Downhill units getting combat advantage

[FactionHeir]

If this is true...ouch!

I did get suspicious when my hashashim got obliterated charging down the slope at spear militia and mailed knights. Even more so when the mailed knights while engaged in melee since a while suddenly started charging through my hashashim at my crossbows a bit further up the hill and literally running over both xbows and hashashim!
But that's another story.

[AussieGiant]

Hell!! ...that is bad.

I hope that was picked up by QA.

[Molinaargh]

Wow!

Units marching uphill still get really tired, tough, so that might balance it out in campaign battles.

[Pass]

The terrain bonus still appears to be working as expected when identical soldiery face off.

Carpathia

Very Hard difficulty, small unit size
60 Vanilla Armored Sergeants defend atop a long steep incline
AI : 60 Vanilla Armored Sergeants attack

AI marches up incline and charges without resting. No player intervention, defender holds and AI routs as expected. 33 casualties for defender, 41 attacker.

600 vs. 600

AI attempts to focus at two points along the line. I allowed it to do so. AI manages to charge twice. Total casualties extremely close to the above ratio, with defenders causing roughly 25% more casualties than attackers.

VH difficulty, huge unit size
1500 vs 1500

AI attempts to stack an assault of 3 units (450) against 1(150) defender along the line and manages to rout that defending unit. What is interesting is that the TW rules behind the battle seem to still be strongly enforced. Only the 1st attacker was actually involved in the fighting on-screen and animating attacks. The remaining 300 attackers were simply stacked up behind in a ridiculously deep AI formation.

Even the loss of 1 routing defender wasn't enough to change the overall ratio of casualties or the predictability of the battle however. This of course says nothing about a possible problem with specific units that would require patching.

I did notice that for whatever reason terrain was effectively slightly less decisive in R:TW and this looks to have been repeated for medieval 2.

[Myrddraal]

In discussions like this, it would really help to know if the TW engine factored in 3D collision detection based on the models, or if all animation is basically derived from the algorithms, as a final display. Do we know how this works?

It doesn't. There is basic 2D collision to make sure soldiers don't overlap, but actual hits are not governed by 3d collision detection. Having collision detection on all those units would be a huge drain on the computer...

[Daveybaby]

It doesn't. There is basic 2D collision to make sure soldiers don't overlap, but actual hits are not governed by 3d collision detection. Having collision detection on all those units would be a huge drain on the computer...

Seconded, i've definitely seen this stated explicity by CA (but cant find a link right now).

Note: The rate at which a soldier attacks (i.e. how many times per minute he attempts to hit an opponent) does seem to be determined by the animation system, but whether each attack is successful or not has nothing to do with the 3D animations. If anything it's the other way round - the game calculates whether the hit is successful via stats and random numbers etc, and then plays the correct animation based on the result.

[zulukiller]

@pass try it the other way round were the person with the high ground is attacking.