Khalifate gameplay thread (feedback on balance and bugs)

[Wintermute]

An update: I posted a list of replays wanted in a rough order of need in the first post. The OOS issues that have happened with the 1.8 addon are likely a result of having the old "default+Kalifa" era. Delete that if you have it. You should have default+Khalifate version 0.6 currently. It's possible that some other eras are also causing conflicts (ageless, EE, other eras that have old Kalifa units) but it's not narrowed down yet.

[Rigor]

drakes v kh on the den of onis, cute action in the caves, and felt more or less ok. i thought that drakes have very little to do against jundis in the caves because of their insane mp modifiers that give them the edge. put a healer behind a few of the units, and they can not only take out a defender in the cave but also fight ferociously until the end.

my ideas for the balance include:

jundis should be neutral and have a strong and dextrous trait, to add variety in their use. because they have a lot of hp (up to 42 ive seen) they are waaay too strong as a defender because they can withstand everything and be replaced by others in the next turn while giving only very limited chances to break the line. plus, TOD stalemates are less probable.

nafta throwers are waaaaaaaaay too strong when played against the ud. in retaliation the only impact dealing unit (corpse) receives insane 11 dmg in retaliation plus if u try to zap them (the other thing u could do against them), it not only hurts u back a lot but u need many adepts which are horribly damaged and in no way can be used against other units. my suggestion: fewer hp for nafta throwers, perhaps 6-8 hp for starters. and a lot less damage per strike for the lvl2, perhaps minus 2 for one of melee or ranged or similar.

falcons really pay off every time we have a wide map and for the laughable amount of 10 gold only u make two units run after him while ur focus should be elsewhere. as units that fight they are not as useful as other lvl0s, but then again they can finish off others because of their cute dual attack. i noticed they got 60% on most hexes also villages on plains, and think they should be very very weak on villages like the bat. change price to 13 gold for starters because of all those reasons please.

khalifa horses that are actually camels would add to the falvour when they r slower and different resists but it still feels quite odd, perhaps kolbur could suggest some more concrete things here to help the kiss principle.

R out

[Velensk]

Also, as a side note, the experience requirements for the Khalifa are inexplicably low. Maybe this was intended as a faction feature but it hardly seems like one they need with all their other advantages.

[Noy]

My biggest concern is that I don't understand the resistances at all.

Most of the resistances of most of the units in Wesnoth are flavour-driven, and this is one of their strengths. Minor tweaks might be made for balance reasons, but the mechanics are resonant with our expectations: arrows are poor but maces powerful against skeletons; fire is great against woses; spears are a good weapon against cavalry. For the most part I think that if you got someone to make intelligent guesses about the resistances of units based solely on their descriptions, they would be in the right ballpark.

It might just be the lack of unit descriptions, but I don't understand the Khalifate resistances. Are they human? If so I'd expect the resistances to be closer to 0% all around, so I'm guessing not. So I'd like to understand how the flavour of this race leads to blade resistance and impact vulnerability. If there isn't a good flavour reason (and a convincing one cannot be produced), I think it would be better to change the resistances early on, before a lot of balance work occurs.

In short: KISS.

No offense, I don't believe this is a valid criticism. Human resistances are all over the map too, based on armor. You have everything from HI to thieves. In this case the khalifate carry different armor than what other wesnothian humans do. So do their horses. Simple as that. Absolutely we need better descriptions but this is a project in its infancy and we need help with actually making it workable in the game. Suggestions like this and what rigor has put forward are not the help we are asking for nor are they useful at all... they just serve as distractions.

[Horus2]

Some thoughts so far:

As the person who played the Khalifate in Rigor's Undead experience mentioned, i have to agree that the nafta guys rule the whole match. A slight damage reducing would be a wise decision, on the other hand had to be careful that Knalga might be their toughest rival, and the main choice is nafta against them since Amirs have a hard time with the mass dwarf recruitment.

While not having too much experience with that either, Loyalists and Drakes could make the game stalemateish indeed from the non-kalifa side, but i had a feel that giving the proper negative horse-resistances to the mounted units could already solve this. These factions tend to utilize pierce heavily and that would create an area to strike on. It seems for me with their versality the Khalifate already more likely to have the possibility to choose the time and place of an attack.

