big data-like algorithmic balance helper, when?

"

Draethorian123#0716 a écrit :

Yeah so, idk, we all know that you think that you're good at balancing the game and yeah let's not compare to d4 or anything, but man, when players on day one of patch release, know that armor wont be able to compete with evasion, or that one bow build will again outshine everything, i think it's time you need an algorithmic helper to objectively and quantitatively back your decisions.

have an algorithm that goes through every combination of skills, gear, passives etc for different level thresholds (every 5 levels or maybe more in early and less in late game, take the average of gear stats), aiming for sensible synergies, to at least be able to have a build power comparison to at least catch some outliers.
because currently it mostly feels vibes based. and your vibes almost always favour bow builds while also almost always disregarding melee.

even worse, a lot of skills in poe1 are just dead and a mindfield of wasted time it seems, and idk, you could buff the underdogs the same time you nerf jungroan ;-)

this shouldnt happen. if gamers figure out the most broken build within the first 2 days your qa on balance just isnt up to par.

i know it's hard and i know its a nigh impossible task, but do yourselves a favour and at least admit that you need better diagnostics/analytics/howeverItsCalled tools for this beast of a challenge.

because it will only get harder, not easier. and it will also help you free up more time to introduce more mechanics or focus on other tasks.

those tools most likely need a long time to reach the proper maturity to be useful and i hope that youve already started thinking about an improvement to what you're currently using, but big data, hadoop, etc aren't totally new topics, so consider this a nudge towards the inevitable ;-)

"

Xzorn#7046 a écrit :

Designing Skills is actually kinda easy and yet they fail spectacularly at it.

It's essentially just a formula that considers damage, AoE (if any), base size of AoE, based radius 3+, status, type of status, cone size/angle, Channeling, chain, fork and similar then apply the result to the base animation speed to get the intended point value.

Long as your skills hit your intended point value you should be no more than 15% off.

I've designed over 300 skills in my time and never been as far off as they get unless there was a bug. They just don't seem to bother. If they were we wouldn't get such comical differences.

"

Legioona#1076 a écrit :

"

Xzorn#7046 a écrit :

Designing Skills is actually kinda easy and yet they fail spectacularly at it.

It's essentially just a formula that considers damage, AoE (if any), base size of AoE, based radius 3+, status, type of status, cone size/angle, Channeling, chain, fork and similar then apply the result to the base animation speed to get the intended point value.

Long as your skills hit your intended point value you should be no more than 15% off.

I've designed over 300 skills in my time and never been as far off as they get unless there was a bug. They just don't seem to bother. If they were we wouldn't get such comical differences.

Yep, it's a formula game.

Somehow GGG fails to understand why lightning arrow is so prevailing -> it nearly always delivers maximum damage when there are enough enemies, and delivers it instantly -> maximum dps.