A scoreboard is a design decision, and it throws something away
Every scoreboard makes the same choice. It shows who is ahead and by how many sets, and it compresses everything that produced that state into one or two characters.
That compression is normally invisible. A record of 116,382 professional tennis matches makes it visible, because it shows what the discarded part was worth.
One set down, six ways
A player who loses the opening set goes on to win 16.52% of the time. That is the figure commentary repeats, and on the scoreboard every one of those players looks identical. One set to love down.
Split them by how the set was lost and the single number comes apart:
– lost it 7-6: wins 21.92% of the time, from 14,265 matches
– lost it 6-4: 19.41%, from 25,546
– lost it 7-5: 18.89%, from 10,252
– lost it 6-3: 17.39%, from 25,970
– lost it 6-2: 13.30%, from 19,739
– lost it 6-1: 11.06%, from 14,269
– lost it 6-0: 7.18%, from 5,833
Top to bottom that is a threefold spread. The scoreboard carries the margin in its second line, and the summary that people quote drops it.
The comparison that makes the point
Look at the 7-6 group and the 6-1 group. They hold 14,265 and 14,269 matches, four apart in samples of over fourteen thousand each.
Sample size cannot explain the gap between them, which is the usual way a pattern like this gets waved off. One group recovers twice as often as the other. Same sport, same period, same measurement.
The ordering never breaks either. Every time the margin widens by one game, the recovery rate falls, through all seven scorelines without exception.
What this has to do with how things are displayed
The useful version of the headline is not a better average. It is one more field: by how much.
The same choice shows up everywhere a system reduces a state to a glyph. A dashboard tile reading 3% conversion hides the channel converting at 9% and the one at 0.4%. A progress bar hides whether the remaining work is the easy part. A rating out of five hides the shape of the distribution underneath it.
fontsarena.com, part 2 of 2. Follows straight on from part 1.
In each case the fix is the same, and it belongs to design before it belongs to statistics. Find the dimension the display averaged over, and give it somewhere to live on the surface.
Where the detail has to survive
None of these splits would exist if the underlying record had been stored the way the scoreboard presents it. They exist because the data was kept point by point, with every event retained. Nobody decided in advance that first-set margin would be the thing that mattered.
It mattered, and the detail was still there to check. That is the argument for aggregating late. The question that makes a number useful is usually one nobody has asked yet, and a pre-summarised record can only answer the questions somebody anticipated.
The figures here come from the Live Tennis API, a JSON service for developers. It covers live scores and fixtures, with players, match events and point-by-point history behind them. They are published openly under CC BY 4.0, free to reuse with attribution.
Method: completed best-of-three professional singles matches, 1 January 2023 to 14 September 2026. Retirements and walkovers excluded. Counted from the point-by-point record.