Battle AI too easy

[QuintusSertorius]

No, that 0.01 is 1%.

[Jurand of Cracow]

I'm all at sea then... Gigantus in his 1648 mod uses the following number:

<captain_modifier float="12.4"/>

What does it mean? 124% probability? Is there another modificiation? Is it a kind of logaritmic scale? I don't understand

[Krampus]

Woops, yeah that's what I meant, 1%. Came out wrong.

[yonyu]

Soo i was wondering how battles in Third age and Terra land of the great torment,are alot harder? Major changes to vanilla battle config? Higher morale? Can i edit myself some txt file to change it?
I like to add that in 150 turns i attacked only rebel and the selekuids,and took from them only one city arbela.

[BHL 20]

I think playing around with higher morale values would be your best shot. I don't know if there's a way to change morale modifiers like the effect of fatigue, if so, that could also be a factor.

[YourMadDoc]

Soo i was wondering how battles in Third age and Terra land of the great torment,are alot harder? Major changes to vanilla battle config? Higher morale? Can i edit myself some txt file to change it?
I like to add that in 150 turns i attacked only rebel and the selekuids,and took from them only one city arbela.

I believe harder battles are mostly thanks to higher morale. It was long time I played those but I think they had higher morale in general and also lots of units which boosted morale of nearby units, which made them even harder to break. Some units almost never routed and they were extremely strong.

Another thing is, that most faction had generals and their bodyguards with epic stats. One particlular faction in Terra, I dont remember name but it was medieval faction, they lived on small island in middle and they resembled Knights of King Arthur to me, they had bodyguards so strong that their 10stargod-king killed 1/3 of my Spartan army just by himself when I tried to capture their capital. Most of my battles in EBII ends when enemy general/ captain dies which happens quite often, considered that most faction have cavalry bodyguards and they use it for flanking attacks.

[QuintusSertorius]

I should note something I've just discovered, which impacts morale. AI characters get a bonus trait on Hard and Very Hard campaign difficulty (NOT battle difficulty) called AIGeneral, which amongst other things gives at least +2 morale (1 point of Command, 1 point of Chivalry). This was, however, disabled for the Rebels, meaning on Hard and Very Hard, they are easier to fight than other factions. That's now been fixed.

Note that morale and fatigue interact with the balance of a battle. If morale is too high, you get grinding slogs of attrition, where tactics mostly don't matter because unit won't rout until they're so worn down they can't fight on. If morale is too low, they break having suffered barely any casualties at all, and insta-routs are possible. However, there's a middle ground between those two, because a unit that routs earlier, preserving more of it's numbers, is more likely to rally and return to the fight. That's the logic of setting morale lower in EBII than in some other mods, so that there is the possibility for that rallying to happen. This is especially important with regards skirmishing cavalry, who are fast enough to escape, regroup and return.

[BHL 20]

This argument only works for cavalry not so much for infantry. Once an infantry rout starts the battle is over in most cases because too many of them will be killed to rally successfully. If the opposing infantry line is tough to break down, it gives the impression the battle is harder despite the AI having the same level of intelligence.

[QuintusSertorius]

This argument only works for cavalry not so much for infantry. Once an infantry rout starts the battle is over in most cases because too many of them will be killed to rally successfully. If the opposing infantry line is tough to break down, it gives the impression the battle is harder despite the AI having the same level of intelligence.

As far as the "never rallies" that works the same for all units. But it's more pertinent for cavalry, which is why they have lower morale, on average, than infantry.

[YourMadDoc]

Low morale for skirmishing cavalry works great, they will regroup and attack or I have to chase them all around the map, wasting my own forces on it so I consider their tactics valid. Buuut, it does work well for other units in my game. I personally can't resist routing infantry. If I see some unit run, I have skirmishers on them right away while all my infantry is engaging/flanking rest of army and cavalry is hammering them from rear and it ends in minutes. AI also forms large mass of units, which makes them easier to surrond.

Low morale for units makes also a huge difference between player's and IA armies, as a player can (and I believe most players do) make much better use of it. I can flank them much more efficiently, I hunt enemy general, protect my own general and I always have one and he is good at his job. AI lacks in those areas and players armies are rarely affected by it.

I actually don't think morale is too low in EBII, when I fight in streets, units can hold adequately long, but I think flanking and death of general (and AI generals and captains have strong suicidal needs) have too much impact on morale. More confident troops would make battles more hack'n'slash thing, but I think it would lower differences between player and AI a bit and with higher causalities, it would make whole campaign harder.

[QuintusSertorius]

Again, infantry has higher morale than cavalry. Obviously skirmishers have low morale, because they're generally levies.

