When to Slam Items in TFT, and When to Hold Components

TL;DR: Slam when the item changes your next fights and a unit you’ll keep fielding can hold it. Waiting costs up to 3 player rounds until your next chance at a component, and that’s up to 15 HP in Stage 2, 27 in Stage 3 and 30 in Stage 4 at 3 enemy units left. A random drop only has about a 1 in 8 chance of being the exact partner you want, if each kind is equally likely. So hold when the fight won’t flip anyway, when no upgraded unit can use the item, or when a carousel you’ll pick early from is next. A wrong holder costs one Remover. A wrong recipe is what usually stays with you.

“Slam” here means building an item now instead of waiting for a better recipe. Riot’s notes use the word without defining it. The worked examples are hypotheticals: sums on Riot’s numbers, not records of real games. Riot publishes no drop odds and no data on slamming, and neither do we, so the calls here are reasoning. For what each item does, see our itemization guide.

What does a slam lock in?

Think of it as three steps, because each one closes a different door.

Where the item is	Still open	How you’d change it
Two loose components in your bag	The recipe and the holder	Nothing to undo
Combined in your bag, which Riot’s 14.23 notes allow	The holder	Put it on a unit first, since by the game’s text a Reforger acts on a unit’s items, and take a random result
On a unit	Nothing, until you act	A Magnetic Remover moves it. A Reforger changes it at random. A Salvager Wisp breaks it apart, if your shop offers one

By the game’s text, a Magnetic Remover unequips all of a champion’s items, and a Reforger unequips them and turns them at random into new ones close in type and quality. The way back to components we found in Set 18 is the Salvager Wisp: by the game’s text, units you bench lose their items and those items break apart, and Riot’s 18.2 notes refer to it doing that. It only helps if your shop offers it, so don’t plan a slam around it. The rules for the Remover and Reforger are on senseos.gg’s consumables page.

So the two parts of a slam don’t weigh the same:

• The holder is cheap to change. Riot’s 14.18 notes (September 2024) started handing out Item Removers from PvE orbs, and pitched that as a way to slam with more confidence. 14.23 then dropped one from every PvE round even if you already hold one. We found no later line that changes that, so check your bag.
• The recipe is what you’re really choosing. Once two components are one item, they aren’t two options any more.

What does holding cost?

Fights. A component in your bag doesn’t play. Your next regular chance at another component is the next carousel or PvE round. By the game’s round list, each stage from 2 on goes three player rounds, carousel, two player rounds, then PvE. So your next chance is at most 3 player rounds off. Here’s the bill if those rounds are all losses. A loss costs a flat amount by stage, and each enemy unit left adds 1, which adds up from Riot’s 14.9 and 16.1 lines to 5, 9, 10 and 13 HP in Stages 2 to 5 at 3 enemy units left:

About to play	Next chance	Player rounds before it	HP at stake, at 3 enemy units
2-1	Carousel, 2-4	3	15
2-5	PvE, 2-7	2	10
3-1	Carousel, 3-4	3	27
3-5	PvE, 3-7	2	18
4-1	Carousel, 4-4	3	30
4-5	PvE, 4-7	2	20
5-1	Carousel, 5-4	3	39

Three things to keep in mind when you read that:

• It only counts if the item flips those fights. If you’d have won or lost anyway, holding costs nothing, and that’s the first thing to check.
• There’s gold on top. A win pays 1 gold, and a lost round on 3 wins costs 2 gold by itself and 6 over the next three rounds, by the streak guide’s table.
• A PvE round isn’t a promise. Riot’s 14.5 notes hand every player orbs of equal tier and count, but the contents differ, and they can be gold or units. So you may wait through a PvE round and get no component at all.

How likely is the piece you’re waiting for?

Lower than it feels. The game’s data files list eight base components, not counting Spatula and Frying Pan. Riot prints no odds for which one drops. If each kind were equally likely, here’s the chance of getting a partner you want. This is a hypothetical, and a generous one, since it treats every drop as a component:

Partners that would do	Next random component	Within 2	Within 3
1 exact partner	13%	23%	33%
2	25%	44%	58%
3	38%	61%	76%
4	50%	75%	88%

Holding out for one exact piece from random drops misses two times in three, even across three drops. And by then you’ve played through most of a stage or more without the item. So if you’re holding, hold for several acceptable partners, not one.

