Optimizing Character Builds in Genshin Impact
Most people approach artifact optimization backwards. They pull a new character, run a dozen domain runs, and then realize they've been mainifying energy recharge on a hypercarry who never triggers their burst. It's frustrating and it wastes resin. The actual process is simpler once you stop thinking about stats in isolation.Here's how I actually set up a new build when a character drops: I determine the role first, then the stat priorities, then farm accordingly. Not the other way around. The trick nobody talks about is that most characters don't need perfect artifacts to be functional. They need correct artifacts. A 4-piece Noblesse Oblige on Yelan with four random substats will still outperform a 2+2 setup with perfectly rolled pieces, because the set bonus is a flat multiplier to everything she does. Beginners obsess over mainstats and ignore set bonuses entirely. That's the first mistake. Second mistake: treating artifact substats as equally valuable across all slots. They're not. Sands is a binary choice between HP%, ATK%, or ER. Goblets are element-specific or ATK%. Cups are almost always DEF% or ATK%. The only slots where you're genuinely comparing substats head-to-head are the flower and plumage, since those rolls directly affect your top priority stat. This means when evaluating a piece, I check the mainstat first, then the set, then see if any of the three substats are in the "good" range rather than trying to optimize a single 5-star piece at the expense of two other viable ones.
I ran into a specific problem with Kazuha recently. Everyone tells you to stack swirl damage and Elemental Mastery on him. Standard build. I did that, ran Spiral Abyss, and realized my damage was underwhelming compared to other players running nearly the same builds. The issue was team composition. My team already had ample EM support from other characters' passives and weapons, making my personal EM investment heavily diluted. The workaround was switching to an Anemo Skill Damage set with an ATK% goblet instead of Elemental Master, which actually increased my overall contribution because it played to where my team was weak rather than stacking onto what was already strong.
The Resin Math Nobody Does
One domain run takes roughly 90 seconds. With 160 resin per day, you get about 10 runs. That sounds like plenty until you realize a single set requires 24 unique artifact pieces to complete. Even assuming perfect luck — which never happens — you're looking at dozens of domain days for one full set. In practice, the math works out differently. I track my drop rates over rolling batches of 20 domain runs per set. The data is consistent: you'll get a usable 4-piece with reasonable mainstats within about 60 to 80 domain runs if you don't obsess over perfection. That's roughly a week to ten days of daily resin spending. Anything beyond that point is diminishing returns unless you're already in the top percentile of players chasing tournament-grade numbers.
Get the Full Details
Weapon Priority Over Artifacts
This is the counter-intuitive part that separates players who hit floor clears from those who barely do. A character with a 4-star weapon and half-optimized artifacts will regularly outperform the same character with a signature 5-star weapon and fully refined artifacts. The weapon provides a flat baseline boost to base stats and often a passive that directly amplifies the character's primary damage source. Artifacts are multiplicative on top of that, but you can't multiply zero. I specifically remember struggling with Xingqiu for months because I kept farming artifacts for him before securing him a decent weapon. Ran Domain of Guyun maybe forty times. Nothing gelled. Then I pulled a R1 Singleton for him, equipped whatever artifacts I'd farmed without caring about substats, and suddenly my shields were actually meaningful in Domo. The weapon passive alone gave him more consistent damage output than the artifact upgrades ever would have at that stage.
Substat Tiers That Actually Matter
Not every substat roll is equal. Here's what I use as a practical threshold system: Excellent rolls: Critical Rate above 6.2%, Critical Damage above 12.4%, Elemental Mastery above 14, ATK% above 5.5%, HP% above 5.5%, DEF% above 5.5%. These are the rolls that make a piece worth keeping long-term. Decent rolls: Critical Rate at 4.7-6.1%, Critical Damage at 9.3-12.3%, any ER roll above 4.1%. These are acceptable on pieces you're using while waiting for better options.
Dead weight: Anything below the thresholds above, plus any rolls in stats the character doesn't use. A Hydro DPS with three DEF% substats is not "close to good." It's replacement material. The mistake beginners make is investing enhancement materials into decent pieces instead of selling them and using the mats for excellent ones. Ten pieces of decent artifacts don't equal one excellent piece. Period.
When to Stop Farming a Character
This is probably the most important piece of advice and the hardest to follow. You need a hard rule for when a character investment is sufficient. My rule: if the character can clear content they were designed for with their current build, and the current build meets the minimum stat thresholds for their role, stop. Do not keep farming. Move to the next character. I've seen players spend three consecutive versions farming artifacts for a 4-star character who was already clearing everything. The resin could have been spent on a new 5-star that was coming next patch. Opportunity cost is real and most players ignore it completely. There are exceptions. If you're consistently failing floors in Spiral Abyss that most casual players clear, that's a signal the build genuinely needs work. But for 95% of content — Domovoi Boss, Ruin Graders, weekly bosses, most Abyss floors — a "good enough" build is indistinguishable from a "perfect" build in actual gameplay. The difference only shows up in frame-perfect speedruns or leaderboard chases.
Constellation Math
Some constellations are statistically more valuable than half your artifact set. C2 Bennett gives his burst a flat buff that scales with his max HP. If you already have high HP on your artifacts, that constellation is effectively giving you another 40-50% ATK on everyone in your team. That's not an exaggeration. It's a direct multiplication of damage output. Meanwhile, C6 Zhongli is primarily a quality-of-life change that removes his shield's downtime window. For most players who already have spare resin, the resonance value is negligible. Understanding which constellations actually change a character's damage profile versus which ones just polish existing functionality saves a lot of primogens. The practical takeaway: if you're considering spending primogens on a character, look at their constellation chart before buying anything. C0 to C1 is usually where the biggest jump happens. C2 to C3 is often small. C6 is usually cosmetic at that point.
Team Synergy Over Individual Builds
The strongest builds are the ones that make other characters better. Baizhu's passive healing scales with his max HP, which means investing in HP on him indirectly strengthens whoever he's healing. Faruzan's shield from her Elemental Skill is stronger when she has high Energy Recharge because she can use it more often. These are invisible contributions that don't show up in damage meters but literally determine whether a team clears a floor or doesn't. I stopped measuring my builds by individual damage output about two years ago. Now I measure by team survival rate and floor clear consistency. The numbers don't lie. Teams built with synergy in mind consistently clear floors that solo-damage-focused teams fail on, even when the damage-focused team has slightly higher DPS numbers on paper. That's basically it. Farm smart, stop farming when it's enough, prioritize weapons before late-game artifact upgrades, and build teams that cover each other's weaknesses rather than stacking the same stat across five characters. Most of the frustration with builds comes from investing in the wrong direction, not from a lack of effort.
