pillars isn't a D20 system... just divide your accuracy and deflection by 5 and truncate. My +n bonuses in pillars looks real similar to my 4e character progression.
Since I was aware of the large amount of extra XP available in game, my first playthrough skipped Raedric and most of Od Nua. It was on hard. My second playthrough was POTD with expert mode, and focused on gaining xp via bounties, Raedric and Od Nua. 
Both playthroughs have been done avoiding degenerate tank and spank playstyle and over-levelling. I'm going to give a somewhat critical overview of what positive things POE brings to rtwp as compared to the older games, having avoided the generally declaimed pitfalls experienced by a lot of players. 
So first off, I enjoyed combat in Pillars at least as much as I did in both IWD's and BG2. In terms of the mechanics of combat, I think Pillars is a richer, more tactical evolution of rtwp. The classes in general all bring some interesting unique active mechanics that require tactical consideration during a fight. To a certain extent, I agree that Pillars suffers from some lack of varied encounter design, decent pathing and AI and clunky engagement mechanics. However, I think I appreciated the encounters more then most who either were OP for most of them or Tank'n'Spanked all the way through. 
So without further ado, where I think Pillars innovates on old-school IE rtwp, without any ordering whatsoever:
#1 consistent, understandable action/reload/recover timing mechanics that correspond with the animation reality in game. ( Obviously the big caveat here is engagement attacks, crappy vfx) 
One of the aspects of Pillars combat I enjoy is synchronizing buffs/debuffs with other powers that exploit them. Being able to understand the timings helps to do this and feels good when these combos work out. Having the animation correspond to the actual in-game mechanics helps to target and get visual feedback in game. It's nice to be able to read about weapon speeds, armor recovery damping, then actually be able to see what that means through visual feedback. 
#2 Real-time opponent feedback. 
Discovering the defensive weakness of enemies and also seeing the debuff/buff durations/impact during combat enables informed tactical play. This could be even better if there were more unique enemies not residing in the bestiary that required in-combat discovery. However, the bestiary was very useful for informing combat tactics before a fight, moreso then as an XP source, and I thought it was well done. Perhaps we could have the bestiary as the *average* stats of a monster, but get deviations, like a particularly large xaurip etc.. 
#3 Existence of repositioning powers. 
I really enjoyed using the push spells, place switching powers, etc. This is remniscent of the tactical positioning available in 4e tabletop. If anything, pillars needs more of this, particularly some cool pulls and transporting items/scrolls. I'd like to be able to force transport that skeleton wizard into the middle of my party and bash him down... I'd also like to see that as an AI tactic.
#4 Damage being a combination of to-hit defenses with varied DR ratings.
Optimizing damage relies alot on understanding and exploiting your ability to debuff an appropriate defense in order to hit an enemy with an attack on its lowest DR rating. Again, this requires understanding the powers /weapons at your command, and then tactically applying them to the battle at hand. 
#5 Multiple limited active powers to choose for each class.
limited resource, whether through per/encounter/rest or more fancy monk and cipher mechanics requires thought and planning... and strategy. For example, I often had my cipher cast pain link on my monk, who was propped up with priest heals to take a ton of damage. 
#6 Health/Endurance Mechanic
Having players get knocked out introduced a whole new stage of combat to me... When you're trying to punch above your level, many times fights come down to a last couple characters vs a few enemies. The loss of a character and their accompanying skills means you have to adapt, and I really like it. In IE games, I hated having a character go down, even with resurrection, it meant I had to pick up all their crap and have resurrection scrolls/spell slots available. (not to mention unresurrectable races.. yechh). Generally it meant a restart from save.
However, there are funky times when you run into a place where you don't want to heal a character, or would prefer to have them knocked out. This is especially raw in POTD where if you heal that last bit of endurance up to the bar on your wizard in a tough fight, you know he's going to die. I think the quick fix for this is to incur an immediate max endurance penalty to health when your character is knocked out. In this case, converting those last bits of endurance when death is on the line becomes really important. 
Now, onto a more critical point of view, hopefully with some useful suggestions:
Engagement is horrible. I understand that it comes from the desire to convert the notion of adjacency with opportunity attacks into rtwp. Originally I thought it could add something to the game, since opportunity attacks provided some interesting tactics in 4e, but it instead detracts as a weird hard agro mechanic that supports tank n spank. Also, opportunity attacks are impossible to tactically use because engagement attacks generally only go off when weird ai/pathing stuff goes on, because some ai has lost a target and starts moving. You cannot play the game with the engagement ai turned off, because you proc attacks just by going to close to attack an enemy. Engagement is passive and pretty much too overbalanced, making tanking too easy and too available and making most movement in a group a mess of random engagement attacks/weird pathing problems.
Engagement is bad because detecting adjacency is ill-posed in continous 2d environment as opposed to discrete tile based combat. I'm convinced that making an equivalent useful measure that's tactically enriching in 2d is actually harder then getting a good pathfinding algorithm. Let's do pathfinding first eh? 
