TFT ranks
Ten tiers from Iron to Challenger, how LP moves between them, and — from 99.9M recorded unit-games this set — what actually changes about the game as you climb.
The ladder
Ten tiers. Iron through Diamond are split into four divisions each, numbered from IV at the bottom to I at the top, so there are 28 steps before Master. Master, Grandmaster and Challenger have no divisions — from Master upward everyone sits in one continuous LP pool, and the two tiers above it are cutoffs rather than thresholds: a fixed number of the highest-LP players in each region, which means the LP needed moves as the ladder does.
- 01 Iron IV · III · II · I Where a new account starts if placements go badly.
- 02 Bronze IV · III · II · I
- 03 Silver IV · III · II · I
- 04 Gold IV · III · II · I The middle of the ladder by any measure.
- 05 Platinum IV · III · II · I
- 06 Emerald IV · III · II · I Added in 2023, between Platinum and Diamond.
- 07 Diamond IV · III · II · I
- 08 Master no divisions No divisions — one LP pool from here up.
- 09 Grandmaster no divisions A cutoff, not a threshold: the top slice of Master by LP.
- 10 Challenger no divisions The top slice above that. Both cutoffs move with the region's ladder.
How LP works
The top four gain LP and the bottom four lose it. That is the whole system, and it is why TFT rewards consistency over spikes: a player who reliably takes 3rd and 4th climbs steadily, while one who alternates 1st and 7th moves sideways. First and eighth are the extremes, and everything in between scales.
How much you gain is not fixed. It depends on your hidden MMR against the lobby's, so the same 4th place can be worth noticeably different amounts in two different games — and a strong player on a fresh account gains far more than the tier suggests until the two numbers meet. This is the reason to track your average placement over a rolling twenty games rather than your LP: LP mixes how you played with who else was in the lobby.
Exact LP values per placement are not published by Riot and change between patches, so you will not find them here. Anything that quotes them to the point is guessing.
Elo, MMR and why your LP gains differ
“Elo” is what players call it; what TFT actually runs on is a hidden matchmaking rating. It chooses your lobbies and sets the LP a placement is worth. When your MMR sits above your visible rank, wins pay more than losses cost and you climb quickly; when it sits below, the reverse. Your rank catches up with your MMR over time, which is why a hot streak on a new account feels so fast and a long losing run is so hard to climb back from.
Double Up and Hyper Roll ranks
Double Up has its own ladder, separate from solo ranked and using the same tier names. Your Double Up rank is personal — it moves with your team's result, but it is yours, and nothing in Double Up touches your solo rank. The comps that win the two-player queue are not the solo meta; they are on the Double Up comps page.
Hyper Roll has no LP and no Iron-to-Challenger tiers. It uses a separate Hyper rating with colour tiers — Gray, Green, Blue, Purple and Hyper — because a fifteen-minute mode with no levelling decisions is measuring something different.
Where the games actually are
This is the share of the 99.9M unit-games recorded in Set 18 so far, by the rank they were played at.
This is not a player distribution, and it is worth being clear about why. These are recorded games, and the sample comes from players who use a tracker — and tracker users are not spread evenly across the ladder. Iron shows up at 0.44% here, which is far below its real share of the player base; Master and above show up at about 5.1%, which is far above theirs. Read it as where the data on this site comes from — which is the honest and genuinely useful reading, since it tells you which rank the tier lists are really describing.
What changes as you climb
Every stats site collapses the ranks into one number before publishing a tier list. The underlying data has all ten, so here is the comparison nobody runs: Master, Grandmaster, Challenger (5.1M unit-games) against Gold and Platinum (51.9M unit-games), for how often each unit is played.
The pattern falls straight out of it. The units played far more at the top average 4.0 gold; the ones played far more in the middle average 2.0. High elo levels and plays expensive units; the middle of the ladder rerolls cheap ones. That is the single biggest behavioural difference between the two halves of the ladder, and it is not a matter of opinion.
Played more at the top All champions →
Played more in the middle Meta comps →
The part that surprises people
Look at what happens to the average placement of the units high elo prefers. Sentinel goes from 4.17 in Gold/Platinum to 4.34 at Master+ — it gets worse at the rank that plays it most.
That is not a contradiction, it is the pool. A unit played in 46% of games is contested in nearly every lobby, so the same unit yields less to each player holding it. The lesson is the one scouting keeps making: a strong comp played contested beats you less often than a weaker one played open. Copy the top of the ladder's habits — levelling, economy, positioning — rather than its unit list.
Where to go next
Where these numbers come from. Placements, win rates and pick rates are aggregated from ranked Teamfight Tactics games — the same dataset the NoxuTFT desktop app reads, refreshed every six hours. Champion costs, traits, item recipes and trait breakpoints are game data from Riot. Tier letters are relative to this set: S is the top 10% by average placement, not an absolute threshold, and champions are ranked inside their cost bracket because a 5-cost's average only looks good for having been reached from a winning position. NoxuTFT is not affiliated with Riot Games.