Is the FAQ mistaken about combat?

[Taxi Driver]

Or is it just me?

Quote: "How does combat work?

* Each battle is divided into rounds.
* In each round, a random generated number is used to determine which unit looses 1 hitpoint. The result depends on the A (attack) value of the attacking unit, the D (defense) value of the defender and on defense bonus (river, terrain, fortification and so on).
* The fight ends when one unit is out of HP and dies (exception: fast units fighting slow units can sometimes withdraw!)


That's not how I thought it worked. I'm pretty sure I've seen rounds where neither unit gets damaged, and rounds where they both do.

I'm under the impression it works like this:

The units take turns taking shots at each other, and the defender always has the first chance to strike. It becomes a little clearer when you notice that the defender never loses a hit point until at least one swing has been taken against the attacker (and hit or missed).

If this is true, then defending units have an even greater advantage than the "defender combat bonus" would suggest...

Anyway, I realize I could be completely wrong but I'd like it if someone would try to explain it..

 

[Jaybe]

Sorry, it's just you. During each round, there is a 100% chance that the attacker or the defender loses a HP.

 

[Tacit_Exit]

Your problem may be if you're basing this on observations of multiplayer combat.

Sometimes the 'resolution' of each 'round' of combat seems to take longer than it takes the Attack.flc to run; thus you get an apparent 'null result' round of combat where I think the info packets are still being sent to other players.Since the attack is still in progress the attack.flc will continue to loop.

This has never occured in single play for me.

 

[DiamondzAndGunz]

Originally posted by Taxi Driver
The units take turns taking shots at each other, and the defender always has the first chance to strike. It becomes a little clearer when you notice that the defender never loses a hit point until at least one swing has been taken against the attacker (and hit or missed).

If this is true, then defending units have an even greater advantage than the "defender combat bonus" would suggest...

Anyway, I realize I could be completely wrong but I'd like it if someone would try to explain it..

Well, first of all, you can't really trust the animation you see on screen to give you an accurate portrayal of who attacks first, etc. (if you want proof of this, just look at online multiplayer- the animation might go through several "rounds" before either unit loses one HP). There's always a chance that the attacker will hit first, and damage the defender. I believe that combat works exactly as described in the FAQ.

 

[Taxi Driver]

Even if the defender always gets the first chance to strike there would always be a chance the attacker will get the first hit.

What I'm basing my observation on is this:

Whenever a HP is lost, it is always just after the *other* unit has taken a visible animated swing. And I could swear I've seen two or three of these animated swings go by without any HP being lost.

I guess I'm wrong but it would be nice to have some way of knowing exactly how it works.

I have only the single player version..

 

[Pembroke]

Is the animation even related to the battle rounds? Or are the pics just looped through more or less continuously until the battle is over? I mean the easiest way to code it would be to just start an animation loop, do the battle rounds, and then stop the animations. Also, the animations don't have the same number of pics in them and if they are run asynchronously then the two figures seen fighting would even be quickly "out-of-sync" with each other...

Anyway the battle probabilities are *exactly* as if the combat is done the way explained in the FAQ. That's pretty strong evidence that the actual combat internals are executed that way, too. There have been many threads about the AI cheating which is why combat simulations and tests have been run many times by many players. The AI may cheat in some ways but actual combat isn't one of them.

If #A is the attacker strength and #D is the modified defender strength then in every round the attacker hits the defender #A times out of #A+#D times while the defender hits #D times out of #A+#D times and so there can't be ties. Thus two veterans will fight for a minimum of 4 rounds and at most 7 rounds unless one of them retreats.

If the combat was done in the "first one then the other" way it would affect the win probabilities and would show up in large number statistical runs.

Trust the FAQ. The FAQ is your friend.

 

[Gothmog]

Yes, the statistics don't lie. Pembroke alone has posted enough charts to convince me that someone takes damage every round. I must admit though that I previously thought the same thing as you, Taxi Driver, based on the animations.

 

[cromagnon]

I think what Taxi Driver is seeing is a discrepancy between the animation of an attack, and the drop in the HP bar. I see this all the time, but I just chalk it up to my slow processor.

This is most noticeable when a tank attacks another tank, and each fires off a salvo. Nothing happens; both bars are still green. Then, just before the second salvo (or simultaneously with it), one of the tanks loses a HP.

I think the animations do not correspond to the 'rounds.'