Also i worry about their easy levelup. I think an intelligent jundi should level at 25 xp, because his lvl2 is so untouchable because of his versability that even the standard 28 (elvish fighter) or 29 (spearman) is looking dangerous, so 30 might be better with the 70% xp modifier (that means 44 should be written in wml iirc). Similar increase should happen to the naffat.

Arif obviously shouldn't have 2 mps on shallow water.
Or should he? One thing what concerns me and this time by standing on the side of the kalifa: is their poor naval and aerial weaponry which is all representated in that falcon. I didn't see any testgames on Weldyn channel yet, such should happen in the future. I'm curious what they can do, looks if they're quite lost there...

That's all for now.

[Wintermute]

Not a pretty game, Velensk's Knalgans punish me for early mistakes. My recruiting is poor, as I went harder for horse power and ended up wishing for more naffat and arif.

[Wintermute]

A game where troll spam seems to work well until crushed by naffat counterattack.

[Scatha]

Hi Noy,

I'm sorry, it wasn't clear at all to me that this kind of comment was not being looked for (though it was clear that this wasn't the only thing being looked for, I thought I could offer more insight here than elsewhere).

That said I do think that there's more of an issue here than you imply. The Khalifate are very appealing to me in some ways, and I really appreciate the great work the developers do on this game, but there appears to be an (actually pretty small) amount of sloppy design, and if that is ever going to be addressed (I live in hope!) then the earlier it is done, the better.

Of course armour can offer some explanations, although I don't think it's as obviously flavour-driven in the case of the Khalifate as HI and fencers. Some nice writing can perhaps make it clear, but this does seem nontrivial to me.

For instance: do you have an idea of what type of armour the Khalifate wear? From some of the pictures (Arif, Khaiyal among the level 1s) I would imagine perhaps a heavy chainmail (which makes sense in terms of providing blade resistance but not helping much against impact), but the other level 1s look essentially unarmoured but still resist blade and are weak to impact. Of course some of them might wear armour under their clothes, but this is asking us to assume quite a lot.

Furthermore, there is just a lot of complexity:
No two recruitable Khalifate units have the same resistances. That's seven different recruitable resistance profiles, across seven recruitable units, without a driving reason. It really seems uneccessary complexity.

(Comparison of other factions:
Drakes have 3 resistance profiles among level 1s across 6 units
Knalgans have 4 resistance profiles across 8 units
Loyalists have 6 resistance profiles across 8 units
Northerners have 3 resistance profiles across 7 units
Rebels have 5 resistance profiles across 7 units
Undead are the worst offenders with 6 resistance profiles across 7 units; however these are very flavour driven)

Hmm, look, I'm going further down this path than I'd intended. I should also note that I'm deliberately overstating my case, in that I'm not presenting the flip-side of things. Of course I understand that there is a certain amount of inertia towards not making changes which will further upset the balance, but it seems to me that while the balance isn't perfected is the perfect opportunity to get the mechanics to line up and feel right. I'll stop there, and I'm somewhat sorry that I'm taking up room in the thread with comments not in the direction that was wanted. But I care too much about the (beautiful) mechanical aesthetics of Wesnoth to remain silent!

[Horus2]

My last game as randomized orc vs the chosen khalifate of enthusiastic Khalifate-balancer Rigor (who's testgames are REALLY helpful) ended up miserably. Even with knowing that i should buy no grunts as all but trolls instead (which would be rather the opposite against random) i had close to zero chances which covers what actually happened. I stole some village for a while, but he kicked back efficiently and for the next stage with arriving naffats things just looked more grim for the distressed northern damage-dealers...
Replay is here:

Khalifate_Rigor_COTB.gz

Troll like surprise... not.

(32.38 KiB) Downloaded 265 times

[negusnyul]

I've played some test games with the newest SVN, and I'm really impressed. The new faction works very well.
Some feedback:

• In my opinion, the cavalry, the resistances and the Arif's movepoints are good in ther current state, maybe the resistances need (not-so-big) changes.

• The Khalifate is definitely overpowered against the undead, in my opinion they are overpowered a little bit against the elfs (and maybe other factions-I haven't played enough) too.

• The different/weird things mentioned before (resistances, slower cavalry) need to be explained in the unit descriptions, the whole faction needs nice and detailed description/writings.

