Dsy wrote:Goodspeed wrote:I think you overestimate the importance of such data. For balancing it's pretty simple, you just need to look at units' frequency of use and at the contexts in which they are used. Generally you can trust high level players to use strong units, and not to use weak units.
There are some issues there.
1. objective vs subjective decision making
Some high level players (its known) dont want to change anything. Even if they know a unit is useless they say: its fine. So how you deal with that?
2. even if you know what units are unused (it means probably they are bad) (plus id like to mention high level players will probably say: dont change its good)
How you change its stat?
Yeah, many units aren't viable. You can show this either with maths or simply by looking at frequency of use in high level games, which is both a more reliable and more available (in the memory of anyone who plays or follows the game) stat. Either way, if a unit is not or rarely viable that doesn't mean it should be changed. The method by which you decided that the unit isn't viable isn't relevant when deciding whether or not you should change it. In that discussion, you need to look at the impact of the viability of this unit on civ balance and on the meta, and you need to gauge the popularity of such a change.
Why is frequency of use more reliable? Because it takes into account things that a direct cost-effectiveness comparison cannot: Shadow techs, availability, effectiveness at certain micro techniques, upgrade (card) potential, compositional advantages etc. These can be very significant, as with the rifle rider which mathematically looks much worse than its frequency of use would imply.
For example here is the halberdier.
I can show it with numbers that its worse than a musket put in melee combat. Still people here say: its fine.
You can show this with numbers, but you can just as easily show it by looking at frequency of use (near zero). So how do maths help here? People don't say halbs are fine, they say they don't want to change them. That's different.
I can initially set a number (for example for its attack) where it becomes technically viable and know almost all the consequences it brings to the game without testings.
Not by doing maths. You'd estimate the consequences by having an understanding of the meta and taking an educated guess as to which game scenarios the unit would become viable in, then guessing how much of a difference the unit would make in those scenarios. No amount of maths will tell you the impact on civ balance or the meta, because these are not things you can even describe with numbers. I mean, you could, but it would take a lifetime.
Plus i mentioned it million times. Using a calculator doesnt exclude being a good player.
So the main: Why are you still refuse to use it basicly?
I get that you want to plug your tool here but the data it provides is just not as significant as you think. Historically the patch team and players in general have always found it both more convenient and more reliable to simply look at their own experience and frequency of use in high level games. Take it or leave it.