How skating is scored

Each skate gets a total segment score (TSS), and it is the sum of two halves minus anything taken off:

TSS = TES + PCS − deductions

TES, the technical element score, adds up every jump, spin, step sequence and lift: what it was worth on paper (its base value), plus or minus what the judges thought of how it went (its GOE). PCS, the program component score, is the judges’ view of the whole skate as a piece of skating and performance.

A competition has two segments: the short program and the free skate (in ice dance, the rhythm dance and the free dance). The two segment scores add up to the final result. On this site a program is the season-long work, one music and one choreography; a performance is one outing of it at one competition.

Every element has a fixed price before anyone judges it, set by the ISU in a table called the Scale of Values. A triple Axel is worth more than a triple toe loop because it is harder, whoever does it and however it goes. Spins, step sequences and lifts are priced by level, 1 to 4 (B for basic), which the technical panel awards for the difficult features it sees.

Jumps landed late in a program are worth 10% more, marked with an x on the protocol: since 2018–19 that is the last jump element of a short program and the last three of a free skate. A jump that is short of rotation is worth less: an under-rotated jump (<) is valued at about 80%, and a downgraded one (<<) is scored as the jump with one rotation fewer.

What the jumps are currently worth, as the protocols in this catalogue print them:

The most common value printed for each jump over the last three seasons. The ISU revises the scale from time to time, so older protocols can differ.

Each judge grades every element from −5 to +5 on how well it was done: height, speed, a clean landing, how it fits the music. Each step on that scale is worth roughly 10% of the element’s base value, so +3 on a jump worth 10 points adds about 3.

The highest and lowest grades are thrown out and the rest averaged, so one generous or harsh judge cannot swing the result alone. The protocol’s Panel column shows every judge’s grade. Before 2018–19 the scale ran from −3 to +3, and those older protocols are scored on it.

Where TES scores the parts, PCS scores the whole. Since 2022–23 there are three components, each marked out of 10:

• Composition: how the program is built and uses the ice and the music.
• Presentation: how it is performed, its expression and timing.
• Skating skills: edges, flow, speed and control in everything between the elements.

From 2005 to 2022 there were five: skating skills, transitions, performance, composition and interpretation. Each mark is multiplied by a factor that depends on the discipline and the segment, which is how PCS is weighed against TES.

Points come off the total for falls, for running over time, for costume or prop violations and for illegal elements. A fall costs at least a point, and more as falls accumulate. The fall also hits the element it happened on, through its GOE.

The technical panel annotates elements that were not quite what they claimed to be:

<<

Downgraded — at least a full rotation short

<

Under-rotated

q

Landed on the quarter

e

Wrong edge

!

Unclear edge

F

Fall

*

Not counted

V

Spin or lift of reduced value

REP

Repeated jump

>

Extended lift — held longer than allowed

x

Late in the program: base value increased by 10%

Protocols name every element in a compact notation. It is built from parts, so once the pieces are familiar any code can be read.

Every protocol on the site spells its codes out underneath, so none of this needs memorising.

Every skate gets two verdicts: where the judges placed it, and how the people watching rated it. Most of the time they agree. Robbed lists the skates where the crowd ranks a skate well above where the panel put it — the judges placed it 7th, say, and the crowd rates it among the best of the field. Overscored is the reverse: the panel was kinder than the room.

A skate needs at least three ratings before it can appear, so one person with strong opinions cannot decide the list.

For the curious: both verdicts are turned into standard scores so they can be compared — the judges’ against the rest of the same segment, the crowd’s against the rest of the discipline — and the gap between them is the number on the card. The bigger it is, the more the two disagree.

See the Robbed list →