Latest Posts

Topic: Amazons balancing

tothxa
Avatar
Topic Opener
Joined: 2021-03-24, 12:44
Posts: 501
OS: antix / Debian
Version: some new PR I'm testing
Ranking
One Elder of Players
Posted at: 2023-10-03, 21:49

OK, I've added quartz to some armours, and some rope where some more fine tuning of costs was needed. I reverted the total amount of gold to 3, so now they need 5 quartz and 3 gold.

I reverted the probability of empty goldmines to 5%, but kept the lower water consumption to not increase total water too much.

I also reverted rations from training sites, and added some extra fish/meat to bring it to the same level as other tribes.

I also reverted the amount of logs for a charcoal to 6, but decreased the time some more. I kept the rare tree kiln as it was. (I came to like it a lot, it just fits the tribe so nicely. And its total build-up costs justify the efficiency improvement. I think I'll be able to make the necessary changes to the blender models too, so graphics won't be a problem either.)


Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 2760
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2023-10-04, 21:02

well I checked the new values and most of them seem reasonable now. However I still don't see the need for the selected firestones. I'd like to propose to have a normal charcoal burner as a second worker instead and add another quartz to the enhancement cost of the rare kiln instead.
So I vote to see how the already commited changes work in the next tournamant round and if Amazons are still too strong we might have this branch in an instant.


Top Quote
tothxa
Avatar
Topic Opener
Joined: 2021-03-24, 12:44
Posts: 501
OS: antix / Debian
Version: some new PR I'm testing
Ranking
One Elder of Players
Posted at: 2023-10-04, 23:08

hessenfarmer wrote:

However I still don't see the need for the selected firestones. I'd like to propose to have a normal charcoal burner as a second worker instead and add another quartz to the enhancement cost of the rare kiln instead.

A second worker would be as alien to Amazons as requiring experience for Atlanteans. face-wink.png

I may remove selected firestones though, if nobody likes the idea.

So I vote to see how the already commited changes work in the next tournamant round and if Amazons are still too strong we might have this branch in an instant.

Cool. Thank you. face-smile.png


Top Quote
WorldSavior
Avatar
Joined: 2016-10-15, 04:10
Posts: 2144
OS: Linux
Version: Recent tournament version
Ranking
One Elder of Players
Location: Germany
Posted at: 2023-10-27, 20:52

We lack testing for the Amazons of the version tournament-round-4, but in theory Amazons should now be the weakest tribe by far, so I think that things got worse.


Wanted to save the world, then I got widetracked

Top Quote
tothxa
Avatar
Topic Opener
Joined: 2021-03-24, 12:44
Posts: 501
OS: antix / Debian
Version: some new PR I'm testing
Ranking
One Elder of Players
Posted at: 2023-10-27, 21:40

WorldSavior wrote:

We lack testing for the Amazons of the version tournament-round-4, but in theory Amazons should now be the weakest tribe by far, so I think that things got worse.

Do you mean they're weakest by far in current master (included in Tournament4)? But that didn't even get tested, because everybody is still playing with the T3 version. Or had they already been the weakest before adding quartz? But then why does everybody (including you) play them? When do you mean things got worse?

Sorry, but I also don't understand why you think they're weaker. Could you please explain what makes them so weak in your opinion?

So far (current master/T4) they only need quartz for some infrastructure (mostly tools, some buildings). The "running costs" (soldier training) only have 2 quartz for 1 fully trained soldier, and that only requires 1 piece of charcoal, while other tribes need 1 coal for 1 iron. So they're not even disadvantaged by not being able to mine coal in the end.


Top Quote
WorldSavior
Avatar
Joined: 2016-10-15, 04:10
Posts: 2144
OS: Linux
Version: Recent tournament version
Ranking
One Elder of Players
Location: Germany
Posted at: 2023-10-29, 09:43

tothxa wrote:

WorldSavior wrote:

We lack testing for the Amazons of the version tournament-round-4, but in theory Amazons should now be the weakest tribe by far, so I think that things got worse.

