Advertisement

Designing a good feel ARPG melee combat system

Started by October 05, 2014 06:06 PM
12 comments, last by wodinoneeye 10 years, 3 months ago

Thanks for taking the time out of your day to offer your advice!


The "attack" animation doesn't really look like an attack; a simple change that might work with your current animation would be to have the weapon extend forward as it swings down, giving the impression of a thrust.
While you do have a flinch animation, it might help to make it more exaggerated, and more connected to what's happening to the character. For example, if you're going to keep the knock-back effect, it might be worth taking that into account, whether by depicting them staggering back or an anime-style backwards skid. (It also looks a little slow to my eye.)
Consider having the sword animation stop short on a hit, to give the impression of the weapon fetching up against--or burying itself into--the target.

Yes, better animations should make it feel better.


I checked the stream (StarMire: just forward to 21:30). I agree with the comments from Thaumaturge

also I see a complete lack of sense of physics and biomechanics. you can spam the attack and get maybe 3 attacks in one second. there's no cooldown, no inertia, no penalty for missing an attack, no way for the enemies to block your attacks, or any other interesting factors beside attacking. the fact that the enemies run towards you like zombies doesn't help either



for some games having a sword attack is enough to call it a 'melee fighting system'. it usually boils down to something like "I was hit 8 times but I hit the enemy 15 times, so I won". Skyrim goes just a little bit beyond this (with blocking, stamina and staggers) but the "I can pause to drink potions" feature ruins all of this (plus aiming and body movement / distance is meaningless)

personally I have to say that a sword attack is only the beginning, and the real fun starts when attacks can be blocked causing a flow of back-and-forth attack attempts and blocks, making all combat decisions much more meaningful.

I'm not sure what to do to remedy the fact that enemies run towards you. Maybe they should get close, then do something like boxers do when they circle. Adding a random time delay to their hits might also help to make it more interesting (although some games have extremely predictable enemies yet stay quite entertaining),

Giving the enemies code to dodge your attacks (or block) might make it better as well. This systems can get frustrating when they automatically do it, so no matter what you miss.

On physics and biomechanics, I didn't think about things like that because of the great Realism vs. Fun argument. I am way more into fun than realism on this project, as opposed to your game, which seems to try to accurately portray swordfighting. I will still add some more physicality though.

And yes, I am aiming more for the 'A' part of ARPG. This is more casual FPS-style.


Building on what Chosker says, an idea:

As Chosker suggests, allow the player to block incoming attacks if the block is performed with the appropriate timing. Give the block an animation as you have with attacking, and make both the attack and block animations fairly long--say about 0.6 seconds each. In addition, only allow a block or attack if neither the block nor the attack animation is playing. However, when an attack lands or a block is correctly-timed, end the animation prematurely. (When the player's attack is blocked, similarly end the animation prematurely, but impose a brief pause--say 0.1 seconds.)



Thus well-timed attacks and blocks make it easier to keep attacking, while poorly-timed blocks and missed attacks leave the player at a disadvantage. If you want to add an additional danger to the player, perhaps allow enemies to counter-attack more quickly when the player mis-times a block.

Good suggestions! A system like this sounds good, especially when the enemies can get a second hit in on a failed swing or poorly timed block (and the player could as well, if the enemy did the same).

I think I will keep recording my progress and iterative cycles so that maybe this thread could become a helpful reference for anyone creating anything similar. In the next recording I plan on tweaking the animations and adding the blocking system suggested by Thaumaturge. More complex enemy movements (as suggested by Chosker) may come later.

Also, I ask that you guys talking about Chosker's game start a new thread or PM each other.

