Topic: Rebalancing the tribes?
king_of_nowhere
Topic Opener
Joined: 2014-09-15, 17:35
Posts: 1668  One Elder of Players |
Posted at: 2016-02-16, 22:59
I think the balance of the game should be modified a bit. For admission of the programmers, the balance was decided by number crunching and eyeballing, and it worked fine enough. But that was before it became known that with some micromanagment it is possible to get fully promoted soldiers quite early in the game. In those old times, balancing focused on lower level soldiers, but at level 10 there are some imbalances. First thing, barbarians are too disadvantaged. As explained in that thread, https://wl.widelands.org/forum/topic/1929/, they lose 67% of times against atlanteans, and 63% against imperials. That's about two times out of three, meaning that in a one-by-one fight the barbarian player can expect to lose two soldiers for every one he kills. And barbarian soldiers are already more expensive that others in term of metals needed for the equipment and food needed for training. And their deep mines are more expensive to operate. it's only fair that the strongest tribe in early game get short shrift later, but there's such a thing ass overdoing it. Especially because, among players that know the micromanaging, "late game" starts after two hours. In most maps, that's before even making contact. On the other hand, atlanteans have a distinct advantage: they can get an almost fully promoted soldier without making any mone, just with the metal they have. They have to skip the defence promotion (actually you can even get defence 1 if you do things very accurately, but it requires a lot of attention, if you make a mistake you lose the change to get the attack promo, and the benefit is minimal anyway; not worth the effort imo), but an atlantean soldier with full attack, evade and healt promotions is already fully capable of laying waste on rookies. And atlanteans can get that soldier in less that one hour. When the empire needs 90 minutes, and the barbarians 100, to get their first level 10 soldier. If the atlanteans had to make mines, they would need 90 minutes to get the first fully promoted soldier, which would be more fair. So, I have two (easy) proposals to fix those two imbalancements that would have no influence on the rest of the game. 1) For the barbarians, raise the attack gained per level from 7 to 8. It would allow barbarians to kill another soldier with 4 hits instead of 5, and that would bring the odds in favor of the barbarians to 52% against atlanteans and 56% against empire; close enough to be fair, and barbarians would still be at a disadvantage because their soldiers are much more expensive. If you think that's too much of a boon, it would be possible to increase cost on some barbarian equipment. I was thinking even adding a 6th level of attack (with every attack giving +7, so same result of killing in 4 hits, but more expensive) but that would require massive recoding. Raising the attack gained from 7 to 8 would require changing a single digit inn the config file. 2) for the atlanteans, make the labirynth cost 2 gold, or the dungeon cost 4 gold. That would force the player to spend all his gold for the training sites (like the other tribes) and would prevent him from making level 10 soldiers until he has a gold mine. it wouldn't affet the rest of the game. P.S. Atlanteans are my favourite tribe, so I'd like them to stay strong. Barbarians are my least favourite, but I use them in challenges to make it more difficult for me, so I'd like them to stay weak. I am definitely not suggesting this rebalance to favor those I like. P.P.S. I doubt there are more than a dozen people or two that knows how to micromanage this game, so it doesn't really make much difference for everybody else. Edited: 2016-02-16, 23:03
Top
Quote
|
GunChleoc
Joined: 2013-10-07, 14:56
Posts: 3324 One Elder of Players Location: RenderedRect |
Posted at: 2016-02-17, 09:48
The Atlantean trainingsites have 2 diamonds and 2 quartz in their building costs. How about changing that to 1 diamond, 1 quartz, 2 gold? Or 1 diamond, 2 quartz, 2 gold for one, and 2 diamonds, 1 quartz, 2 gold for the other? This way, we would get the gold cost without raising the overall building cost too much. I agree we should do something about the Barbarians as well, but I am not sure which is best. Giving them more attack would certainly be the easiest solution coding-wise. Busy indexing nil values
Top
Quote
|
|
einstein13
Joined: 2013-07-28, 23:01
Posts: 1118 One Elder of Players Location: Poland |
Posted at: 2016-02-17, 10:24
Barbarians are famous of great attack. Giving them more of it will stick to the tribe einstein13
Top
Quote
|
|
king_of_nowhere
Topic Opener
Joined: 2014-09-15, 17:35
Posts: 1668 One Elder of Players |
Posted at: 2016-02-17, 11:11
yeah, exactly. I thought, in popular culture, the "barbarian horde" concept is associated to two things: a big army of untrained fighters, attacking in an uncoordinated human wave; and the barbarian warlord, a hulking brute with a huge axe that can bisect a man in heavy armor. by giving the barbarians cheap but weak level 0 soldiers, with strong but expensive level 10, we get both sides. As for atlantean training site cost, I agree with removing a couple oher wares from the cost. By the way, a crystal mine produces the same amount of quartz and diamonds, but diamonds are less used, making them actually cheaper than quartz. So I'd remove quartz from the cost.
