- Not sure if your referring to input lag or the time it takes for the swords to swing, but the input lag is minimal, and the sword swing is probably less than a second. So animations, or the way people swing to get momentum?
- Well its not going to be 100% realistic, and none of those moves are so telegraphed besides the Ltr's.
- There is trusting, overheads, ripostes, remise, moving with swings, LtR, etc. There is that portion of, take swing, don't get hit by swing. But thats pretty fundamental.
- Addressed in video I suppose.
-Not input lag, but the time for the physical action to happen.
-
Of course it's not going to be 100% realistic. That's not the point.
- All of those moves fall under "take/block/avoid a big swing".
Let me rephrase that. It's been five years since I practised any combat sport, and one year since I found time to do sport with any regularity.
I would still, right now, take my chances against them in the arena. (Just to be on the safe side, I'd allow myself half a day getting familiar with the weapon.)
They are that much below RL human level.
The characters don't misinterpret your inputs
When my character takes special care to hit the opponent square in the weapon/shield when more than half the body is unprotected, I call that misinterpreting my input. Either that, controls being broken or needlessly requiring hand-eye coordination rarely found outside of top FPS progamers (if you have to actually follow a on-sceen 1mm head moving unpredictably for one second to have a chance at head hit). All three are bad, but I suspect the first one is the best description : when directing a character from above with the mouse, it has to be smart enough itself. But...
The character needs to be smart enough to take away strategy?
is the exact opposite of what I tried (without success, it seems) to write.
The character has to be smart enough so you can
concentrate on strategy.
Which becomes even more true if they can move at any sort of half-reasonable speed, as you won't have time to make the mental calculations to try and compensate their drunkenness as is the case now.
Note that, on that point, you still wouldn't really need faster reflexes - what you would need is the ability to take more decisions at a given time. Reducing the 2s feedback loop (decision/order->action->recovering from action) to, say, 1s or even .5s is still way below "lightning fast" reflex level, and at a range most people should be comfortable with (gameplay tests would be necessary to determine the exact limit, though).
But the fundamental problem is this:
Imagine you have several racing games.
One sort is arcade-like, with unrealistic, stylized physics and art, and even sometimes arrowhead-shaped hovering cars zooming around. Pretty fun, even if unrealistic, though not in the segment we are interested in there - but they are the most common, so we have to mention them.
Another sort is sport car simulation. To varying degrees of fidelity and simplification, they are trying to simulate real physics and use real principles of drive, car inertia, tire adherence, engine consumption, heating and damage... It can be challenging and pretty fun for the more simulation-minded players.
Related, there are those taking the same physics/realism-based gameplay, but apply it to heavy trucks. The gameplay is quite different: due to their enormous mass and lower speed, you have to anticipate each and everyone of your actions. While true to an extent with the previous ones, here it is amped up to eleven.
Sure, you couldn't recklessly take a sharp turn at the last instant, but now you have to anticipate things several times before to even get a chance to not crash.
Now there is this game - it is not a Formula-One simulator, but it still has mid-to-high-end cars. It also has this interesting control scheme with the mouse and a tactics-like view from above, which gives an emphasis on decision-making over pure reflexes like the other games.
But when you play it, you discover that cars are going half their real speed, and have ten times the inertia they should have.
(Now, controls aren't quite as good as they could be, sometimes you click on the road and it comes crashing in the street lamp it thought you had clicked on, but those should be fixed at some point, so let's not worry about that right now.)
What feeling does it gives? The driver is drunk. The cars are underpowered.
In fact, what is really happening is that it is using the heavy truck model of the previous game, while depicting cars.
And while players can bludgeon their intuition to obedience and ultimately work with that gameplay, it is still too much of a discrepancy.
Remaking the entire game to display trucks is out of the question, so instead you have to tweak the physics from "heavy truck" to "about car-like".
No need to make it like the Formula-One simulation above. Not only would you need to simulate too many details, but the control scheme would also make that too hard to play. Have you tried to steer a 650kg at 300km/h in narrow streets
with a mouse from above?
But right now, the cars are about 20 000kg and top at 80km/h, which is respectively about 20 times and half what they should be in the best cases.
I doubt many would be against seriously tweaking that.