Trials of Osiris
What Trials players are actually bringing, and what it is killing with.
135 Trials matches
(809 player entries) from the last
30 days, crawled from
108 players,
most recent 2026-07-28.
A further 131 older
matches are held but not
ranked, so this stays a read of the current meta rather than an average of
every patch since.
Weapon types
Rolled up by type rather than by individual weapon, because that is the
measure this sample supports: single weapons fragment across dozens of
similar rolls, while Hand Cannon versus Shotgun is stable — and it is what
people mean by the Trials meta anyway.
Individual weapons are listed separately.
| Tier |
Type |
Usage |
Kill share |
Kills / use |
Precision |
Weapons |
|
S
|
Hand Cannon |
50.7%
|
44.3% |
3.11 |
73.7% |
53 |
|
A
|
Shotgun |
32.4%
|
15.3% |
1.68 |
25.0% |
39 |
|
C
|
Pulse Rifle |
13.6%
|
10.7% |
2.82 |
69.0% |
34 |
|
C
|
Submachine Gun |
11.4%
|
9.0% |
2.84 |
64.0% |
20 |
|
D
|
Sniper Rifle |
6.7%
|
3.5% |
1.85 |
90.0% |
13 |
|
D
|
Auto Rifle |
6.3%
|
5.7% |
3.22 |
62.8% |
23 |
|
D
|
Scout Rifle |
6.1%
|
4.3% |
2.51 |
76.4% |
13 |
|
D
|
Rocket Launcher |
3.5%
|
1.3% |
1.36 |
0.0% |
8 |
|
D
|
Fusion Rifle |
2.8%
|
2.0% |
2.57 |
30.5% |
8 |
|
D
|
Sidearm |
2.5%
|
1.9% |
2.75 |
41.8% |
13 |
|
D
|
Machine Gun |
2.1%
|
0.8% |
1.29 |
54.5% |
8 |
|
D
|
Grenade Launcher |
0.9%
|
0.3% |
1.29 |
0.0% |
5 |
|
D
|
Sword |
0.6%
|
0.3% |
2.0 |
0.0% |
2 |
|
D
|
Combat Bow |
0.5%
|
0.5% |
3.5 |
35.7% |
4 |
|
D
|
Glaive |
0.2%
|
0.1% |
1.0 |
0.0% |
1 |
Usage is the share of player entries that brought the type at all, so it
does not sum to 100% — a loadout carries several weapons. Tiers grade usage,
not effectiveness: a match report says which weapons got kills, not whether
something else would have done better. Kills per use and precision rate are
shown alongside so the two are not confused.
Bungie publishes no leaderboard, so unlike the other sections here this is
not a ranking. These matches come from a crawl that starts
at arbitrary players and follows each match report to the players in it, so
the sample is “people who play Trials”, not “the best players”. Trials is
self-selecting for skill, which is what makes it worth measuring at all —
but treat it as a sample, not a ladder.