Difference between difficulty settings?

[Longasc]

All the AI gets is probably some right now unknown boni - it does not react smarter or different than on easy / medium on high / very high.

[GeWee]

Ok, thanks. So there's a difference from Medieval then?

Because as far as I can remember in Medieval it was something like:

Easy: Player gets morale bonus and the AI uses dumbed-down tactics.
Normal: Normal AI tactics and no bonuses to anyone.
Hard: Better AI tactics, still no bonuses to anyone.
Expert: Beefed up AI as well as a morale bonus to it.

From what I hear here the AI in Rome gets various bonuses even on hard.
I don't like cheating AI, let it use the best tactics it can but don't let their peasants be stronger than mine... :P

[Spino]

Ok, thanks. So there's a difference from Medieval then?

Because as far as I can remember in Medieval it was something like:

Easy: Player gets morale bonus and the AI uses dumbed-down tactics.
Normal: Normal AI tactics and no bonuses to anyone.
Hard: Better AI tactics, still no bonuses to anyone.
Expert: Beefed up AI as well as a morale bonus to it.

From what I hear here the AI in Rome gets various bonuses even on hard.
I don't like cheating AI, let it use the best tactics it can but don't let their peasants be stronger than mine... :P

No, I recall a developer spelling it out for us here in the Org. In Medieval Hard was the AI at it's very best with a small morale boost for its troops in combat. Expert difficulty had the same AI opponent as Hard but with a large morale boost for AI troops.

[GeWee]

Oh ok, my bad.

[hoof]

I posted this in an earlier thread:

Don't play on Hard if you want a fair fight. Here are some stats from some controlled tests I did:

A few experiments (all battles fought until 1 side routs, large battle, syrian flats during winter, no weather):

first number is my troops left at rout time, second is theirs, lower number always represents the loser

Hastati vs Hastati:

easy level:
(the AI never threw pila on easy, but my troops did)
61/10
65/32
60/13
66/15
62/14

medium:
35/6
13/21
13/54 (my general died early)
18/2
5/22

hard:
6/41
5/52
2/39
3/52
4/48

very hard:
(on very hard, my troops broke and ran much earlier than on hard, which is why there are more of them left at rout time)
17/64
11/57
18/58
11/66
24/61

As you can see, Hastati vs Hastati, all else being equal, the difficulty level does appear to dramatically affect killing ability.

As it stands now, you cannot expect your troops to hold their own vs equal enemy types. If that is how you like to play, that's great, but especially since Generals no longer improve unit killing/surviving ability, I don't like the fact that a unit of AI hastati might be really a unit of AI principles or heavier stats-wise on hard or very hard. I want to fight the best AI tactics, but on equal terms, unit wise. Why couldn't CA have stuck with MTW's model of leaving the stats the same, but increasing morale?

I first noticed this problem when I saw what should be equal/lesser troops walk right through hoplites defending a gateway. Now I know why.


Next I tried Cataphracts (the regular horse type).

Again the first number is my guys left, the second is the AI's, and each side starts with 54 horsemen.

Easy:
45/11
44/12
42/8
47/12
47/18

Medium:
30/20
27/18
5/29
1/18
24/11

Hard:
1/32
16/52
3/45
18/50
15/48

Very Hard:
(again, like with the Hastati, my men broke much quicker than on Medium)
22/47
19/49
20/50
19/47
17/48


I think people don't really understand how much of an edge Hard gives the AI in combat. Maybe MTW spoiled us, but I think a lot of the frustration going on is due to the artificial combat bonuses the AI troops get.

[econ21]

Verty interesting, Hoof. Do you know how a level of experience difference would affect the combats on medium?

The reason for asking is that I vaguely recall expert in MTW gave the equivalent of +1 valour, more or less. That seemed about right (particularly given the big bonuses you could get for general command stars).

Your experiments seem to show a bigger edge than that in RTW, but to be sure one would need data on the effects of experience differences on medium (i.e. fair) comparable to what you have produced here.

[maestro]

So for a fair challenge playing on expert campaign with medium battles would be right????