Do you mean they're weakest by far in current master (included in Tournament4)?

Yes

But that didn't even get tested, because everybody is still playing with the T3 version.

Normally Nordfriese and hessenfarmer like to test new versions... @ them: maybe you can tell why you don't test it in the tournament?

Or had they already been the weakest before adding quartz?

No

But then why does everybody (including you) play them?

I don't like playing them with the new version

When do you mean things got worse?

When the new version was introduced. It's called balancing change, but maybe it's an imbalancing change?

Sorry, but I also don't understand why you think they're weaker. Could you please explain what makes them so weak in your opinion?

Well, the quartz. It's quite expensive and needed for a lot of tools, so the growth of the economy is supposed to be much weaker. Also, it makes their early soldier training super expensive so they are the weakest tribe in this regard.

So far (current master/T4) they only need quartz for some infrastructure (mostly tools, some buildings). The "running costs" (soldier training) only have 2 quartz for 1 fully trained soldier, and that only requires 1 piece of charcoal, while other tribes need 1 coal for 1 iron. So they're not even disadvantaged by not being able to mine coal in the end.

You might overlook that amazon coal is much more expensive than coal of other tribes.


Wanted to save the world, then I got widetracked

Top Quote
tothxa
Avatar
Topic Opener
Joined: 2021-03-24, 12:44
Posts: 501
OS: antix / Debian
Version: some new PR I'm testing
Ranking
One Elder of Players
Posted at: 2023-10-30, 01:15

Thank you for clarifying.

I won't rehash the arguments from the earlier discussion in this topic, so I only say that you seem to make the same mistake as the-x did, ie. comparing the new Amazons more to the old Amazons than to the other tribes.

As for the cost of quartz, it is indeed more expensive in total sum of basic raw materials than iron, but it needs much less processing, so the required economy size should be comparable. This is where testing is needed.


Top Quote
tothxa
Avatar
Topic Opener
Joined: 2021-03-24, 12:44
Posts: 501
OS: antix / Debian
Version: some new PR I'm testing
Ranking
One Elder of Players
Posted at: 2023-11-03, 01:30

I was finally able to create the graphics for the rare tree kiln and the charcoal master, so the follow-up PR is ready. I removed selected firestones too. Unfortunately it looks like the mirror bot's down, so the finishing changes are only available on github for now.


Top Quote
tothxa
Avatar
Topic Opener
Joined: 2021-03-24, 12:44
Posts: 501
OS: antix / Debian
Version: some new PR I'm testing
Ranking
One Elder of Players
Posted at: 2023-11-30, 02:00

I've had a test game with @the-x. I played Amazons against his Empire with current master. I'm not yet sure about the conclusion regarding the new relative strengths with quartz. On the one hand, I never had a problem with it. First I struggled with finding medium buildspaces where I needed them, so I was late with rare trees, so initially ironwood was the blocker for spear production. I was a bit late with cassava and cocoa too, but I don't think that mattered much. But later I ran into a big shortage of bread, which I noticed too late. Now, this may have been because of the increased demand for rations for the stone mines, but it may have only been that I hadn't been properly prepared.

Unrelated to quartz, but Amazon soldiers seemed too stong for the training level differences until he came with full attack trained soldiers. I've checked with the fight simulator, and it says emp a2/h1/e1 should be much stronger than amz a2, but it didn't seem so in the first fights.

I sabotaged my own later game though by letting the dressmakery use the 3 gold bars that should have been used for the training glade. Again, part inexperience with Amazons, part assuming that I'd be able to find the gold early but then actually not remembering where it was.


Attachment: application/octet-stream
2023-11-29T22.34.34_netclient_1.wry (1.4 MB)

Top Quote
tothxa
Avatar
Topic Opener
Joined: 2021-03-24, 12:44
Posts: 501
OS: antix / Debian
Version: some new PR I'm testing
Ranking
One Elder of Players
Posted at: 2023-12-14, 01:36

hessenfarmer wrote:

well I checked the new values and most of them seem reasonable now. [...]
So I vote to see how the already commited changes work in the next tournamant round and if Amazons are still too strong we might have this branch in an instant.

