I am aware that Pillars of Eternity is a spiritual successor to Baldur's Gate (which was Real-Time-With-Pause), and that many people hold strong personal preference for TB or RTWP, with most of Obsidian's and Black Isle's games being the latter. This thread is not about about that.
The topic I'd like to discuss is the technical advantages of TB combat for games with fairly complex combat mechanics made on a comparatively low budget (kickstarter).
(TLDR at the bottom)
Throughout this post, I'll be comparing PioE and Divinity: Original Sin for the following reasons:
1) Both are critically acclaimed Kickstarted party-based RPGs with relatively modest budgets, funded at roughly the same time and released within a year of each other.
2) Both have non-trivial combat mechanics and a heavy focus on combat.
3) I've done completionist playthroughs of both games within the last year so they are still fresh in memory.
Based on my experience with the two games, I think PioE would have worked better with turn-based combat, and here is why:
1) It's hard to argue that poor pathfinding is a huge problem with PioE, with enemies often lining up behind each other even in open spaces. It is difficult to implement pathfinding that reliably finds a way around obstacles and environmental hazards. It's even more difficult to make it quick and cheap enough to perform for dozens of characters at the same time. Games with 5-10x PioE's budget routinely struggle with this and/or resort to cheating (disabling clipping, teleporting characters, etc).
Having Turn-Based combat makes it easier to implement accurate pathfinding. You only need to run the algorithm for exactly one character at any time, no matter the size of the encounter. As a result of having far more resources and time to work with, the pathfinding algorithm can be written to not just draw a line from A to B, but also to account for other variables on the way (avoid engagement, detected traps, environmental hazards, etc).
I have rarely seen enemies fail to find their way around a battlefield in DivOS, even though its environments are much more geometrically complex than PioE.
2) Poor (or non-existent) AI is also a big, if not the biggest problem with PioE's encounters. Enemies beeline for the closest and/or weakest character and engage him for the entirety of the fight. They are easily tricked, tend to ignore environmental hazards and make single-minded decisions - nonsensical behaviour like squishy enemy casters running into your frontline to cast a spell at your backline are a routine occurrence.
Turn-based combat makes it easier to script "better" AI for the same reason it allows better pathfinding - the AI scripts need to run for only one enemy at any given time. As a result, each individual enemy can take his time to determine the best course of action - decide to take a disengagement hit to slaughter a nearby squishie, move out of a Wall of Fire, reposition to attack from a side or behind, and so on.
As a result of being given ample time to "think", in DivOS enemies often make sound tactical decisions on the fly: high-health enemies deciding to swallow an Attack of Opportunity to pick a better target; enemies repositioning for better firing/spellcasting lines and moving out of your range before the turn ends; swarmer enemies splitting across several targets to inflict Flanked on all of them; the list goes on. Beating AI enemies that show some semblance of intelligence is so much more satisfying than outsmarting dumb targets that only dwarf you in numbers and stats.
3) PioE has to resort to arbitrary and severely contrived solutions for problems that TB games avoid naturally through game mechanics. Here are a few examples:
I) PioE arbitrarily forbids some abilities to be used outside of combat and allows others. This is an ugly band-aid, but necessary to avoid excessive pre-buffing.
DivOS gets around this by giving buffs that normally last several turns very short durations in real time - every second of the buff outside combat may be several turns inside combat. The game is consistent in its seconds-to-turns conversion across the board, so pre-buffing is balanced by the natural casting and recovery animations taking away from the active buffs' durations.
II) PioE's hard engagement system and player movement penalties feel like band-aid solutions to restrict the player's movement. Coupled with unpredictable pathfinding, breakneck combat speed and restrictive nature of most environments, most encounters allow you to maneuver very little, if at all.
In a TB game, only one character moves at any time, so you have full control over the pathfinding and engagement. You can carefully maneuver around enemy engagement zones and control exactly where your character ends his turn, allowing for a much greater mobility and wider range of tactical opportunities. Since movement costs action points, you cannot recklessly run for the enemy backline, lest you want to end your turn surrounded by enemies.
All of the above combined gives designers more freedom in designing abilities and encounters. DivOS can get away with open-field encounters with big enemy counts, with friendly fire on all projectiles, with abilities that create large obstacles and block attacks, with small-scale encounters against intelligent enemies and much more - all because Turn-Based combat allows both the player and AI enough time to evaluate the encounter and plan intelligent moves.
While Obsidian is obviously going to stick with RTWP for the expansion and sequels, I think PioE would have worked better with Turn-Based combat, and I hope they consider returning to Turn-Based in future RPGs. What do you think?
TLDR: Turn-based combat makes it easier to implement accurate pathfinding and "smart" AI on a small budget. Weak pathfinding and AI severely constricts ability and encounter design. I think PioE would have been a better game if it were turn-based.