Currently Online

Latest Posts

Topic: Frisian Balancing

hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 2646
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2019-12-21, 14:11

ok changes done so far have been merged to master now.


Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 2646
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2019-12-21, 14:18

JanO wrote:

I would adjust the mathematics here, too. But yet I have no idea, where you want to calculate squares. Would it not be sufficient to just multiply the random number by a "probability correction factor" of - lets say - 1.1 or 1.2?

I don't know exactly what you mean but I beleive you meant to increase the random number we start to subtract from. this could lead to no immovable being plant at all as we might never get below zero.

So here is what I meant with using squares:
Instead of adding 70+10+10+10 +10+10 It would fit better to have summed up 4900+100+100+100+100+100.
this results in havin a probability of 4900 / 5400 approx 90% for the 70 % tree instead of 70 / 120 approx 58%.


Top Quote
JanO
Avatar
Joined: 2015-08-02, 11:56
Posts: 177
Ranking
At home in WL-forums
Posted at: 2019-12-21, 18:11

Then I got it the wrong way round and it should be multiplying the random number by 0.9 or 0.8. Calculating the squares of course would do, too - and seems very reasonable. But it means more extra-work, as you have to change both the probability and the random number. To clarify: I meant the one that you called the third step

we create a random number in range 0 to sum of probabilites.

I suggest making it 'create a random number in range 0 to (0.9*sum of probabilites)'.

Edited: 2019-12-21, 18:12

Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 2646
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2019-12-21, 18:53

JanO wrote:

Then I got it the wrong way round and it should be multiplying the random number by 0.9 or 0.8. Calculating the squares of course would do, too - and seems very reasonable. But it means more extra-work, as you have to change both the probability and the random number. To clarify: I meant the one that you called the third step

in fact we just would need to square the values while putting them into the sorted list. Rest of the code would be untouched, as the random is just a rest of a division by the sum of all stored probs.

we create a random number in range 0 to sum of probabilites.

I suggest making it 'create a random number in range 0 to (0.9*sum of probabilites)'.

that would result in an 11% increase of the prob of the best tree. In our example 65% instead of 58%. May be still too low


Top Quote
JanO
Avatar
Joined: 2015-08-02, 11:56
Posts: 177
Ranking
At home in WL-forums
Posted at: 2019-12-22, 12:45

0.9 was just a guess. If 0.8 or 0.7 is more reasonable, just pick that. You could even make the number depending on tribe, win-condition, lunar phase or whatever face-wink.png


Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 2646
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2020-01-12, 20:42

while working on the amazons I found out that frisians have significantly less hammers in starting condition HQ then other tribes. (5 in comparison to at least 11) so i would vote to increase them to at least 10, cause these shaortge limits builders which are essential for a good economy grow up.


Top Quote
WorldSavior
Avatar
Joined: 2016-10-15, 04:10
Posts: 2091
OS: Linux
Version: Recent tournament version
Ranking
One Elder of Players
Location: Germany
Posted at: 2020-01-12, 21:18

hessenfarmer wrote:

i would vote to increase them to at least 10, cause these shaortge limits builders which are essential for a good economy grow up.

sounds reasonable


Wanted to save the world, then I got widetracked

Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 2646
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2020-02-17, 17:33

With the hammers fixed now in our master branch, I feel we could remove the experimental mark from frisian.
Although there may still be hidden problems and balöancing issues I beleive they have reached a state of maturity that would allow us to do so.
Opinions?


Top Quote
Nordfriese
Avatar
Joined: 2017-01-17, 18:07
Posts: 1928
OS: Debian Testing
Version: Latest master
Ranking
One Elder of Players
Location: 0x55555d3a34c0
Posted at: 2020-02-17, 18:34

+1


Top Quote
the-x
Avatar
Joined: 2019-01-19, 13:23
Posts: 967
Ranking
One Elder of Players
Posted at: 2020-02-17, 18:35

+1


Top Quote