• I don't think the stalemate against loy/drakes is really a big problem.

That's all for now, I will continue testing and if I can come up with useful replays I will post them.

Keep up the good work!

[Wintermute]

Thanks for the thoughts. I want to be clear that balancing things comes before lore - we can explain about anything after the fact, but it isn't balanced then what's the point. At this stage in the game, the focus is on balance, and issues like "why is it like that" can be dealt with later. If we start inventing reasons for why things are the way they are, then suddenly there are a whole bunch of constraints on what we can or can't change for balance - and balancing is hard enough without all that stuff on top of it. But unit descriptions will eventually be added and updated, and themes will be worked in, never fear.

Also, the reason I'm trying to put the breaks on changing too many things too fast is because these guys are new enough that everyone is still getting used to them. Tactics are going to evolve and even small changes can have a big impact on that. For example, we see some games where trolls look overpowered. Then naffats are used to crush them. If naffats are changed somewhat to better reflect them as more of an anti-UD counter unit rather than just a good fire archer then we get to see what happens to the orc matchup. Hopefully that will also give more reason to use the Rami as the main archer that it was intended to be. So far I remember seeing orcs win with grunts in a game, so maybe the troll-spam isn't as potent as it seems (given that games are now being lost by troll spamming). My point is that a few lost or one games isn't reason to start changing everything around. But this feedback has been very useful, replays especially. I'd like to see some 2v2 replays also, though.

I hope to get a new version with changed unit ids (to eliminate the OOS errors, hopefully) and to address the obvious biggest problems (like naffat). Maybe tonight or tomorrow if all goes according to plan.

[Scatha]

Thanks for the thoughts. I want to be clear that balancing things comes before lore - we can explain about anything after the fact, but it isn't balanced then what's the point. At this stage in the game, the focus is on balance, and issues like "why is it like that" can be dealt with later. If we start inventing reasons for why things are the way they are, then suddenly there are a whole bunch of constraints on what we can or can't change for balance - and balancing is hard enough without all that stuff on top of it. But unit descriptions will eventually be added and updated, and themes will be worked in, never fear.

Thanks, this makes a lot of sense and I essentially agree with you. I think I got caught up a bit in detail, but I was trying to make the following general design point which goes a little the other way: there are a lot of dimensions available for balancing, so some can be used for flavour (as long as they don't obviously introduce balance problems). Changes of this type cannot sensibly be made after the faction is well-balanced (because as you say balancing trumps lore), but there is an opportunity to make some simplifying modifications in terms of unifying resistances and movetypes (and simplicity is good for the players when it's not removing depth from the game) a little bit across the faction at this point, without upsetting delicate balance (because it doesn't exist yet) or making it any harder to balance.

[Mountain_King]

Playing with Khalifate seems to produce corrupt saves all over the place. I was playing a local game against the ai, made a lousy mistake and decided to start over from turn 1. All the saves were corrupt. I'll try and reproduce and then I can post here, if you want.
Doing 1.9.6 Windows, btw.

[Exasperation]

Playing with Khalifate seems to produce corrupt saves all over the place. I was playing a local game against the ai, made a lousy mistake and decided to start over from turn 1. All the saves were corrupt. I'll try and reproduce and then I can post here, if you want.
Doing 1.9.6 Windows, btw.

I may know what this is. In Wesband, there used to be a unit with a '-' in its unit ID, and we would get corrupted saves (the engine was treating the id as if it ended before the '-', or something like that, I've forgotten the details). Changing that '-' to a '_' (just in the unit id, not in the actual displayed name of the unit) fixed it. I notice that the Khalifate has a unit with id=Qatif-al-nar, so that could be the culprit here.

Edit - did a little poking around to refresh my memory, and I was right. In the statistics section of a saved game, the unit ID is used on the left hand side of an expression for how many advancements to that unit type have occurred. If a unit ID has a '-' in it, this results in a malformed expression and a corrupted save whenever one or more units have advanced to the unit type in question (in this case, whenever a unit advances from Naffat to Qatif-al-nar).

[Wintermute]

You're right, I get a corrupted save if it has that Qatif-al-nar in it. I'll file the bug report on this one, but don't hesitate to file bug reports for anything that comes up!