◦ MOST LIKELY TO · 2026 ◦
Most Likely To Questions
100+ Most Likely To questions specific enough to actually start an argument. For pregames, group chats, friend groups who've been through it. The wheel doesn't care, the table will. Free below.
15 questions that start a group debate
These are pulled from our app's most_likely mode. The common thread: each one references a specific, recognizable behavior that your friend group either does or doesn't do. The wheel will pick. The group will argue.
- Most likely to say 'quick drink' and come home at sunrise.
- Most likely to start a sentence with 'as a Capricorn' and lose all credibility.
- Most likely to delete and redownload Instagram six times in one weekend.
- Most likely to like a post from 2019 and pretend their phone glitched.
- Most likely to become famous for a video they begged everyone not to post.
- Most likely to fall in love because someone remembered their coffee order.
- Most likely to leave a group chat dramatically and rejoin in 20 minutes.
- Most likely to use 'I manifested this' to explain ordinary coincidences.
- Most likely to start a niche podcast no one will subscribe to.
- Most likely to keep using a friend's Netflix login until 2030.
- Most likely to ghost a date and then ask the group if they're being too much.
- Most likely to argue with the bartender about a wrong song on the speaker.
- Most likely to get a tattoo as 'a fun thing' and forget what it means.
- Most likely to have a 14-step skincare routine and bad sleep schedule.
- Most likely to lecture the group on a wellness trend they tried for 2 days.
How to play — group vote version
- Get 3+ people in a circle. Drinks optional.
- Read a "Most Likely To ___" question aloud.
- On 3, everyone points at the friend they vote for.
- Most votes wins (or loses — depending on the question).
- That person drinks, or reads the next one. Group choice.
- If there's a tie, both drink. Or both judge. Or argue about it.
What makes Most Likely To different from Never Have I Ever
Never Have I Ever is private — only you know if you've done the thing. Most Likely To is public — the group decides for you. That shift changes what the questions can ask. NHIE rewards confessions; MLT rewards diagnoses. The fun is in seeing how your friends actually see you.
The friend-group rule
Most Likely To questions work better when they reference shared in-jokes. After the first 10 spins, start riffing — "Most likely to ___ at Sarah's wedding next month" hits harder than a generic question. The app lets you write your own with the Custom Deck mode.
Common questions
What are good Most Likely To questions?
Good ones target specific, recognizable behavior patterns. "Most likely to be late" is boring. "Most likely to say 'quick drink' and come home at sunrise" forces the group to actually think about which friend that's most true for. Specificity is the entire game.
How do you play Most Likely To?
Pick someone to read the question. On 3, everyone points at the person they think is most likely. Whoever gets the most votes takes a sip (or laughs, or both). Optional: that person reads the next question.
How many players do you need?
Three minimum. Two players can play but the voting falls apart. Five to eight is the sweet spot — enough variety for surprising votes, small enough that each person gets called out often.
Can Most Likely To questions get mean?
They can, fast. The fix: keep the focus on quirky behavior, not flaws. "Most likely to forget your birthday" lands wrong. "Most likely to remember the exact lyric of a song no one else knows" lands right.
What's the best Most Likely To question for couples?
Couples Most Likely To is a 2-person variant: "If we both did X, which of us would do it first/worst/most dramatically?" Try: "Most likely to start a fight over which restaurant" or "Most likely to text the group chat first about a wedding we just left."
Get 200+ Most Likely To prompts (and 13 other modes)
The app has 210 Most Likely To prompts plus a Custom Deck mode where you can write your own inside-joke versions. Works offline, 10 languages, $14.99/year.
Install free on Android →