Most League games are lost at the shop, not in lane. You open a build site before champ select, copy the six items in order, then buy them in that order for the next 30 minutes while the enemy Dr. Mundo stacks health and your third crit item does nothing to him.
So here is the short answer. Your champion decides your first two or three core items. The enemy team decides everything after that, and you make that call again at every single back. A build in League of Legends is not a shopping list you lock in at champion select, it is a series of small bets placed with the gold you actually walked back with.
Start With Your Champion's Core Build, Not the Full Six Items
Core items are the two or three items your champion buys in nearly every game, whatever the enemy picks. They are not a preference, they are what makes your kit function: the crit a marksman needs before autos are worth walking up for, the haste a mage needs before the combo actually kills, the health a tank needs to survive their own engage.
Only slots three to six are genuinely up for debate. That is the part nobody teaches, and it is the part that decides games.
What your core chases depends on your archetype:
Marksmen: attack speed and crit, then penetration once armor appears
Assassins: lethality or magic pen, enough to kill before they react
Mages: AP and ability haste, penetration once magic resist shows up
Bruisers: AD with health and haste, so you survive long enough to deal it
Tanks: health, then whichever resistance their damage demands
Enchanters: haste, heal and shield power, actives your allies feel
If you cannot say out loud why your first item is your first item, that is the gap worth closing, and lol-coaching.net lists coaches by role and rank.
Where to Get a Starting Build, and How to Read Tier Lists and Win Rates
Pull a starting build from a stats site. u.gg is fastest for patch-fresh items, runes and skill order. Mobalytics has the cleanest champion guides plus an app giving live recommendations. OP.GG is best for a player's match history, METAsrc has the widest raw data, probuilds shows what pros locked in last night.
Read the numbers properly. Set the rank filter to Emerald+ rather than all ranks, because a low-elo sample records what people buy, not what wins. Check sample size before believing a win rate: 57% across 300 matches is noise, 52% across 40,000 is signal. A tier list is a pick aid, not a build. It says the champ is strong this patch, not what to buy at 14 minutes.
Treat probuilds with the most suspicion. That build came from a far higher-skill solo queue game with different champion pools and different win conditions, so their first item is often nonsense at your rank.
Runes, Summoner Spells and Skill Order Are Part of the Build
Your build is not only items. Keystone and secondary tree carry the same weight as your second item, and so do the shards: adaptive force when you want damage online now, health or tenacity against burst and crowd control.
Summoner spells follow role and matchup. Ignite comes with 40% healing reduction attached, often the fastest answer to a sustain laner, and Exhaust turns an assassin's all-in into a coin flip.
Skill order flexes too. Max the ability that wins your lane, usually the poke or wave clear spell into a longer-ranged laner, not the one the site recommends at every level.
The Four Build Questions to Ask at Every Back
Every time you press B the game hands you a decision. Run these four questions in order and your build will diverge from the site's in two games out of three.
1. What Is Actually Killing Me, Physical or Magic Damage?
Open the scoreboard and read their items, not their class. A Kai'Sa going on-hit AP, a Brand support and a Kayle with Nashor's is a magic damage team whatever the class list says. Buy against the damage genuinely taking your health bar, then check again ten minutes later, because their build changed too.
The plain version of effective health: armor and magic resist do not add health, they multiply the health you already have. So the right buy depends on what you own. Big health pool with thin resistances, take the resistance. High armor on a small health bar, take health. Against true damage neither resistance does anything, so health is the only defense on sale.
2. Are They Healing or Shielding Through My Damage?
Anti-heal is the most-skipped correct purchase in low elo. Grievous Wounds cuts healing by 40% and the components cost 800 gold, so no game exists where you cannot afford one against Soraka, Aatrox, Vladimir or Mundo. Buy it the back their sustain starts winning trades, not as a sixth item after the game is decided.
Shields are a separate problem. Grievous Wounds does not touch them, so you need burst that outpaces the shield, or Serpent's Fang at 2,500 gold, which cuts shields by 50% if you are melee, 35% if you are ranged.
