Deadlock tier list
The current meta, computed from ranked match data
Rat King, Graves, Paige, Seven, Victor currently top the Deadlock tier list. Heroes are ranked by the lower bound of a 95% confidence interval on their win rate, not by raw win rate, so a hero with 300 matches cannot outrank one with 300,000 on a lucky streak.
S tier
A tier
| Hero | Vs 50% | Win rate | Pick rate | Matches |
|---|---|---|---|---|
| Haze | 54.1% | 52.7% | 905k | |
| Kelvin | 53.9% | 17.3% | 298k | |
| Ivy | 52.9% | 26.4% | 454k | |
| Dynamo | 52.5% | 27.9% | 479k | |
| Abrams | 51.2% | 30.7% | 527k | |
| Lash | 51.1% | 43.0% | 740k | |
| Mo & Krill | 51.1% | 28.3% | 486k | |
| Lady Geist | 50.9% | 28.0% | 481k |
B tier
| Hero | Vs 50% | Win rate | Pick rate | Matches |
|---|---|---|---|---|
| Wraith | 50.3% | 33.4% | 574k | |
| Vindicta | 50.3% | 32.7% | 563k | |
| Drifter | 50.1% | 43.4% | 746k | |
| Apollo | 49.9% | 27.9% | 479k | |
| Calico | 49.8% | 22.7% | 390k | |
| Celeste | 49.1% | 31.3% | 538k | |
| Infernus | 48.9% | 43.1% | 741k | |
| Vyper | 48.5% | 17.6% | 303k | |
| Yamato | 48.4% | 27.8% | 478k | |
| McGinnis | 48.2% | 14.5% | 249k | |
| Holliday | 48.0% | 22.9% | 394k | |
| Billy | 47.7% | 32.3% | 556k | |
| Warden | 47.7% | 15.8% | 272k |
C tier
| Hero | Vs 50% | Win rate | Pick rate | Matches |
|---|---|---|---|---|
| Grey Talon | 47.3% | 16.9% | 291k | |
| Bebop | 47.2% | 43.1% | 741k | |
| Rem | 47.1% | 39.9% | 687k | |
| Venator | 46.6% | 45.3% | 778k | |
| Mina | 46.5% | 42.3% | 726k | |
| Silver | 46.1% | 30.9% | 531k | |
| Paradox | 46.0% | 30.2% | 519k | |
| Viscous | 46.0% | 17.0% | 292k | |
| The Doorman | 45.9% | 19.7% | 338k | |
| Sinclair | 45.8% | 17.5% | 300k | |
| 45.4% | 23.8% | 409k | ||
| Shiv | 45.0% | 32.1% | 552k | |
| Solomon | 44.6% | 14.1% | 242k | |
| Mirage | 44.4% | 15.0% | 257k |
D tier
| Hero | Vs 50% | Win rate | Pick rate | Matches |
|---|---|---|---|---|
| Deadman Danny | 36.0% | 0.9% | 16k |
How far apart the roster actually is
Rat King holds the highest raw win rate at 57.8% and Deadman Danny the lowest at 36.0% — a spread of 21.8 percentage points across 41 heroes, around a mean of 49.1% with a standard deviation of 4.0 points.
That is wider than the "every hero sits at 50%" intuition suggests. The average is pinned near even by arithmetic — each match produces one winner and one loser — but the roster around it is not tightly packed: 15 of 41 heroes fall within two points of even, which leaves the rest meaningfully clear of it in one direction or the other. A hero at 54% is roughly a standard deviation and a half above the mean, not a rounding error. Popularity is a separate axis — Haze is the most-picked hero at 52.7% and sits in A tier, which is the reminder that what people play and what wins are different questions.
What the confidence adjustment is doing right now
Honestly: very little, and that is the correct result. Ranking here is by the lower bound of a 95% confidence interval rather than by raw win rate, but the smallest sample on the board is 16k matches — at that size the interval is narrow for every hero, the bound sits just under the raw rate for all of them, and the two orderings agree.
It is kept because it is a guard, not a decoration. It is what stops a newly released hero with a few thousand matches, or one played almost exclusively by specialists, from taking the top of the list on a streak that has not been tested. Deadlock has no such hero at the moment. When it next adds one, this page will not need changing — which is the difference between a stated method and a hand-placed ranking.
Method
- Ranked matches from the last 30 days, from the community Deadlock API's match dataset.
- Heroes with fewer than 200 matches are excluded rather than ranked on noise.
- Ordering is by Wilson score lower bound at 95% confidence, which penalises small samples in proportion to how small they are.
- Tiers are assigned by standard deviations from the roster mean. Deadlock's win rates sit in a narrow band, so fixed cutoffs would produce an empty S tier one week and a twelve-hero S tier the next.
Common questions
- Who are the best characters in Deadlock right now?
- The S-tier heroes on this page have the highest confidence-adjusted win rates over the last 30 days. Rankings are recomputed as new match data arrives.
- How is this tier list calculated?
- Heroes are ranked by the lower bound of a 95% confidence interval on their win rate, so a hero with a small sample cannot outrank a proven one. Tiers are then assigned by standard deviations from the roster mean rather than at fixed win-rate cutoffs.
- Why is a 52% win rate considered strong?
- Every Deadlock match produces one winning team and one losing team, so the average hero win rate is pinned near 50% by arithmetic — no amount of balancing moves the mean. Individual heroes still spread well clear of it, so 52% is around two thirds of a standard deviation above the roster average: a real edge, though not the top of the range. The section above gives the current spread and standard deviation.
- Does a high pick rate mean a hero is strong?
- No, and it often means the opposite for the win-rate number. A hero everyone plays carries the full range of skill and is dragged toward 50%, while a rarely-picked hero is played mostly by specialists and looks better than they are. Pick rate is shown on every row so the sample behind a win rate is visible.
- How often does this tier list update?
- It is recomputed from a rolling 30-day window of ranked matches and refreshes several times a day. Balance patches move heroes with a lag, because the window still contains matches played before the patch.
Win rate measures how often a hero's team wins, across all skill levels. It is not a measure of how strong a hero is in expert hands, and heroes that are hard to play tend to under-perform their ceiling here.