how many times have i said now that we don’t have enough data to determine balance?
That does not matter, it is still a sample which is in this case equal to the total population. This is called total population sampling
The entire data available can be the sample too.
If i have a fair coin and we flip the coin 300 times. It gives us 145 heads (win) and 155 teals (lose). You get those data and you say no the coin is not fair, the probably it lands on head is just 48,3%, not 50% based on the data. You flip the coin even more. At 1000 flips you get 508 heads and 492 teals. Now you got the data and you say the same coin has now a probability of 50,8%. So it is still no fair coin, but now it is unfair the other way around?
In the end, if we could play infinite many games, we will end up with 50% winrate, if the coin is indeed fair. As long as we dont have infinite many games played (we wont get to that point), the winrate will be around 50% if the coint is indeed fair.
At both points in time we had a look at the full available data. We did not change the coin (like we look at the situation where we didnt do balance changes). The winrate went up from 48,3% to 50,8%. So there is uncertainty about the current winrates. In math we can use CI to say something about this uncertainty.
I fully agree. I think we needs the data from some months to really can say something about it. So i dont really like it if the devs keeps changing the same civ every month. Just change the civ and have a look about the civ 2-3 months later. If needed you can have another change after 2-3 months.
2 Likes
When did I pester you on bringing data? Please point to that instance, before making accusations.
Keep the discussion civil(non-personal), and this is for the fourth time, make constructive discussion on the topic given.
Balance decisions like there are not meant to be made on civ winrate data, but on subjective terms.
2 Likes
Why do we need to neglict objective data? I think we have to look at objective data and subjective terms. Both. Objective data shows us what civs are OP, but part is also psychology and thus subjective.
I have heard about stories were the devs said they nerfed something, but in fact they didnt, but the winrate and pick rate for the civ dropped. Just based of a false statement of the devs (they wanna nerf something, but is was just nog part of the patch). In the end they even didnt nerf it at all. I dont know more about this, i even dont no the game.
umm, no. you definitely take data into account when balancing. the thing is though, 2 weeks isn’t enough time to collect adequate data.
Are we playing a different game?
With Castle crossbows and range/dmg upgrade and my scout i have no problem to fight against feudal skrimishers. Thats one of the reasons why archer rush into crossbow works, if feudal skrimishers would hard counter crossbows, this wouldnt be possible without mixed in knights.
(my definition for hardcounter is beating 2x+ the costs with the counter unit, like elite huskarls do against archer or teutonic knights do against infantry, camels and normal pikes arent hard counter for me since you dont beat 10k ressources of knights with 5k ressources of camels or pikes)
2 Likes
Feudal Skirms with Castle Age upgrades do counter Crossbows easily.
You can also go for Mangonels, which destroy Crossbows.
Sorry i meant “Balance decisions like these are done subjectively” , i.e. that balancing of unit lines is not done by Civ Winrate data.
Yes it is. I agree with 100 F and 75 G increase.
1 Like
No. Crosbows are perfectly balanced. They are a good unit but they still get countered by many units.
Any kind of buff to longswordsman will be a direct nerf for meso civs aswell.
1 Like
Only Mayan since Incas and Aztecs would use Champs themselves when relevant (as they already do)
2 Likes
I think Long swordsman are supposed to be situational. Archers have range, cavalry have speed, swordsman are great against trash and eagle warriors. Given the introduction of supplies we have seen a slight increase in infantry, but it’s still not a ubiquitous pick like scouts or knights or xbow is - it’s situationally called for. I think even if Long swordsman was half the cost, we’d still see Long swordsman used about as much as they are currently because you’re sacrificing range and speed for… swordsman.
If supplies was cheaper and the longswordman amd champions upgrade was a bit faster to search I think going longswords + mangonel would be actually decent if you play as japanise or vikings maybe.
As someone said incas still have those slingers and aztec would go for jaguars more often.
You know, I didn’t realize that the LS upgrade took 75 seconds while Xbow took 50. I think supplies should stay where it is - but an argument can definitely be made by making Ls slightly faster to research (in exchange for making 2-h or Champion slightly longer ofc). I think Xbowman should still be faster than LS, but having 15 seconds to clean up men at arms should still be plenty.
1 Like
11 again with turks xD if something lose to turks its balanced xD bythe way u cant counter crosbows with turks skirmish bro not even if u fully upgrade these garbage turk unit. u are calling turk unit with a word ‘‘amazing’’ omg !! u cant counter anything with those garbage. i tried now on editor again and again its not even close as turks range and attack is awfull,awkward,silly
1 Like
You easily can.
Skirmishers with the second armour upgrade, counter Crossbows.
with 1 less range and poor HP, all it takes is some micro to deal with regular skirms
1 Like
i tried in editor even against aı i couldnt counter effectively against generic xbows (15vs15) thank you
1 Like
In imp turk skirms are not the worst but in castle age there is absolutely no use for them. Even after getting the second armor upgrade you still lack armor compared to all the other civs (except for having the same armor as the infamous Burmese skirms) while also having less range and hp.
4 Likes