Nothing we can do about Captain-led armies having low morale, that's another area where giving the AI time to build its forces tends towards less fragmentation, and thus more character-led armies.

Sieges aren't analogous to field battles. Pathfinding is broken, amongst other things which don't work the same.

[Neuchatel]

Bump because the problem is still here

[Gunydii]

Not utterly sure I agree with that assertion. That said i wonder if there is a financial way of skewing difficulty.

I'm about to say this and hate myself (as most of the factions I chose to play, I go for with a view to creating the most pretty, if vaguely historically plausible royal army). If the upkeep costs favored levies then pros then elites (so you get the least bang for your money at the elite end) a bit more I think this could make things more difficult. having any unit (or five) unexpectedly rout could ruin your day on any given field of battle, which is obviously more likely with levy level troops.

My guess is that some of the 'this is too easy' views are informed by having armies that are highly reliable vs ai armies that are often less so.

My current impression is these kinds of diminishing returns happen, a bit. That said making it harder to maintain professional quality standing armies could make armies less reliable and could also make it harder to have armies with large numbers of high experience troops.

[Hummer]

To expand on the above post #73. If the comment raised indeed is the issue.

Currently the upkeep of a unit is set at a certain amount of the cost of the unit, correct? To discourage stacks of professionals and elites maybe it should be set so that the upkeep of elites goes up exponentially.

For example. If levies are 1/8 upkeep to original cost, make professionals 3/8 upkeep to original cost and elites 1/2 upkeep to original cost.

Or something along the like. You get the picture of what I'm trying to say.

[Kull]

If you raise the cost of elites too much, the AI won't recruit them either. But that's not the issue:

1) The human player will always (especially after attaining a kingdom of moderate size) be able to out-plan and out-play the campaign AI. We have a pretty good CAI, but it will never outperform even a decent human player. That's just how it is.

2) Because the human player can out-plan the CAI, the majority of Battles will happen at the human player's behest or with adequate time and space to prepare a response.

3) That means the vast majority of battles will occur when the human player is more prepared (both in unit count and quality) than the AI.

4) There is no Battle AI (certainly not in M2TW) which can overcome those advantages.

5) Accordingly, if the human player wants interesting and/or exciting Battles, they need to impose handicaps on their own play:
- Use the general cam
- No green triangles
- No banners
- Unit count/quality at 75-80% that of the AI

If that's still too easy, try 60-70% - at some point there will be a force ratio that adequately compensates for the difference between your skill level and that of the AI.

OR. Continue complaining that the Battle AI "isn't good enough". The team has done all that we know how to do in this area, and you won't see anything else from us in this area (barring new knowledge on ways to tweak the AI that we are currently unaware of). The rest is up to you.

[Neuchatel]

Other mods like Stainless Steel or Third Age have great battles.
Unfortunately it seems that the EB team trapped itself in a wrong approach.
The fact that AI armies rout in 30 seconds in a lot of battles is a huge blow to an otherwise excellent mod.

[QuintusSertorius]

Watch charlieh's excellent Let's Play, in which you see battle after battle that aren't "30-second insta-routs" (he's playing on H/H). As just one easily obtained example that proves the charge to be nonsense.

[Weltschmerz]

Speaking of battle difficulty, are there any drawbacks in using a setting higher than the recommended M?

[Jurand of Cracow]

Other mods like Stainless Steel or Third Age have great battles.
Unfortunately it seems that the EB team trapped itself in a wrong approach.
The fact that AI armies rout in 30 seconds in a lot of battles is a huge blow to an otherwise excellent mod.

I think all three mods have great battles, but they differ in approach. In the SS the units are more likely to rout earlier (as compared to the EBII) so I don't think this is a problem of the EBII. In the SS the tension of battles consists of the question if your center or weaker wing is able to hold long enough until the other part of the front wins and provides a support. You don't have much time to pull it off and you may lose. I believe, the EBII it's different in the sense that loses are lowers so units can hold longer and there's more time for maneuvers. There're also many other differences that impact on the shape of the battle: the charges of the cavalry are less deadly or there's no Dread of the generals - both of which should make a rout less imminent than in the SS.
Anyway, the approach is a matter of taste and you may find one or the other more enticing.
In both mods for a thrilling battle it's better to come with inferior forces, eg as Kull proposed.

[Genghis Skahn]

Other mods like Stainless Steel or Third Age have great battles.
Unfortunately it seems that the EB team trapped itself in a wrong approach.
The fact that AI armies rout in 30 seconds in a lot of battles is a huge blow to an otherwise excellent mod.

