The crowd memory report
What players
remember.
Every finished pile leaves a map of gaming memory: the reflex answers, the eras that dominate, and the places where recall slows down. These figures come from real, completed Name 100 Games runs—not surveys or generated articles.
3,200
qualifying completed runs
320,000
verified answers in the sample
16,577
distinct games remembered
Sample window Jul 13, 2026–Aug 31, 2026 · updated from the latest qualifying run
Enter the collective memory mapSee which games people recall together01 · The memory shelf
The games that refuse to be forgotten
Ranked by the share of qualifying runs containing each title. A game can appear only once per run, so this measures piles containing the game—not raw repetitions.
- #1
Minecraft
64.7% of runs
- #2
Call of Duty
53.4% of runs
- #3
Fortnite
45.5% of runs
- #4
League of Legends
44.9% of runs
- #5
Dark Souls
39.7% of runs
- #6
ELDEN RING
39.6% of runs
- #7
Mario Kart: Double Dash!!
38.8% of runs
- #8
Overwatch®
36.9% of runs
- #9
The Legend of Zelda
36.5% of runs
- #10
Valorant
36.2% of runs
- #11
Super Mario Bros
35.4% of runs
- #12
Halo: Combat Evolved
35.2% of runs
02 · The starting grid
Which game lands first?
Openers reveal the titles closest to the surface before strategy, series runs, and memory detours take over.
- 110.9%
Minecraft
Opens 350 runs
- 24.2%
Super Mario Bros
Opens 135 runs
- 33.8%
Fortnite
Opens 121 runs
- 43.0%
Halo: Combat Evolved
Opens 95 runs
- 5
2.3%Pokémon Red Version
Opens 74 runs
- 6
1.9%Super Mario 64
Opens 60 runs
- 71.8%
Call of Duty
Opens 57 runs
- 81.5%
League of Legends
Opens 49 runs
03 · The hundred-step climb
Recall gets heavier near the top
Each step shows the median time spent on that ten-game segment. The clock beneath it is the median elapsed time when that milestone lands.
- 10+0:470:47 elapsed
- 20+0:491:38 elapsed
- 30+0:512:32 elapsed
- 40+0:523:29 elapsed
- 50+0:554:29 elapsed
- 60+0:575:30 elapsed
- 70+0:586:29 elapsed
- 80+0:587:31 elapsed
- 90+1:008:35 elapsed
- 100+1:029:40 elapsed
04 · The memory fog
Where the longest pauses gather
Within each ten-game stretch, we found every run’s longest single pause, then took the median across runs. These are the five stretches where a typical run hits its thickest fog.
- Fog bank 1
0:16
longest typical pause · games 91–100
- Fog bank 2
0:15
longest typical pause · games 81–90
- Fog bank 3
0:15
longest typical pause · games 71–80
- Fog bank 4
0:15
longest typical pause · games 61–70
- Fog bank 5
0:14
longest typical pause · games 51–60
05 · The arcade timeline
The eras inside the pile
Share of answers with a known release year, grouped by decade. Release metadata covers 66.6% of all answers in this sample.
06 · The recurring deep-cut crate
Rare, but not a one-off
These titles appeared in at least two completed runs while staying at or below 5% of the sample. That recurrence keeps one-off anomalies out of the crate.
Little Nightmares
161 runs · 5.0%
NBA 2K19
161 runs · 5.0%
PAYDAY 2
161 runs · 5.0%
Dragon Nest
160 runs · 5.0%
Dying Light
160 runs · 5.0%
F-Zero
160 runs · 5.0%
Fire Emblem: Three Houses
160 runs · 5.0%
THE GAME OF LIFE
160 runs · 5.0%
Enter the Gungeon
159 runs · 5.0%
Jak and Daxter: The Lost Frontier
159 runs · 5.0%
Split Fiction
159 runs · 5.0%
Candy Crush Saga
158 runs · 4.9%
How original is a pile?
The median completed run has an average title overlap of 10.9%. The most original 10% average 7.7% or less.
Methodology
We scan at most the latest 5,000 candidate standard SOLO/100 runs under the current rules version; 3,200 qualified for this report. A qualifying run must contain exactly 100 distinct normalized game titles with ordered timestamps and a duration between ten seconds and 24 hours. Internal test users, Daily, Live, Twitch, Ultra, Perma Death, incomplete, malformed, and exact duplicate timelines are excluded.
A “run” is not necessarily a unique person: 2,866sample runs were guests and 334 belonged to accounts, while repeat attempts can remain in either group. Frequencies therefore describe completed runs, not the percentage of all players who remember a game.
Pace figures use median ten-game segment times. Stall figures first take each run’s longest single pause within a ten-game stretch, then use the median of those maxima, so one abandoned browser tab cannot dominate the result. Release-era percentages use only answers with a known year; failed optional metadata lookups never remove a game from any other count. No rejected speech text or audio is used here.
Game rankings are cover-driven: current catalog artwork is used to fill older run metadata, and titles still lacking a cover are omitted from visual ranking slots. Those titles remain in the aggregate run, answer, uniqueness, timing, and era denominators.
Sample begins
Jul 13, 2026
Last updated
Aug 31, 2026
Dated answers
213,085 · 66.6%
Your memory, on the clock