Ballpark

The idea

Ballpark is a daily estimation game. Every day there are five questions, the same five for everyone, and every answer is a number: the height of a building, the price of a painting, the number of bones in a hand. You are not expected to know the answers. You are expected to get in the ballpark.

The rules

1. Answer with a range. Drag the two handles, or type a low and a high number. You can type shorthand such as 750k, 3m or 2.5b.
2. If the real answer is inside your range, you score. If it is outside, even by a hair, you get 0 for that question.
3. Tighter ranges score more. A razor-thin range that hits is worth 100 points. The widest possible range is worth 5. A perfect day is 500.
4. One go per day. A new puzzle arrives at midnight, your local time. Play on consecutive days to build a streak.

How scoring works

The slider is logarithmic: each tick to the right is ten times bigger than the last. That is deliberate. When you are estimating, being off by a factor of two matters the same whether the answer is 40 or 40 million.

Your score depends on how many powers of ten your range covers compared with the whole slider. Most sliders span four powers of ten, so:

• a range from 900 to 1,100 (1.2× wide) scores about 98;
• a range from 500 to 2,000 (4× wide) scores about 85;
• a range from 100 to 10,000 (100× wide) scores 50.

The points you would win are shown under the slider before you lock in, so you always know what you are risking.

The share grid

Your result is a row of five tiles: 🟩 a tight hit (80 points or more), 🟨 a medium hit, 🟧 a wide hit, ⬜ a miss. It gives nothing away about the answers, so it is safe to post. You can also send a friend a challenge link: they play the same five questions and see a head-to-head at the end.

Calibration: the number that matters

Calibration is the share of your ranges that contained the answer. It is the skill Ballpark is really training. If you hit five out of five every day, your ranges are too wide and you are leaving points behind. If you hit one or two, you are overconfident: your ranges are tighter than your knowledge justifies. Good forecasters land around four in five.

Tips for estimating

• Break it down. You may not know how many piano tuners work in a city, but you can guess its population, how many homes have a piano and how often one gets tuned. This is called a Fermi estimate.
• Anchor on something you know. A marathon is 42 km. A jumbo jet weighs about 400 tonnes. The world has 8 billion people. Work outwards from there.
• Set the range, not the guess. Ask yourself: what is the smallest value that would not surprise me? And the largest? Those are your handles.
• Widen when you are clueless. A wide hit beats a confident miss every time, because a miss is worth nothing.

Where the answers come from

Every answer comes from a published source such as NASA, Guinness World Records, national archives or peer-reviewed research, and the source is linked on the answer card. Where reputable sources disagree on a number, the question is not used. You can check yesterday's answers with their sources at any time.