Top
Quote
|
|
GunChleoc
Joined: 2013-10-07, 14:56
Posts: 3324 One Elder of Players Location: RenderedRect |
Posted at: 2016-02-17, 15:46
I have created a branch and merge request: https://code.launchpad.net/~widelands-dev/widelands/balancing_tribes/+merge/286346 The merge request will give us automated builds for you to play with. Busy indexing nil values
Top
Quote
|
|
king_of_nowhere
Topic Opener
Joined: 2014-09-15, 17:35
Posts: 1668 One Elder of Players |
Posted at: 2016-02-17, 17:15
i don't think i can play with those builds because i don't have linux.
Top
Quote
|
|
GunChleoc
Joined: 2013-10-07, 14:56
Posts: 3324 One Elder of Players Location: RenderedRect |
Posted at: 2016-02-17, 17:24
There will be Windows builds available too I can post a link once we have some; it will take a while. The edits are in the Lua files only, so you could also look at the diff and recreate the changes in your current installation, if you have a dev build. Edited: 2016-02-17, 17:25
Busy indexing nil values
Top
Quote
|
|
king_of_nowhere
Topic Opener
Joined: 2014-09-15, 17:35
Posts: 1668 One Elder of Players |
Posted at: 2016-02-18, 00:31
right, I could modify the barbarian tribe by myself on the lua file. The real problem is, I doubt playtesting could yield any meaningful information. The suggested changes are visible only with fully promoted soldiers. The AI is not capable of fully promoting soldiers in a short time. If it has contact with the enemy, it starts to fight, and will never get fully promoted soldiers. Also, the AI is better at playing some tribes than others. It is bad at playing atlanteans, but it also has huge problems with barbarians if it must rely on deep mines from relatively early. So, no amount of playtesting with AI will tell if the changes are good. And if I play with AI, I win anyway; I can't tell from that if the tribes are more balanced or not. So, the only way would be to playtest among humans; and not any humans, but humans who are good and are capable of micromanaging high level soldiers from the early game, and that also are active members of the community, or they wouldn't volunteer to playtest. That seriously restricts the field; how many such players are there? I can only think of me and einstein. There are other strong players, but I haven't seen them posting. Even then, it would be really hard to make some meaningful playtest. Let's assume I play against einstein with barbarians and I lose; does it mean barbarians are still weaker than empire, or does it just mean that I suck with barbarians? (they are, after all, my least favourite tribe). Or if I win, does it just mean I'm a stronger player than him? And if both of us lose with barbarians, does it mean that they still are not balanced, or that both of us are bad at playing with them? In fact, the only meaningful result of a playtest would be if both of us could consistently win with barbarians against other tribes, in which case we could conclude that barbarians are too strong. But I'd bet heavily that this is not going to happen. So, in the end, I think number crunching and eyeballing is still the best way we can decide whether to try this rebalance or not. And I think some more people would need to agree with this before we decide a serious change in the game balance after a discussion of only three people. By the way, the two proposed changes are independent; we can decide to raise the cost of the atlantean labirynth without raising the attack of the barbarians, or viceversa.
Top
Quote
|
|
GunChleoc
Joined: 2013-10-07, 14:56
Posts: 3324 One Elder of Players Location: RenderedRect |
Posted at: 2016-02-18, 07:54
I agree that this is tricky to test - maybe you could play against yourself in LAN mode and have a fight between the Barbarians and the other tribes to see how the soldiers fare with surviving. We could also hack the starting conditions to start a game with fully promoted soldiers? And yes, the two changes are independent of each other. The Windows builds are ready. Go here: https://ci.appveyor.com/project/widelands-dev/widelands/build/_widelands_dev_widelands_balancing_tribes-569 then pick your build type and Operating system, then click on the "Artifacts" tab to download. Busy indexing nil values
Top
Quote
|
|
DragonAtma
Joined: 2014-09-14, 00:54
Posts: 351  Tribe Member |
Posted at: 2016-02-18, 10:13
If the AI is poor at upgrading soldiers (and it is -- all too often they can't even create the equipment for new soldiers!), it means that the AI needs an overhaul. There's a reason I normally set everyone to Fortified Village instead of Headquarters.
Top
Quote
|
I agree with king_of_nowhere in this case.