The simplest solution, in my mind, is to award engagement only when a character has succesfully procced a melee attack (hit or no) against an enemy in range. No multiple engagements, just a single engagement line saying, I'm melee and in range and it's going to hurt you if you run away. You can get an opportunity attack if your currently engaged target leaves range. This covers the "anti-kiting purpose" of engagement, I believe. Thus you choose who you get an opportunity attack on by "targeting". This is well defined in rtwp. I'd suggest that even forced movement should give up an opportunity attack in this case. (where's the fear mechanic that forces an enemy to run away?), leading to combos you can take with other character powers.
As for losing the aggro mechanics, this could be replaced by a 4e tactic called "marking" where the tank has a power that causes the target to lose a large amount of accuracy attacking any other target other then the tank itself. Since this power could be limited, you'd have to make tactical considerations, which would make combat *better* and limit some tank n spank. AI should naturally switch targets based on this info ( I think they would as of 1.04), and even if they don't, it will provide some back line support.
Movement damping of recovery:
Yet another mechanic that doesn't come from the point of interestic tactics, but instead penalizing many things for the sake of one (kiting). This needs to go, or be modified. If you really want to discourage kiting, here's what I think should happen.
When a creature is targeted, it's recovery should be slowed if its attacker is moving towards it.
When a creature is targeted, and it's running away from it's attacker, its recovery should be slowed.
When a creature targets another, and the target is running away, the creature should have its recovery speed and action time increased.
When a creature moves towards its target, the creature should have its recovery speed and action time increased.
All of these mechanics should be increased/decreased by the difference of the resolve of the two participants. This could allow an interesting mechanic, where if your melee character is ready to strike by the time he gets to his target, he can have a charge bonus or something. Tanks could protect the back line by running towards mobs that have slipped through... 
let's *incentivize* movement. 
Weapon balancing:
There's a bit of redundancy that happens with the DR system, the fact that each weapon does at least %20ish percent of damage means that large weapons have generally better bypass then light weapons, even those with bonus like stilleto. I understand this is somewhat balanced with action speed, but given that damage modifiers all benefit heavy weapons more then light weapons, there needs to be a change in mechanics.
1. every weapon should have a DT rating and it should be fixed, no percent min damage.
2. Critical hits should be 2*DT (or 3*DT) instead of bonus damage ( okay, I'd be fine with keeping critical hit damage bonus).
3. perception should increase accuracy.
Encounter Balancing:
4e was interesting in that there was a lot of powerful DM tools for making varied and cool encounters with lots of tactical variety. Can we port this over? With a combat system this interesting, you can get a ton of ways to make cool encounters. 
In general, there seemed to be a swarm element to most of combat encounters, I'm assuming this was necessary to overcome engagement.
There was very few encounters where the target was one or two elites or a terrible solo creature, generally they were accompanied by reams of trash mobs and supporting characters. That being said, the final fight was pretty good example of a few elite characters versus the party. Thaos needed some better stats before buffing himself. 
Hard Counters and Immunities:
I wasn't a big fan of these in BG2, but I think there's enough of a crafting/scroll/magic system in POE that it could be done without making a single class GODS. I'd like to see multiple immunities to different types of attacks and conditions, with a variety of counters, probably spell scrolls that require some interesting, dangerous to find components, and perhaps some high level counter spells. Maybe some soul engine that crafts arcane veils for fighters. Really, the 1.05 arcane veil is near enough a hard counter, it will be cool when/if firearms actually become its counter. Encounters based on various immunities would require more variety and thought. There was some of this in Pillars, where an enemy had a high freeze DR, but we need more then one immunity to really influence gameplay. 
Yes, this has been beat to death. My chime: quality of equipment should be inherent. Higher qualities and unique weapons should support more, or more powerful enchantments. There should be a few, hard to get items in the game that are meaningful, and which make combat tactics dependent on bringing them to bear on your enemies.
Ranged Accuracy dependent on clear line of sight:
This might be over-reaching a bit, but one way to avoid tank n spank would be to give accuracy penalties to ranged attacks that clip through your allies. No friendly fire, just optimal targetting requires shooting lanes, which can open up your back lines to more 
fun ... 
Pathing Problems:
Why is this such a huge problem? I understand pathing is hard, but has there really been so little rtwp since BG2 that we can't get something better then this since early 2000's ? I kind of get the feeling pathing in Pillars is a horrible mishmash of Unity pathing/3d weirdness. It would be cool to get some developer feedback on what the issues are. I'd be interested in joining an effort to support the incline by hacking together some good 2d pathing and ai for rtwp. My brain cycles are on short supply currently, but a nice slim c,c++ rtwp code on github with BSD style licensing could probably help indys and bighouse dev alike.
Thanks for reading... eviscerate at your pleasure.