Want to get to know my work and I better? See my website: Au 79 Games

I wrote General Tips on the Process of Solo Game Development

I'm not sure what to do to remedy the fact that enemies run towards you. Maybe they should get close, then do something like boxers do when they circle. Adding a random time delay to their hits might also help to make it more interesting (although some games have extremely predictable enemies yet stay quite entertaining),

I thought Skyrim is your main reference? play it again, but check out how the enemies behave. you might as well watch some (real) swordfighting videos to understand movement a little bit better, although they might be too static for your liking. circling around would be a good addition IMO.

adding a random time delay would be a good first step as well. it all depends on how deep you want to go for programming your AI, which should translate to how smart or varied you want your enemies to be.

and all of this depends on what you expect your game flow to be like. do you expect the player to kill lots of enemies with little AI (Diablo, Serious Sam, or your best known modern zombie shooter) or is it more about having fewer but more interesting and challenging encounters?

On physics and biomechanics, I didn't think about things like that because of the great Realism vs. Fun argument. I am way more into fun than realism on this project, as opposed to your game, which seems to try to accurately portray swordfighting. I will still add some more physicality though.

And yes, I am aiming more for the 'A' part of ARPG. This is more casual FPS-style.

I don't see why Realism vs. Fun should be an argument, you can have both, or none, and it's also a matter of what's fun for every person. To me it's more fun to play the more realistic combat of Chivalry vs. the less realistic combat mechanics of Minecraft. Many years ago Counterstrike came out and it attempted a more realistic approach of a shooter vs. the more arcadey shooters from back then, and it was much more fun for many people. and if my game's combat wasn't fun (for me at least) I would've changed it long ago.

you also have a different form of realism: you can die. Minecraft's combat was fun because it was realistically dangerous

realism/fun aside, the fact of the matter is that you need to define a limit for the player's actions to an extent, and you already have the most basic example: there's a (small) cooldown between the player's attacks as opposed to allowing every click as a valid attack. it's the same for the guns in a shooter (at the very least there's a minimum fire interval).

from there on you can build up on it, by designing different limitations and actions. for example crouching in an FPS makes you smaller (harder to be aimed) but it also makes you move slower.

it's up to you to design how and where you would want to limit the player's actions, and how to make these interact with eachother (rock-paper-scissors)

Also, I ask that you guys talking about Chosker's game start a new thread or PM each other.

I agree and that's why I initially said I didn't mean to advertise. however even if mine or Thaumaturge's game aren't Skyrim, I would've thought you to be interested in knowing how they work

Chosker - Developer of Elium - Prison Escape

Advertisement
Sorry, "find" was a poor choice of words. I meant when I skip around in the video, it doesn't buffer properly. That player isn't friendly to slow internet connections.

I bit the bullet and downloaded the game, but there's no executable here and I see I need to download and install Blender to play it.
So, to anybody else who wants to try this: It's not even potentially dangerous, but you do need blender.

makuto,

I've only been able to see a couple seconds of gameplay in the video, but I agree that circling would help a lot. The animation issue is complicated, but as I understand you're going for a bit simpler gameplay, you might not want to represent sword swings too literally. A few variations in attacks and block regions would go a long way, though.

On blocking by the enemies, the simplest way to do it is block by default. Enemy uses a shield, you kick to knock them off balance and then time and aim an attack right before they can bring up the guard again. This would create a much more interactive experience, and isn't too complicated. Also, you can have them raise and lower their guard somewhat randomly as they circle, or before they attack, to let in attacks and interrupt their attacks.
I agree that, for what you're doing, you probably shouldn't have them respond to an attack by blocking, which could become frustrating if you want to deal with more enemies (which I assume you do, based on the presentation you linked to).

A system with a progressive gamble payout (like those Las Vega Slot machines) where a progression of actions leads to a much better result, but with the risk of being interrupted/disrupted (And secondary actions/preparations involved to prevent that).

Unpredictable results/surprise can increase player interest (particularly with fancy optional visuals ???)

The player has the choice to do things simple (more certainty) or more complicated (with more risk but better reward). Balancing between those paths to prevent loopholes will probably be a chore.

--------------------------------------------[size="1"]Ratings are Opinion, not Fact

This topic is closed to new replies.

Advertisement