The carousel is the exception, with a catch. Riot’s Space Gods review (August 21, 2026) says the ring lets you choose from almost every component, and our carousel guide covers how to pick. The catch is order. Riot’s 16.2 notes have first place picking last, and Riot describes the round as a comeback mechanic, so expect a healthy player to pick later. That lines up neatly with the rest of this page:

Your spot	Your carousel	The slam call
Winning	Later pick, so don’t count on one exact piece	Slam. You also have the most to lose by dropping a round
Losing, and the item would flip fights	Earlier pick, but you’re paying HP now	Slam
Losing, and the item wouldn’t flip fights	Earlier pick	Hold, and pick well

When should you slam?

When all of these are true:

1. It changes a fight soon. You’d lose without it and win with it, or your win streak needs it.
2. A unit you’ll keep can use it. Our flexible play guide has the rule: items follow your upgraded units, not your final board.
3. You can’t name a better recipe that’s close. “Better” is not a plan. “An early pick at 3-4, next round, taking the Tear” would be.

The later it is, the less the third test matters:

Stage	Default	What moves it
2	Slam for a 2-star. With no upgrades, holding is fine	A loss is cheap, at 5 HP to 3 enemy units, and options are worth most now
3	Slam for your best upgraded unit, unless your pick is next round and you’ll get it	Base damage has gone from 2 to 6, so waiting costs 18 to 27 HP
4	Slam everything your board can use	Your line should be known, and waiting costs 20 to 30 HP
5 and later	Slam it all	Waiting costs 26 to 39 HP in Stage 5 and more after it, few rounds remain, and a component in your bag is worth nothing when you’re out

When is holding right?

• The fight won’t flip. A single item on a weak board may not change the result. If you’re losing anyway at Stage 2, keep the options.
• Nothing can use it. A flexible slam on a 1-star 1-cost is unlikely to matter, as our flexible play guide says. An AD item on a board with no AD unit doesn’t help either.
• Your pick is next, and you’ll get it. You’re going early, on Stage 3 or before, and a specific component would make the item your upgraded carry wants. Our carousel guide covers the Stage 3 ring.
• You’re choosing between pairs. You have three components and two items’ worth of use is coming soon. Build the one you’re sure of, and hold the third.
• An augment or a trait pays you to hold. Read the card. Some effects look at what’s on your bench, as the next section shows.

Don’t hold just because an item isn’t best-in-slot. That list describes a finished board, and our item guide shows why several items can do a job. If your line changes later, a Reforger or a new carry can pick up the rest.

Which pair should you build?

Start from your upgraded unit: build what it uses. If that doesn’t settle it, prefer the item more carries can use, because it leaves you more places to go if your line changes.

A hypothetical: you hold B.F. Sword, Sparring Gloves and Tear of the Goddess, with no upgraded unit that decides it. By the game’s recipes, those make three different items. As an example, here’s how many of the 21 lines built from our own match crawl (patch 18.6 data) list each one in their carry’s build:

Pair you combine	Item	Different carries listing it	Lines listing it	What stays loose
Sword + Tear	Spear of Shojin	7	11	Gloves
Sword + Gloves	Infinity Edge	3	5	Tear
Gloves + Tear	Hand of Justice	0	0	Sword

With nothing else to go on, the Shojin keeps more doors open. But that’s a count of builds, not a rating of strength. Riot’s 18.4 notes rate Crownguard above its other tank items wherever a unit can build it, and only 1 of those 21 lines lists it. A 2-star on your board that wants Infinity Edge beats any count. When Set 19’s lines arrive, the names change and the question stays the same.

Our carousel guide has the sum for how many pairings a component adds, which helps you choose what to leave loose.

Who should hold it?

Go through these in order:

Question	If yes
Is it the carry’s item, and the carry is on your board?	Put it there
Is it a tank’s item, and your tank is 2-star?	Put it there. Our tank guide has what it needs
Will this unit be on your board next stage?	A fine holder for now
Is there no good unit yet, but you’re sure of the recipe?	Combine it in your bag, or use a stand-in and a Remover later
Does your best unit already hold items that overlap?	Try someone else. Our damage guide shows why filling an empty bucket does more than adding to a full one

A word on that last row. The damage guide’s sum is reasoning, built on Riot’s 18.4 line that Damage Amp multiplies with AD, AP and crit. We found no Riot page that sets out how every stat stacks, so read your unit’s tooltip as you add items.

A stand-in holder costs you one Remover and a little attention. That can be worth it, since an item on a 2-star you’ll replace at Stage 4 can still win fights in Stage 3. Just check your bag before you plan around it. If you’d like a second opinion on any of this in the moment, TFTSense reads the items in your bag and tells you what to slam and who should hold it.

What does that look like in a game?