  • Like 5
Link to comment
Share on other sites

A good review and a good read. From what I've seen of patch 1.04 I hope that more AI tweeks are in store for us. But a good read, one that shines the light on not only its flaws but also what it did right. Sometimes it seems I'm seeing the spotlight on its flaws and the absence of its good even if they enjoyed the game.

Link to comment
Share on other sites


1. What you see as a turn vs RTwP issue, I see as a UI issue instead. Generally, turn-based uses tiling or similar stuff to clearly show "engagement" range, but just imagine a turn-based without such clear indicators. Divinity: OS doesn't have an AoO mechanic but otherwise fits; imagine that game with no added visual indicators for AoOs, but with AoOs nontheless.


Now imagine PoE where the red circles around enemies in the UI expanded to clearly indicate engagement range.

2. Kiting was simply hated on. Engagement is overtuned with not just free attacks, but free attacks with more damage and better interrupt. While these things may be fine for specialists (for example, a tank Talent to buff disengagement damage), making them standard was ridiculous.


Sawyer wanted kiting impossible and got his foolish wish: movement is weak in combat.


Weapon balancing:

Disagree with just about everything you wrote. Accuracy from an attribute would be boringly mandatory for all characters. DR penetration has no need to exist when you have the ability to alter your damage type, therefore it shouldn't.


That said, armors should have more variable DRs, and enemies should have less all-around DR (perhaps more base health instead). The bias against fast weapons seems to be the result of monster defense design, rather than inherent to the system.


Soft counters would be good.


No strong feels regarding OP's other points.

Edited by scrotiemcb
  • Like 1
Link to comment
Share on other sites

Yep, I noticed kiting when I played as well. I also noticed another effect like a seesaw. This was pretty easy to do with having one of your characters move in and out all the time, and the enemy couldn't get to any of my characters while the rest of my characters could shoot from range and kill the enemy.







  • Like 1
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Create New...