I'll try to keep this as brief as possible, and if I seem to go into too much details at times, it is because I have the maybe-too-ambitious goal to help the developers not just with suggestions for a current problem, but to propose them a better general approach to a more abstract problem. Ok, with that said, my main argument is that the problems we are all having with combat right now (v333) are due not to deficiencies in the game mechanics/ruleset, but to deficiencies in the User Interface.
PoE is to use a "real time with pause" system for combat1. The main problem with this system is the risk to overload the player with information2, which breaks his Observation-Orientation-Decision-Action loop and prevents him from choosing the optimum set of actions for his current situation. Time pressure isn't an intended feature of the combat mechanics of the game3.
Currently this is exactly what happens, the player is overloaded with information when playing out battles in realtime, which practically forces him to use the Slow time feature (press the S button at any time to toggle it on/off), and to pause the game manually at regular intervals and/or to use the autopause options. The problem I see (there may be more) with this playing style being forced onto the player (because of a UI deficiency), is that while it makes combat playable, the player receives only a piecemeal experience of it, and misses out the dynamics and the emotional impact provided by the beautiful graphics and by the sound effects4.
I see two main changes that could be made to remedy this problem (with the piecemeal experience of combat), and they are basically pointed at preventing the information overload, thus restoring a functioning OODA loop with a humanly-feasable realtime combat speed.
- Firstly, I propose that the animation speeds are cut to about half when in combat, how much precisely is a matter of trial and error, with the resulting about x2 increase in action cooldown periods across the board, for all weapons, spells and talents5.
- Second, the current halos above character heads provide a sort of a meter for the action cooldown of characters (the little horizontal lines), but I'm pretty sure these meters, although identical across characters, don't progress with the same speed for each character. This radically reduces their usability while the player is in paused mode, because he can't get an idea of how fast the lines are shortening for everyone. And if he unpauses in order to get an idea of who will be the next character to initiate action, it often may be too late for him to react. Additionally, trying to visually compare the length of lines is inconvenient, even if they were moving with the same speed.
The solution here is simply to provide a form of unambiguous visual feedback of which will be the next character who will be able to take action. This would give us a combat system which we may call "desynchronised turn-based" Off the top of my head, there are three ways in which this may happen. First, make all the cooldown indicators grow shorter with the same amount per second, but increase the on-screen length of the indicators. Then getting a hint of which character's "turn" would come first would be as simple as finding out who has the shortest line. Alternatively, you can make it so that when in paused mode, the section of the cooldown indicator that will shorten will blink. Then when paused, the player will check which of the characters has all his remaining cooldown indicator blinking, and know that this is the character who will next be able to preform an action. The third approach is I think the best - simply have the game make the calculation for the player and write the order somewhere in the halo. I've provided a modified screenshot illustrating that. One problem that always exists with this solution is that if one of the AI characters, or the player suddenly decides to change the action he was about to take when his turn comes, this may require a re-calculation of the whole queue and changing the numbers accordingly, which the player still has to keep an eye out for.
A more radical solution to the problem would have been to provide all the information needed by the player, for any hovered character, in the center bottom of the screen instead of in a halo, doing away with the frankly redundant set of buttons there. I may make the argument for such a change in a future post6.
Thank you and let me know what you think.
1 - Personally, I would have preferred a full turn-based system, but I guess it's unrealistic to ask for such a change at this point as this would require reworking much of the combat part of the ruleset, introducing action points in some form, etc.
2 - In Information Theory they would call this "Denial of information through oversaturation of the channel", but I won't go into that here.
3 - After all, the audience's expectations are for an IE games successor, not a Diablo clone
4 - In my opinion this way much of the artists' and designers' work's effect is lost, and this emotional impact is one of the main goals of a game designer - I'm oversimplifying, but you generally prefer people to associate playing your game with tense tactical combat, not with spending 90% of their time in combat in paused mode, sifting through spells and talents and squinting at the action cooldown indicators (the little horizontal lines).
5 - Unrelated to combat, I think the default animation for characters should be walking, not running, as in BG/IWD. Movement speed of the characters on screen should then be slightly adjusted if needed, but I think it's about right as it is. I also join in Sensuki's request that movement animations be desynchronized to prevent the party look like marching soldiers.
6 - I've worked as a QA on software with a halo-based UI though, so I may be biased here, but my impression was that a UI relying on halos has something inherently wrong and providing sufficient information without cluttering the screen is an unwinnable fight.
Edited by Gairnulf, 01 November 2014 - 04:33 AM.