Question? How do ranks improve battle capabilities.

[PBI]

From what I've read elsewhere it seems fairly conclusive that experience doesn't affect the effectiveness of missile units, either in terms of accuracy or attack strength. It only upgrades their melee ability and morale.

This means experience is almost useless for missiles and artillery. So there's no point in keeping a gold-chevron ballista once you don't need it, just disband it and build a new one when you need it.

I believe this is different to MTW where even one chevron led to a very noticeable increase in accuracy. I remember seeing this during battles, my artillery suddenly became much more accurate as soon as it garnered enough kills to get promoted.

It raises a point about gaining experience: Towards the end of a siege, when the enemy have only 60 or so men left in the town center, my instinct would normally be to shoot them down for free with my remaining missiles, rather than lose any more men. But perhaps it would be better to send in the heavy infantry, as the experience will be very useful for them, but not for the archers.

[Ars Moriendi]

...experience doesn't affect the effectiveness of missile units, either in terms of accuracy or attack strength....This means experience is almost useless for missiles and artillery....

That may not be entirely true.

I suggest the following simple experiment :
Custom battle, grassy plain, 1vs.1 peasant archers, one unit 9 exp, the other 0 exp.
Order your archers to attack, hit "C" when in range (AI will do the same), then hit "CTRL-T" to speed it up.
Run it ten times, it will only take you 5 min.

My results : the gold unit won each time, average losses 35%. So, there must be some advantage for ranged units in those chevrons. (M2TW 1.02)

[Ramses II CP]

The advantage you're noting is the improvement to their defense 'skill.' Still, if you were the archers with the chevrons every time then you were also taking advantage of the well known 'the AI Just Sucks' exploit, wherein the AI makes incredibly foolish decisions in the middle of battle, like readjusting it's position in the midst of a archer fight, causing it's men to fire fewer volleys and take far more losses.

Run it again with yourself as the no chevron archers an equal number of times and, though the defense skill will still help the AI, I'm willing to bet you'll get better results with the weak archers than the AI does.

[Zim]

I thought that defense skill was ignored by ranged weapons. Or is he having them go into melee after they run out of ammo?

The advantage you're noting is the improvement to their defense 'skill.' Still, if you were the archers with the chevrons every time then you were also taking advantage of the well known 'the AI Just Sucks' exploit, wherein the AI makes incredibly foolish decisions in the middle of battle, like readjusting it's position in the midst of a archer fight, causing it's men to fire fewer volleys and take far more losses.

Run it again with yourself as the no chevron archers an equal number of times and, though the defense skill will still help the AI, I'm willing to bet you'll get better results with the weak archers than the AI does.

[Ramses II CP]

I thought that defense skill was ignored by ranged weapons. Or is he having them go into melee after they run out of ammo?

If that's true it's news to me. The morale difference might still have some effect, though I'd expect it to be minimal since companies under fire rarely rout until they're almost destroyed.

Still, as repeated experiements have shown, the player's soldiers have an advantage in almost any 1 vs 1 engagement. To get a valid test you have to take both sides.

I would also hope for an attempt to count volleys, as it's quite possible that the veteran archers shoot faster but still not more accurately.

[Yaropolk]

Missile attacks ignore defense skill

[Abokasee]

Missile attacks ignore defense skill

Unless its shields

[Ars Moriendi]

I thought that defense skill was ignored by ranged weapons. Or is he having them go into melee after they run out of ammo?

Yes, it is and no he isn't. Besides, the battles were over long before ammo was depleted.

....you were also taking advantage of the well known 'the AI Just Sucks' exploit

Good point. So, in order to remove the "dumb AI" factor, I ran a different kind of test, with me playing the dumb part.

I took a unit of swordsmen and just stood there taking arrows, stopping the battle after exactly ten volleys from the AI's peasant archers.

Results :
- when against 0exp archers, average deaths per volley : 0.64
- when against 9exp archers, average deaths per volley : 1.44

test ran in same conditions as before, 5 times each case.

Conclusions :
- 3-gold experienced archers are twice as good as fresh recruits
- it's not because of morale or xp-added attack/defense skill - these never came into play (units never meleed, morale never changed during the test)
- it's not because of faster firing
- so it must be something else like accuracy or lethality

EDIT :

I ran another, more dramatic test, using Scots Guard instead of peasant archers and counting to 15 before stopping.
When against 0exp , average deaths / volley was 1.7
When against 9exp , I never really got to 15 - by the time the Scots fired 10-13 volleys, my swordsmen were all dead and/or routing.