Latest Posts

Topic: Reducing soldier strength parameter space (for easier comparison).

Forums » Player Forums » Playing Widelands » Reducing soldier strength parameter space (for easier comparison).
WorldSavior
Avatar
Joined: 2016-10-15, 04:10
Posts: 2112
OS: Linux
Version: Recent tournament version
Ranking
One Elder of Players
Location: Germany
Posted at: 2018-09-20, 23:03

hessenfarmer wrote:

WorldSavior wrote:

By the way, if that program is not wrong and I've not miscalculated it could be possible to balance the values by increasing the maximal attack of Fri0s and decreasing their minimal attack.

I don't understand this, cause the meridian of attach will stay the same if increased and decreases by the same value. Perhaps you could provide your values for test.

Because the values are close to each other, the probability is high that soldiers need one more hit even though the opponent is almost dead. If the difference between the values increases this probability gets lower.

I don't have some values, but I experimented a bit and it seemed to work.

Wanted to save the world, then I got widetracked

Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 2727
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2018-09-20, 23:37

I have experimented a bit and increasing the increase value per level for defence from 15 to 18 would deliver the following values:

vs. bar_00 emp_00 atl_00 fri_00 bar_10 emp_10 atl_10 fri_10
bar_00 56.9% 41.6% 31.6% 39.0% 0.0% 0.0% 0.0% 0.0%
emp_00 70.3% 56.1% 43.0% 47.4% 0.0% 0.0% 0.0% 0.0%
atl_00 78.5% 68.0% 55.7% 52.8% 0.0% 0.0% 0.0% 0.0%
fri_00 72.1% 63.5% 57.9% 55.2% 0.0% 0.0% 0.0% 0.0%
bar_10 100.0% 100.0% 100.0% 100.0% 54.1% 52.3% 50.4% 49.0%
emp_10 100.0% 100.0% 100.0% 100.0% 55.3% 53.5% 49.3% 48.3%
atl_10 100.0% 100.0% 100.0% 100.0% 56.6% 57.2% 53.0% 52.6%
fri_10 100.0% 100.0% 100.0% 100.0% 61.7% 60.1% 55.4% 57.9%

For me they look acceptable. Opinions?

Top Quote
einstein13
Avatar
Joined: 2013-07-29, 00:01
Posts: 1116
Ranking
One Elder of Players
Location: Poland
Posted at: 2018-09-21, 09:05

61.7% is way too high. Now Frisians are the strongest soldiers. Almost 2 out of 3 fights are win.

What is your target for Frisians? What balance do you need?

einstein13
calculations & maps packages: http://wuatek.no-ip.org/~rak/widelands/
backup website files: http://kartezjusz.ddns.net/upload/widelands/

Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 2727
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2018-09-21, 22:07

einstein13 wrote:

What is your target for Frisians? What balance do you need?

Well, I would say the y should be in the range of atlantean soldiers maybe slightly better. I have experimented with my excel this afternoon and I got the following results leaving the level 0 values untouched.

vs. bar_10 emp_10 atl_10 fri_10
bar_10 54.0% 52.3% 50.7% 49.7%
emp_10 55.4% 53.2% 48.9% 54.1%
atl_10 56.6% 57.1% 53.2% 52.7%
fri_10 60.6% 54.9% 56.1% 59.0%
Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 2727
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2018-09-21, 23:10

further finetuning . Opinions?

vs. bar_10 emp_10 atl_10 fri_10
bar_10 54.2% 52.0% 51.1% 50.8%
emp_10 55.1% 53.2% 48.9% 54.0%
atl_10 56.6% 57.2% 52.9% 53.0%
fri_10 59.7% 54.9% 56.0% 58.8%
Top Quote
king_of_nowhere
Avatar
Joined: 2014-09-15, 18:35
Posts: 1668
Ranking
One Elder of Players
 Posted at: 2018-09-22, 06:54

einstein13 wrote:

61.7% is way too high. Now Frisians are the strongest soldiers. Almost 2 out of 3 fights are win.

What is your target for Frisians? What balance do you need?

don't forget that''s the caseof the frisian starting first, while if the barbarian start first the odds are closer to 50%. So total odds are around 56% for the frisian, which is fine.

it is no surprise that the two tribes with the lowest evade are also the ones where attacking first makes the greater difference.

Which brings me to my suggestion for hessenfarmer: could you give us an average value instead of two values depending on who attack first, and therefore leaving half the table blank? it would be easier to read

Top Quote
Nordfriese
Avatar
Joined: 2017-01-17, 18:07
Posts: 2044
OS: Debian Testing
Version: Latest master
Ranking
One Elder of Players
Location: 0x55555d3a34c0
Posted at: 2018-09-23, 13:51

hessenfarmer wrote:

further finetuning . Opinions?

vs. bar_10 emp_10 atl_10 fri_10
bar_10 54.2% 52.0% 51.1% 50.8%
emp_10 55.1% 53.2% 48.9% 54.0%
atl_10 56.6% 57.2% 52.9% 53.0%
fri_10 59.7% 54.9% 56.0% 58.8%

I think some of these chances are still a bit too high. IMHO we should aim that none of the frisian values are much higher than the highest chance of the other three tribes, that is, keeping them well above 52% but below 58%.

Top Quote
WorldSavior
Avatar
Joined: 2016-10-15, 04:10
Posts: 2112
OS: Linux
Version: Recent tournament version
Ranking
One Elder of Players
Location: Germany
Posted at: 2018-09-23, 16:26

I agree with Nordfriese, I would like it if there would be values which are even better...

Wanted to save the world, then I got widetracked

Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 2727
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2018-09-23, 18:08

Ok I'll do my very best fiddling with the numbers. It is not that easy without touching the other tribes. And touching them is not an option as it would be much more work on top.

Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 2727
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2018-09-23, 20:55

Now I reduced frisian health uograde a little further. the results in einsteins program are as follows: (with the uncertainty due to only 10000 fights each)

Battles win:

vs. bar_10 emp_10 atl_10 fri_10
bar_10 54.1% 52.2% 51.1% 52.3%
emp_10 55.4% 53.5% 49.4% 54.2%
atl_10 56.7% 57.4% 53.4% 52.9%
fri_10 58.2% 55.0% 55.9% 58.8%

Attackers health:

vs. bar_10 emp_10 atl_10 fri_10
bar_10 19.6% 19.6% 19.3% 14.7%
emp_10 22.3% 19.6% 17.7% 18.9%
atl_10 24.5% 21.7% 19.7% 21.1%
fri_10 19.4% 21.1% 22.2% 23.5%

Defenders health:

vs. bar_10 emp_10 atl_10 fri_10
bar_10 14.7% 17.7% 19.7% 13.3%
emp_10 15.3% 15.9% 17.7% 15.4%
atl_10 14.8% 14.1% 15.9% 16.3%
fri_10 10.2% 14.0% 16.0% 14.4%
Edited: 2018-09-23, 20:55
Top Quote