but that's not how it works in real life.
in real life, up to certain point in an attack, say half way between attack start and weapon impact, you can "change your mind" and react to what the opponent is doing. your "new move" won't be as effective as if it was your original move, cause you don't have time to fully execute it, just sort of. you would not have enough time to react twice, but you would be able to react once if done quickly enough. speed or dex checks might come into play here as well. if you tried a second move after the time window had passed, it would be ignored, and you would simply continue the original move.
yes, but only so much. in real life you have to "commit" otherwise you end up with a very bland move that isn't as effective as it could be.
unless you rely on being a 'feinty' type, someone that tries to wear down the opponent with lots of small hits rather than a few powerful and effective hits.
in either case IRL there's often situations where you end up in a bad place where you cannot recover from. I do my share of sword sparring with a friend and it's the case :)
bottom line: for me "changing your mind" is a perfectly valid choice to give to players, just not in a way they can abuse. M&B handles this with feints (which I do in my game as well)
moves should never be queued, only polled at high frequency. queued moves prevent you from changing your mind when you should still be able to. "oh, look, my pea shooter won't go though this guys armor! ok, i'll switch to my bmf gun. oh wait, i clicked fire a bunch of times, and this POS queued them up. now i sit here for a couple seconds watching myself shoot and reload to no effect before i can change weapons. oh wait, i'm dead! thanks a lot - GD queued input! who the h--l wrote this s--t? oh yeah, bethesda - shoulda known". i experienced this as recently as last night playing fallout new vegas. that's a definite "bad designer - no twinkie". i don't think that's the kind of gaming experience anyone wants to deliver to their users. yet companies do it all the time. i guess they just do think about the ramifications of their design choices enough - (or at all?).
I only queue moves when they are performed almost at the end of the current move.
this came from player feedback that it was hard to time a subsequent attack efficiently because in an animation-based combat system (with your standard animation blending system) it's really hard to know exactly when the attack finished. when I didn't have any queuing at all, on a playsession I could see the player effectively spamming the attack button for the next attack (which I wanted to avoid)
>> the veteran hero or the shaolin master can still fail due to different factors (multi-enemy fights, fatigue, etc). yes they will prevail most times but if it's like Assassin's Creed then it already feels wrong
all i'm saying is more experienced combatants would be better at "counter-moves" or "reaction moves". obviously, you can only dodge so many bullets at once or some finite number of bullets before fatigue overcomes you.
all good there. it's all a matter of, if this is part of the player skill or the character skill.
and no I don't mean button mashing skill. I mean skill through the game mechanics, like a veteran player from shooters of the 2000's (like Quake3) would completely destroy an average player by his player skill - not only aiming and shooting but moving, positioning, situational weapon selecting, jumping, bunnyhopping, etc
the reference to assassin's creed is lost on me. i don't play games that aren't first person view - unrealistic - too arcade-ish - too easy - a crutch for those who can't mentally track a target outside their field of view - and thus would suck at combat in real life. sad but true. not everyone is a natural born fighter pilot.
in asssassin's creed enemies take turns to hit you :D
right now i use direction of attack (direction vector from target center to attacker center) to determine which "side" is hit (front, back, left, right, top, bottom), and then use a hit location table and die roll to determine the area hit. the next step is to use the impact point of the weapon at the time of attack resolution (time of weapon impact) to determine which area on the appropriate table is hit, instead of using a die roll. if the attack is from the front against a bipedal target, if the impact point is high and center its a head hit, mid and center is body, mid left and right are arms, and low left and right are legs. this way the player could go for head, body, sword arm, etc. odds are i'll be adding this to caveman at some point. along with the "counter move" mechanic.
sounds good.
if/when you add this, you're likely to encounter that between the aiming, positioning of the enemy and the "delayed" nature of the hits of an anim-based melee combat (vs. the instant hits from a gun in a shooter) locational damage becomes much less predictable than what you'd first imagine (predictability and precision expectation are more accurate in proportion to the speed of your combat)