Anti-heal buy | Cost | Stats | Suits | Upgrades into |
|---|---|---|---|---|
Executioner's Calling | 800g | 15 attack damage | AD carries, bruisers, junglers | Mortal Reminder or Chempunk Chainsword |
Oblivion Orb | 800g | 25 ability power | Mages, AP assassins | Morellonomicon |
Bramble Vest | 800g | 30 armor | Tanks holding an auto-attacker | Thornmail |
Morellonomicon | 2,850g | 75 AP, 350 health, 15 haste | AP champs spending a full slot | Finished form of Oblivion Orb |
3. Am I Ahead, Even, or Behind?
Ahead, buy the item that ends the game before enemy carries scale: damage, tempo, whatever takes the next tower now. Behind, buy the cheap thing that keeps you alive and farming, because a 3,500 gold spike you die twice before finishing is worth zero.
This is the back-with-what-you-have rule. Walk in with 1,200 gold and boots at 300 plus an 800 gold anti-heal component beats leaving with 1,200 unspent and a dream. Gold in your pocket has no stats, and dying with 1,600 banked is how low elo players donate teamfights.
4. What Does My Team Not Have?
Look at your four allies and name what the teamfight is missing: frontline, engage, peel, disengage. If nobody can hold a fight, your fifth slot solves it. A tank item on a bruiser wins more games than a fifth damage item that only matters if you survive to second four.
One concrete pivot. You are an ADC two items in, and the enemy Malzahar has ended four fights by suppressing you on sight. The build says Infinity Edge next. Instead you back on 1,400 gold, buy Quicksilver Sash (1,300 gold, 30 magic resist, cleanses the suppression), and finish the crit item next back. That trade is what ADC coaching sessions get spent on.
Build for Power Spikes, Not the Perfect Six Items
The full six-item build almost never happens. Games are decided by who is stronger at 14 minutes and at 22 minutes, so build for those two clocks instead of a fantasy inventory.
That means finishing the component that spikes now over hoarding for the ideal item later, and it means boots on time. Boots are their own decision this season: 2026 role quests hand the mid laner a free tier 3 upgrade, while the completed bot lane quest moves boots into a quest slot and quietly gives marksmen a seventh item.
The crit carry is the cleanest example. Base crit damage sits at 200% this season, and Infinity Edge costs 3,500 gold for 75 AD, 25% crit and +30% crit damage, pushing crits to 230%: biggest spike in the build, slowest to reach. If grubs and dragon fights land at 12 minutes, the cheap attack speed spike (Berserker's Greaves went from 25% to 30% in patch 26.16) wins the objective IE would have won at 17.
Build Mistakes That Cost Low-Elo Games
Copying a probuild into a completely different draft
Full damage into a three-tank comp with no penetration anywhere
Never buying anti-heal, then calling the enemy Aatrox broken
Buying resistances at 12 minutes against a carry who is 0/6
Dying with 1,600 unspent gold, repeatedly
Tier one boots at 25 minutes
Picking an item for its stat line and ignoring the passive, which is usually the whole reason it exists
Running last season's build after a patch deleted the item, and 26.9 removed Opportunity and Trailblazer outright
Your rank changes which of these matters. Iron to Gold, the win rate lives in boring stuff: finish boots, buy the 800 gold anti-heal, spend your gold before you fight. Emerald and above, that is assumed, and the edge moves to matchup-specific first backs and landing your spike before a dragon rather than after it.
Turn Itemization Into a Habit You Can Check
Make it measurable. Load one replay from last night and write two columns: the game clock at every back, and the clock when each item completed. Then mark every purchase after your core as either a reaction to the enemy team or autopilot.
Two numbers jump out fast. How much gold you walked back to lane with unspent, and how late your first completed item landed. After 14 minutes in a normal game, your problem is spending, not farming.
Three games of that gives you a list of habits, which is what a VOD review diagnoses in one sitting. A coach pauses on the back where Executioner's Calling should have gone in and builds a personalised improvement plan from your own replays, so get your VODs reviewed by a Challenger coach when itemization stops improving on its own.
Next game, before you buy anything, run the four questions. Two backs in you will already be playing a different build than the one you copied.
