Lessons

Short lessons about one in-game decision each. They are about how to think, not what to look up, so they stay true after the comps that illustrate them have rotated out.

Set 18, patch 18.2

Balance your three augments

Aim for one econ, one item and one combat augment. Take too many item augments and you end up putting items on units that do not change much. Take only gold and you will hit your comp but lose to a max cap board. The exception is a fast 9 line, where board strength balances it, and that one usually wants two econ augments.

What a streak is actually worth

The math is simple. If you do not hold your econ intervals, and your games go loss, win, loss, win, you finish with at least 50 gold less than a player who streaks properly. That is what an arco-iris costs you.

When to level, by the line you are on

Most lines want a level timing you commit to before the game starts, not one you improvise. If your comp caps at level 8, level at 4-2 and keep levelling on tempo from there. Dying at 7 with gold in the bank is the most common way a good board never happens: the units you need are not in a level 7 shop often enough to find them. Decide which line you are on by 3-2, then hit its timing every game, even in the games where you are behind.

Do not buy the stat your synergy already gives

Check what the synergy is already handing your carry before you build. Aphelios at Rapidfire 2 wants Guinsoo's Rageblade, because he still needs the attack speed. At Rapidfire 5 the synergy is already giving him plenty, so another attack speed item adds almost nothing, and you want Deathblade, Infinity Edge and Kraken's Fury or Striker's Flail instead. Same champion, same items on the shelf, opposite build.

aphelios_rapidfire · elderwood_aphelios

What slow rolling actually means

Slow rolling is how you hit a 3 star on a low cost unit. It is not rolling a lot: it is holding your gold at 50 and spending only what comes in above it, so you roll a few gold a turn and never break your interest. You keep earning 5 gold a turn while you search. Two things justify going below 50 and giving that up: your health is getting low, or you need the board now to hold tempo. A common rule of thumb is to slow roll until around 4-2 or 4-5, and by then you want the 3 star.

When to play for the win, and when to play the next fight

Which question you are answering changes the moment top 4 is safe. Until it is, every decision is about the very next fight: the gold, the level and the item that get you through this round, because nothing else matters if you are out. Once top 4 is guaranteed that question is answered, and a different one replaces it: how do I win this spot? That is when you spend, push the level, and take the upgrade that beats the strongest board left rather than the one that holds the place you already have. The value players lose most often is by never switching questions, and playing the whole game as though survival were still in doubt.

Levelling to 4 at 2-1 is a bet, not a default

Levelling to 4 at 2-1 is worth it when you think the board in front of you can win five in a row. It buys you the units to hold that streak, and the streak pays the level back. When you do not believe the streak is there, the level is just gold you no longer have: save it, econ up, and take your interest instead. The mistake is levelling at 2-1 out of habit and then losing the fights that were supposed to pay for it.

How a 2 cost reroll is generally played

Most 2 cost rerolls follow the same broad shape. You usually level to 6 at 3-2 and roll until you hit about three upgrades, enough that the board can hold. Then you stop, econ back up to 50, and slow roll from there for your 3 stars. The step players tend to skip is rebuilding to 50: rolling to nothing at 3-2 and staying broke is how the comp arrives two stages late, and a 3 star that lands at stage 5 has usually missed the games it was meant to win. Lines and lobbies vary, so treat this as the default you deviate from on purpose.