1,2. https://www.youtube.com/watch?v=pWdd6_ZxX8c

More on 1, I've personally played TATW as the Orcs of Gundabad(or in other words, arguably the weakest faction in the game--let's not even address how EBII's factions are more balanced than TATW's, IMO) for a few hundred turns. It's a fantastic mod, but comparing it's battles to EBII's is quite ludicrous. Since it's a fantasy mod(seriously, does any more need to be said about the differences in battle systems between EBII and TATW besides this?), it has a number of stunningly OP units(most elven units, the dunedain riders and trolls--that's not even mentioning heroes and the Balrog, etc. etc. etc.--this is forgivable once again because it's a fantasy mod...), and unlike EBII(especially since it's a mod with medieval weaponry) cavalry charges are still ludicrously OP like in vanilla M2TW. I've confirmed this by my mass use of warg riders(pre-military reform, most of my orc units were total garbage save for the warg riders--general BGs included. By the way, I found TATW's, and many other mods, infantry bodyguards to be ridiculously crappy when compared to our MUCH more balanced EBII infantry BGs), as well as using a very small number of dunedain BGs to wipe out an entire orc garrison(don't even get me started on how underpowered orc units are--but that's definitely intentional AFAIK). The only cavalry who come close to the levels of TATW's OP cavalry units are cataphracts, who are few and far between(and cost a fortune) and who are also quite slow moving.

TATW's battle mechanics are fundamentally different from EBII's: the devs had totally different intentions, use totally different stat systems and the BAI and battles in general behave completely differently. As well, since EBII is a historical mod vs a fantasy mod expecting EBII's battles to be more like TATW is just plain silly. TATW is excellent for what it is, there is no denying that, but it is also fundamentally different from EBII. In case you didn't realize(try reading Thucydides and see how often routs are mentioned vs the deathmatch slugfests that people seem to prefer in this thread) routs in ancient warfare were far from a rare occurrence--they were the norm AFAIK. The team has repeated itself multiple times stating how most casualties during a battle occurred during routs, and not during actual fighting. We have also mentioned again and again that making units more easy to rout allows them to more easily reform and return to the battlefield(this phenomenon is also mentioned by Thucydides, there are several occurrences mentioned where soldiers rout, then reform, and end up defeating the enemies who initially routed them since they chased after them in disorder). Interestingly enough, that last tidbit has personally happened to me in battle, with respect to elite units like Hypapistai and Agema Phalanxes--I've had them rout before, then they reform, and in the heat of battle after chasing them, they end up routing or slaughtering my troops in return. That's why I feel a LOT of pressure to keep a routing unit routing, because if they do reform they can cause a lot of trouble for you.

3. This kind of pure exaggeration really doesn't give you any analytical credibility whatsoever. I've played thousands of battles, all on 6x speed, and I've virtually NEVER seen "armies rout in 30 seconds in a lot of battles". Have I seen reasonably quick routs? Yes, but never 30 second routs(and on 6x speed, that's saying something). Sample bias much?

If you raise the cost of elites too much, the AI won't recruit them either. But that's not the issue:

1) The human player will always (especially after attaining a kingdom of moderate size) be able to out-plan and out-play the campaign AI. We have a pretty good CAI, but it will never outperform even a decent human player. That's just how it is.

2) Because the human player can out-plan the CAI, the majority of Battles will happen at the human player's behest or with adequate time and space to prepare a response.

3) That means the vast majority of battles will occur when the human player is more prepared (both in unit count and quality) than the AI.

4) There is no Battle AI (certainly not in M2TW) which can overcome those advantages.

5) Accordingly, if the human player wants interesting and/or exciting Battles, they need to impose handicaps on their own play:
- Use the general cam
- No green triangles
- No banners
- Unit count/quality at 75-80% that of the AI

If that's still too easy, try 60-70% - at some point there will be a force ratio that adequately compensates for the difference between your skill level and that of the AI.

OR. Continue complaining that the Battle AI "isn't good enough". The team has done all that we know how to do in this area, and you won't see anything else from us in this area (barring new knowledge on ways to tweak the AI that we are currently unaware of). The rest is up to you.

If anybody can properly counter the points outlined here, I will be impressed. Since no one has given this post a proper reply, I'm going to assume that your arguments simply aren't good enough to counter the points presented here.

I think all three mods have great battles, but they differ in approach.

Exactly. It may have sounded like I was trashing TATW at some points(I certainly hope any TATW team members don't feel that way) but TATW is fantastic for what it is. I was REALLY impressed upon seeing it for the first time, even if I ended up choosing one of the worst(and unbalanced) factions in the game