Hypothetical one: winning, and tempted to wait. It’s 3-1. You’re on 3 wins with a 2-star AD unit, and you hold Sword, Gloves and Tear. You’d like to wait for a better recipe. The carousel at 3-4 is 3 player rounds away, and you’ll pick late. If the item holds those fights, you keep 27 HP at 3 enemy units, plus 6 gold over three rounds on your streak if the first would have been the one you dropped. Against that, one exact partner is a long shot from where you’ll pick. Slam now, and let the loose component find its partner later.

Hypothetical two: losing, with no one to hold it. It’s 2-5. You’re on a loss run, with only 1-star units and a Bow and a Rod. A Guinsoo’s on a 1-star wouldn’t flip these fights. Waiting costs at most 10 HP before 2-7, and you’d lose those HP anyway. You’ll also pick early at 3-4. Hold.

Hypothetical three: your line is known, and the piece isn’t there. It’s 4-1. Your carry wants a recipe you’re one component short of, and you hold two other components that make a decent second item. 30 HP is at stake before 4-4. Build the decent item now. If the piece comes, you’ll have another use for it, and if it doesn’t, you didn’t play three rounds without an item.

What in Set 18 changes this?

These are lines from Riot’s notes and the game data at each patch, so read the cards in your game.

• Opening encounters. 18.1 added encounters that start everyone with one or two Completed Item Anvils. When those come up, the whole lobby has items early, so a board with no slam falls behind faster.
• Slammin’. It’s on Set 18’s augment list in the game data, and it pays XP after each player combat if your bench holds no items other than consumables. That’s a game effect paying you directly not to hold.
• Emblems. 16.2 raised them to about 3 components of power and described them as slams for second or third carries. Riot hasn’t restated that for Set 18, so read the item.
• Tank items. Riot’s 18.4 notes changed Crownguard and Spirit Visage, so check the numbers on our item guide or in your tooltip before you pick by memory.

What changes when Set 19 lands?

Riot’s schedule lists 19.1 for December 1, 2026, and warns that dates can change. The reasoning here doesn’t rest on a set: a component doesn’t fight, a recipe is hard to change, and a holder is easy. Re-check four things in the 19.1 notes: player damage, since 16.1 changed it at a set launch; whether the carousel stays, since Set 17 replaced it; the round list; and the recipes. The pair table is an example, so rebuild it from Set 19’s lines.

What should you do on your next item?

Your spot	The call
Item flips the next fights, and an upgraded unit can use it	Slam now
Winning, and waiting for a better recipe	Slam. You pick late, and you have the most to lose
Losing, and the item wouldn’t flip anything	Hold, and pick well on the carousel
Only 1-star 1-costs to hold it	Hold, unless a unit you’ll keep can use it
Your pick is next round, you’ll get it, and it makes your carry’s item	Hold for that one round
Three components, nothing deciding it	Build what the most carries use, and keep the third
Sure of the recipe, unsure of the holder	Build it anyway. A Remover fixes the holder
Stage 4 or later, components in your bag	Build everything your board can use
Stage 5 or later	Build it all

Sources

All read November 4, 2026. Riot publishes no page on when to slam and no component drop odds, so each number links to the Riot page that sets it.

• Riot Games, patch 14.5 notes, published 2024-03-05, for loot orbs.
• Riot Games, patch 14.9 notes, published 2024-04-30, and patch 16.1 notes, published 2025-12-02, for player damage.
• Riot Games, patch 14.18 notes, published 2024-09-10, and patch 14.23 notes, published 2024-11-19, for Item Removers and combining items in the bag.
• Riot Games, patch 16.2 notes, published 2026-01-07, for first place’s last pick and Emblem power.
• Riot Games, Dev TFT: Space Gods Reviewed, published 2026-08-21, for the choice of components and the comeback mechanic.
• Riot Games, Teamfight Tactics: Hyper Roll, published 2021-04-14, for win gold.
• Riot Games, patch notes for 18.1 (2026-08-25), 18.2 (2026-09-09), 18.4 (2026-10-06) and 18.6 (2026-11-03), for encounters, the Salvager Wisp, Crownguard, Spirit Visage and Damage Amp, and the absence of player damage changes.
• Riot Games, Patch Schedule, Teamfight Tactics, for the 19.1 date.
• The game’s data files, read November 4, 2026, in a version not yet updated for 18.6, for the round list, the base components, recipes, Slammin’ and what the Magnetic Remover, Reforger and Salvager do.
• Our own match crawl: the 21 comp lines built from it (patch 18.6 data), for the counts of items. Those counts and the chances in the partner table are our own arithmetic: 1 minus (8 less the partners that would do, over 8) to the power of the drops.

Update log

• 2026-11-04: First version.