A chart guessing game is any game where the answer is the shape of a line. Sometimes the format is multiple choice: which of these four graphs is real? Sometimes it is a score: how many of these trends can you call? Extraplot's version is a drawing game — you are given the real line up to a cut-off year and you draw the rest, year by year, in front of an audience of one.
Why draw instead of choose
Multiple choice is a test of recognition. You rule options out, and often the real chart is the one that looks least dramatic — which is a skill, but a narrow one. Drawing asks for the direction and the magnitude, and it asks for them without revealing the options.
It also removes the biggest loophole in guessing games: process of elimination. In Extraplot there is nothing to eliminate. Your prior about the world is the whole answer, and the reveal tells you exactly how good that prior was.
The house line is the standard
Every chart is scored against two lazy baselines: a flat line that holds the last visible value, and a trend line that carries the last ten visible years straight ahead. Those are the two guesses a player makes when they are not really thinking. The game shows both at reveal, scored with the same curve as your drawing.
Beating the house is a low bar on a warm-up and a real achievement on a bold call. The badge for beating both lines on all five charts of a single day is called House Beater, and it is the closest thing Extraplot has to a rank.
Reading a trend in five questions
- What kind of quantity is it? A share of a fixed whole behaves differently from a count. A price behaves differently from a rate. A stock differs from a flow.
- Where are the limits? Percentages stop at 100. Shares stop at zero. Populations age. Physical prices rarely return to their old nominal level once they have moved, while speculative ones can.
- Is the visible slope a trend or a one-off? A small line that has been straight for a decade has a different future from a small line that just jumped and held.
- What happened after the cut-off year? This is not cheating — it is reading history. A chart cut in 2019 about travel is a different question from one cut in 2014 about a device category.
- What would surprise you? Sketch the surprise. If the chart is a bold call, the honest line may be the one that looks least like the last ten years.
Guess with units, not vibes
The axis is labelled, and the label is a clue. A chart measured in millions of monthly users can fall as well as rise. A chart measured in dollars per person is about a different thing than a chart of total dollars. One of the fastest improvements new players make is reading the unit and the note before they start drawing.
The chart's title is written as a question for a reason. It tells you the subject; the note gives the qualifier — the country, the metric, the qualifier that stops the chart from being ambiguous. If you find yourself drawing a shape that does not fit the title, re-read both.
Seven categories, seven habits
- World — country series are mostly slow and bounded. Look for demographic transitions and policy breaks.
- Names — popularity curves have a rise, a peak and a long decline. The peak is usually the question.
- Search — pageviews spike on events and then decay. Expect a spike shape more often than a smooth curve.
- Culture — physical formats collapse faster than you think; digital ones compound longer than you think.
- Tech — adoption is an S-curve, and the top of the S is usually nearer than the last slope suggests.
- Nature — global series move slowly and steadily; the surprise is usually in the pace, not the direction.
- Money — prices mean-revert in real terms and rarely in nominal terms. A sudden move often partly unwinds.
How to practise the skill
The archive is a better training ground than the daily set, because you can pick a category and play ten charts of the same kind in a row. Patterns that take a season to learn from mixed daily sets can be learned in an hour of archive runs. Your daily streak is not affected, and your daily total is not touched.
Second, read the reveal note every time, even when you scored 950. The note is the expert commentary: it names the year, the number and the reason. Over a season it becomes a decent mental catalogue of how different kinds of data behave.
Third, draw with the keyboard at least once a week. Nudging values year by year is a different exercise from dragging, and it forces you to think about each year rather than the overall gesture.