What rounds do you get augments in TFT?

Three times: at 2-1, 3-2, 4-2. Each time you are shown three augments and must take one. Their rarities follow a pattern fixed at the start of the game, so your first augment's rarity tells you something about the two still to come.

Should I always take the highest-rated augment?

No. Ratings measure how an augment performs across everyone who took it, and people take an augment when it fits their board. A gold augment that doubles your existing carry's damage beats a prismatic that asks for four units you don't have and can't reach. Fit first, rating as a tie-break.

What does a silver first augment mean in TFT?

That a stronger augment is coming later, because the three rarities follow a pattern. A silver at 2-1 means you should stay flexible and avoid committing to a direction on the strength of a small bonus — the game still owes you something bigger at 3-2 or 4-2.

Guide · Chapter 6 ·12 min read

TFT augments guide: how to read the offer

Three choices, three times a game, and each one quietly sets the ceiling on everything after it. Augments are the highest-leverage twenty seconds in TFT.

You will make roughly two hundred decisions in a game of TFT. Three of them are augments, and they account for a wildly disproportionate share of the outcome — because unlike every other decision, an augment cannot be undone, re-rolled or played around. You live in it for the rest of the game.

How augments work

At 2-1, 3-2, 4-2 the game pauses and offers you three augments. You pick one. Each has a rarity — silver, gold or prismatic — and in Set 18 there are 73 silver, 114 gold and 73 prismatic augments in the pool.

The crucial structural fact: the three rarities you'll be offered are decided at the start of the game, not per round. You don't know the pattern in advance, but you learn it as it reveals itself — and what your first augment's rarity tells you about the second and third is real, actionable information that most players throw away.

Rarity patterns

A few shapes recur. The pattern matters more than any individual augment, because it tells you how much of your game plan should be riding on the augment in front of you.

Pattern	What it means	How to play it
Silver → Gold → Prismatic	The escalation. Each is bigger than the last.	Stay flexible early. Don't build around the silver; it's a stat stick. The prismatic at 4-2 is the one to plan for.
Prismatic first	Your biggest augment arrives before you have a board.	Commit hard. The later two will be smaller, so this one is the game. Take the direction it points and force it.
Gold → Gold → Gold	No single game-defining card.	Three solid picks that compound. This is the pattern where economy and tempo augments shine, because nothing else is carrying you.
Prismatic last	You'll be settled by the time it lands.	Play stages 2 and 3 straight, then look for the prismatic that amplifies what you already built rather than one that asks you to change.

Read the rarity before you read the cards

A silver augment at 2-1 is telling you "don't decide yet". A prismatic at 2-1 is telling you "decide now, and mean it". That one signal changes how aggressively you should commit for the next two stages, and it's free.

What each round is actually asking

2-1 — what kind of game is this?

You have three or four units and no plan. This augment is not choosing your comp, it's choosing your shape: are you an economy game, a tempo game, or a specific-trait game? Prefer augments that stay good across many outcomes — econ, item generators, generic combat stats — over ones that name a trait you have one unit of.

The exception is a prismatic that names a trait. Those are strong enough to justify building the comp around them from nothing, and this is the only round where you have enough time left to do it.

3-2 — what am I actually playing?

By now items and units have narrowed things. This is the augment that should commit: it's the last one that arrives early enough to change your comp and late enough for you to know what your comp is.

Trait augments and unit-specific augments are at their best here. Economy augments are past their peak — an econ augment at 3-2 has ten rounds to pay back, where the same one at 2-1 had eighteen.

4-2 — what wins the fight in front of me?

Your board exists. Your comp is decided. This augment should be the biggest immediate power spike available, full stop. Economy augments are nearly worthless here. Anything that says "over the next 10 rounds" is a card you will not get paid on.

This is also where you should be most willing to take an augment that looks narrow: narrow is fine when there's no future left to be flexible for.

Why fit beats rating

Augment tier lists are the most misread data in TFT, and the reason is survivorship. Every recorded pick of an augment was made by someone who thought it fit their board. So the rating measures the augment plus the situations people take it in, not the augment.

Which produces two systematic distortions. An augment that's only ever taken when it's perfect — because it's obviously bad otherwise — reads as incredible. An augment that's frequently the least-bad of three reads as terrible even when it's a fine card. Neither number describes what happens if you take it now.

The usable version: compare the three you're offered, for the board you have. Not against the tier list, against each other. Three questions, in order:

Which of these three works with what's already on my board? Not what could be on it. What is.
Which closes the gap I actually have? If your problem is that your frontline melts, damage doesn't fix it.
Only then: which has the better numbers? Use the rating as a tie-break between two that both fit — that's exactly what it's good for.

Per-comp augment shortlists — which augments perform best specifically inside each meta comp, rather than overall — are on every comp page. That's the version of the data that answers the real question, because it's already conditioned on the board you're building. Overall numbers by round are on the augments page.

The archetypes

Specific augments change every patch; the categories don't. Knowing which box a card falls into gets you most of the way to a decision.

Type	Best at	The trap
Economy gold, interest, income	2-1	Taking one at 4-2. There aren't enough rounds left to be paid back.
Item generators components, full items, anvils	Any round	None, really. These are the safest picks in the game and are usually undervalued.
Trait augments emblems, trait bonuses	3-2	Taking one at 2-1 for a trait you have one unit of, then never finding the others.
Combat stats flat AD/AP/health for the team	4-2	Undervaluing them early because they're boring. A big stat augment is never dead.
Unit-specific one champion gets a new ability	3-2 or 4-2	Taking it for a unit you don't own yet and might not hit.
Comeback / risk pays out on a condition	Situational	Taking a condition you can't control. If the payoff needs you to be at 20 health, you are planning to be at 20 health.

The NoxuTFT augment overlay showing tier ratings on the three offered augments — **What the overlay adds — and what it can't.** It puts the tier and placement data on the three cards while the timer runs, so you're not memorising 257 augments. What it cannot do is the first two questions in this chapter: whether the augment fits *your* board and closes *your* gap. Those stay yours, which is why the rating belongs at step three and not step one. How the in-game helper grades augments →

A note on the clock

You get a short timer and it's easy to spend all of it reading three cards you've never seen. A habit that helps: eliminate first. Find the one that clearly doesn't fit and mentally discard it, then you're comparing two. Most augment mistakes aren't picking the wrong card of three — they're panicking and picking the familiar one of three.

Augments narrow your game down. Next is the chapter about what they narrow it to: choosing, flexing and pivoting.