Based on this comment and my tests (see below), I'd like to appeal to include my updated PR in v1.2.

The current state of Amazons was merged with the hope that the tournament would provide more testing, but unfortunately that didn't happen, so it remained not widely tested. Based on some testing with the Tournament 2021 win condition (10 heroes), I've found that getting charcoal production in shape is rather painful currently. This is partly because coal should have been increased in the starting conditions, but also because charcoal production didn't get any help in the quartz PR. So I updated the PR to correct these issues, and did the test with it too. My impression was that it's much more fun to play Amazons again, in spite of the increased equipment costs.

So my conclusion is that the current state is half baked and painful to play while the PR is well rounded and nice to play, in addition to being better balanced with the other tribes.

To provide more comparative testing with the other tribes, I played test games with each tribe, first on Three Warriors, then on Escher Goose Fish building up the economies to serve one of each kind of trining site for the given tribe as well as I could make it, recruiting 20 new soldiers (30 for atlanteans to get the same total amount), and training 10 heroes. I don't present the results on Three Warriors, because that's unfair to tribes needing a lot of logs. Goose-Fish is a much easier map where all tribes can easily get what they need. So I provide my results on that. Later I continued the games up to 20 heroes, with economies mostly on auto-pilot, so that's included too. The build-ups are not perfect, but I hope they're quite reasonable. There were various hiccups and mistakes for most tribes, so the results don't purely show the actual strengths of the tribes, but to some extent my mistakes should even out.

So the final stats:

old amz amz master amz pr atl bar emp fri avg. of tribes
0 to 10 heroes
time (minutes) 129,5 155,6 146,9 143 139,9 142,6 172,4 149,0
relative to avg. -13,1% 4,4% -1,4% -4,0% -6,1% -4,3% 15,7%
land 1186 1366 1347 1461 1191 1353 1537 1377,8
rel. to avg. -13,9% -0,9% -2,2% 6,0% -13,6% -1,8% 11,6%
buildcost 829 1027 1181 1184,8 952,4 1747 1516,4 1316,32
rel. to avg. -37,0% -22,0% -10,3% -10,0% -27,6% 32,7% 15,2%
used wares 5308 7994 7173 7328 5820 6386 7965 6934,4
rel. to avg. -23,5% 15,3% 3,4% 5,7% -16,1% -7,9% 14,9%
avg. of rel. values -21,6% -0,9% -0,2% 0,5% -14,0% 2,6% 11,2%
0 to 20 heroes
time (min) 174 195,2 192,2 173,5 192,6 190,6 201,9 190,2
rel. to avg. -8,5% 2,7% 1,1% -8,7% 1,3% 0,2% 6,2%
land 1186 1366 1347 1504 1191 1387 1537 1393,2
rel. to avg. -14,9% -2,0% -3,3% 8,0% -14,5% -0,4% 10,3%
buildcost 829 1033 1200 1296,8 1032,4 2009 1678,8 1443,4
rel. to avg. -42,6% -28,4% -16,9% -10,2% -28,5% 39,2% 16,3%
used wares 7683 10587 11143 10252 9213 9977 10575 10232
rel. to avg. -24,9% 3,5% 8,9% 0,2% -10,0% -2,5% 3,4%
avg. of rel. values -23,3% -6,2% -1,2% -0,6% -11,8% 6,8% 6,8%
10 to 20 heroes
time (min) 44,5 39,6 45,3 30,5 52,7 48,0 29,5 41,2
rel. to avg. 8,0% -3,8% 10,0% -25,9% 27,9% 16,4% -28,4%
used wares 2375 2593,0 3970,0 2924,0 3393,0 3591,0 2610,0 3297,6
rel. to avg. -28,0% -21,4% 20,4% -11,3% 2,9% 8,9% -20,9%

Attachments with the replays, saves, scripts for the stats and the calculation sheet in the next posts.


Top Quote