How quiz scoring works on omequiz: base points, time bonus, streaks and estimations

A correct multiple-choice answer is worth 100 points plus 2 points for every second left on the clock. A question runs for 20 seconds, so an instant answer is worth 140 and an answer at the buzzer is worth 100. A wrong answer is worth nothing, and there is no penalty beyond that — you cannot go below zero, and guessing is never worse than staying silent.

Correct answers in a row raise a multiplier: 10 percent per answer, up to five in a row, so a perfect run is paid at 1.5 times. One wrong answer resets it. Estimation questions are scored differently: up to 1,000 points depending on how close you were, with no multiplier and no effect on your streak either way.

That is the whole system. The rest of this page is what those rules mean in practice, because the difference between a good score and a great one is not knowing more — it is understanding which of the three levers is worth pulling.

The base is 100 points for being right. The bonus is 2 points per second remaining, and the clock starts at 20 seconds for every multiple-choice question in the pool. So the bonus ranges from 0 to 40 — never more than 40 percent on top of the base.

That ratio is deliberate. If speed were worth as much as correctness, the winning strategy would be to slam the first plausible answer and hope, and a quiz between friends would turn into a reaction test. At 40 percent, speed decides between two people who both knew the answer, and never between somebody who knew it and somebody who did not. A player who takes fifteen seconds and gets it right (110 points) always beats a player who answers in one second and gets it wrong (0 points).

The practical rule: read the question, and if you know it, tap. Do not agonise over the last two seconds — they are worth four points. Do not rush a question you are unsure about to save time — the four points you gain are nothing against the 100 you lose if the rush was wrong.

A quiz that pays the same for the eleventh question as for the first gives nobody a reason to stay sharp for the eleventh. The streak is what changes that. Each consecutive correct answer adds 10 percent to a multiplier: the second correct answer in a row is paid at 1.1 times, the third at 1.2, the sixth and everything after it at 1.5. One wrong answer drops the multiplier back to 1.0.

The cap at five steps is the important design decision. A multiplier that kept climbing would decide a game by question four — the person who started well would be unbeatable, and everyone else would stop trying. At 1.5, a streak is worth about a question and a half of extra points over a full ten-question round. It is felt, it is worth protecting, and it is still catchable.

What this means for play: a streak is worth more than any single question. If you are on five in a row and hit a question you genuinely do not know, the expected value of guessing is still positive (you cannot lose points), but the expected cost of being wrong is now the multiplier on every question that follows. There is no way to "skip" — the clock runs out and counts as wrong — so the real advice is simpler: on a streak, use the full 20 seconds. The bonus you forfeit is small; the multiplier you keep is not.

An estimation question asks for a number — how tall the Eiffel Tower is, how many bones an adult has — and nobody is expected to hit it exactly. Instead the answer is scored by how far off it was, against a tolerance of one tenth of the question's range. An exact answer is worth 1,000 points. Every tolerance you are away costs 100. Ten tolerances away — the whole width of the range — is worth zero.

Concretely: if a question about a mountain has a range of 0 to 10,000 metres, the tolerance is 1,000 metres. Answer within 1,000 metres of the truth and you score 900 or more; within 2,000 metres, 800 or more; and so on. The scale is linear, so being half as wrong is worth exactly twice as much.

Two things follow from this. First, an estimation is worth up to ten times a multiple-choice question, so a round with estimations in it can swing on them — a leaderboard lead of 300 points is not safe if two estimations are still to come. Second, there is no multiplier: your streak neither breaks nor grows on an estimation. It sits out that question and picks up where it left off, which is the only rule that reads the same to everyone.

The practical rule for estimations: never answer a round number you have not thought about. The scoring rewards being close, so a moment spent anchoring — the Eiffel Tower is taller than a football pitch is long, and shorter than the Empire State Building — is worth far more than the seconds it costs. And do not leave it blank: an answer anywhere inside the range scores something.

A round is 5, 10, 15 or 20 questions, drawn from up to three categories. When the difficulty is set to "mixed" — the default — the draw is 40 percent easy, 35 percent medium and 25 percent hard, played in that order: the round warms up and then tightens. Every question in the pool has a 20-second clock; community quizzes written in the editor use the same clock.

The order matters for the streak. Because the easy questions come first, a streak is usually built early and tested late — which is where the 1.5 multiplier does its work. The last five questions of a mixed 20-question round are the hard ones, and by then a player on a long streak is being paid half again for each one. That is where rounds are decided, and it is why a slow, careful finish beats a fast, sloppy one.

Points are the round's score and nothing else. They do not carry between games, and they are not XP. Experience points and levels are a separate, slower system: a game awards XP only when at least two people played it, so a solo round is practice and never a way to climb levels. Winning a room you opened yourself is worth less XP than winning a matchmaking game against strangers, because one is a game you set up and the other is a game you were dealt.

And the score does not know who you are. There is no bonus for the host, no handicap, no adjustment for level. Two players who know the same answers and answer at the same speed score exactly the same — which, in a game you play with friends, is the only fair way for it to work.