What the picture proves, and what it never proves

A picture feels like the most reliable thing on the screen, and in a dispute it is the least authoritative. Three sources describe the same event, and only one of them decides a bet. The picture is not it.

Desk spec

sources: 3
that govern: 1
disputes: 100
settled by picture: 0

the pictureA live picture is a licensed broadcast shown beside a market, not a window onto the event. It has an owner, a territory and a cost, and only 96 of 300 events in the month carry one.

the two clocksThe event happens at 0.0 s, the price refreshes at 0.2 s and the picture arrives at 12.0 s, so the number beside the picture is 11.8 seconds newer than the picture it sits beside.

the one sourceThree things describe an event - the picture, the data feed and the official result - and only the official result settles a bet. The picture looked decisive in 96 of 100 disputes and settled none of them.

Direct answer

Three sources describe a streamed event: the live picture, the data feed and the official result. Only the official result governs a bet. On the samples, of a hundred disputed results the picture looked decisive in ninety-six and decided none of them, because the rules name the official result as the source and a camera is not a rulebook.

The three sources

The picture and the feed are both fast and both partial, and the official result is slower and complete. The rules of a market say which source it is graded on, and for almost every market that source is the official result rather than the broadcast.

Sample H - the three sources and what each one is for
Source	Arrives	What it is	Does it settle a bet
the live picture	12.0 s	a licensed broadcast of the event	no
the data feed	2.5 s	a running set of facts about the event	no
the official result	after the event	the statement the rules name	yes
three sources	2 fast, 1 slow	one rulebook	1 of 3 governs

sample H - the picture against the result disputed results examined = 100 cases where the picture looked decisive = 96 -> 96.0% cases where the picture governed the settlement = 0 -> 0.0% cases where the official result governed = 100 -> 100.0% so a picture that looks decisive is wrong about its own authority 96 times in 100, and right about it 0 times in 100.

Why a decisive frame settles nothing

The frame that looks clearest is often the one furthest from what the rules measure. A ball on a line, a foot in front of a defender, a hand a fraction before a whistle: the rules name a measurement, an official and a protocol, and a picture at one angle is none of those. The picture is a reason to ask the question; the result is the answer.

sample H - where the disagreement lives 100 disputes, by what the picture appeared to show: clearly one way 96 -> 96.0% and decided 0 of 96 unclear 4 -> 4.0% and referred to the result, 4 of 4 official result and picture agreed 96 of 100 -> 96.0% official result and picture differed 4 of 100 -> 4.0% and in all 4 the official result was the one paid, so the picture was right about the event 4 times in 100 and relevant to the settlement 0 times in 100.

Check which source a market is graded on

Does the market rule name the official result, the broadcast, or a specific body?
Is the thing you are disputing a measurement the rules define, which a picture cannot settle?
Has a result already been amended after the event, which a picture could never have shown?
Does the operator publish the source it grades on, or only the result it reached?
If the picture and the result disagree, which one do the rules say to use?

What the picture proves, and what it never proves

The three sources

Why a decisive frame settles nothing

Read next