Consensus Scoring in Taekwondo: When the Judges Have to Agree

Some events can’t be reduced to a single counter at a single tablet. Poomsae demonstrations, technical breaking, panel-judged sparring at federation level — these are events where multiple judges score independently and the system has to combine their views into one number. That’s consensus scoring. It looks like more work than point counting, and it is, but it’s the right tool when no single official should be the sole decider.

This is the second of three articles on scoring methods. Part 1 covered point scoring — the counted-actions system. Part 3 covers visual-pick scoring — when the eye decides. Consensus sits between them: a number, but a number agreed by a panel.

What consensus scoring is — and isn’t

Consensus scoring means: each of N judges scores independently, and the result is computed from the panel’s scores using a defined rule. The rule is the part that matters. It’s not enough to say “average them”. Real systems specify exactly how the panel’s scores combine, what gets dropped, and what counts as agreement.

The common patterns:

• Average of all judges — simple, vulnerable to one rogue score.
• Drop high and low, average the rest — the gymnastics-style rule, used in WT poomsae.
• Majority decision — each judge picks a winner; whoever has 3 of 5, or 2 of 3, wins. Used in panel-judged sparring.
• Threshold + tiebreak — a minimum agreement level required; if not met, a head judge or technical chair decides.

The poomsae case: drop-high-drop-low

The most common consensus method in taekwondo competition is the WT poomsae rule: 5 to 7 judges score on accuracy and presentation, the highest and lowest scores are dropped, and the remaining scores are averaged. Here’s how it actually plays out:

Two things to notice. First, the rogue scores got removed (J1 was harsh on accuracy, J5 was generous; J2 was harsh on presentation, J4 was generous). Second, the final number is calculated per criterion, not on the totals — so a judge who’s consistently strict on accuracy doesn’t get dropped on presentation, and vice versa.

This is why “just stick the totals in a spreadsheet” doesn’t work for poomsae. The rule operates on each criterion separately, and getting it wrong means the wrong competitor advances.

The panel-sparring case: majority decision

In panel-judged sparring (still common in ITF and at some federation championships), each of three or five judges picks a winner at the end of the bout — not a number, just a flag. The result is whoever has the majority.

Simple to describe, harder to run well. The risk in panel sparring is the asynchronous reveal — if judges call winners verbally one after another, later judges hear earlier ones and consciously or unconsciously align. The fix is making all judges submit their decisions simultaneously and privately, which is exactly what an electronic system gives you.

The five things consensus scoring breaks on (when run on paper)

1. Slow tabulation

Five score sheets per match per criterion, collected, walked to a calculator, dropped, averaged, written up. A single poomsae match can take three minutes of tabulation after the performance ends. Multiply by 200 entries.

2. Rounding inconsistency

Some tabulators round per-judge, some round only the final. The same panel scores can produce two different ranks depending on which calculator was used.

3. The judge who keeps changing their mind

Pen-and-paper judges scribble, scratch out, re-write. By the end of a long division it’s impossible to tell what the original score was — which matters when results are queried.

4. Tied final scores with no tiebreak rule loaded

Two competitors finish on 9.50. The rule says “next-highest dropped score”. Nobody on the day knows what that means, because nobody read the rulebook that closely. Now you’re writing policy at the medal table.

5. No reviewable record

A coach asks “what did Judge 3 give my athlete on accuracy?” on the way home. Nobody can say. The averaged total is on the sheet, the per-judge breakdown isn’t.

What proper electronic consensus scoring looks like

Each judge has their own tablet (or scoring panel) and submits their score privately. The system applies the rule — average, drop-high-drop-low, majority — instantly. Tied results trigger the configured tiebreak. Per-judge scores are stored against the match record.

What this means for organisers

The biggest hidden cost of consensus events isn’t the judging — it’s the tabulation bottleneck. A four-ring sparring division can run hot all day; a four-judge poomsae division spends as much time tabulating as performing. Electronic capture closes that gap to zero. The maths happens the instant the last judge hits submit.

If you’re running a mixed event — sparring brackets and poomsae rounds in one day — you almost certainly need software that does both, because nobody has the budget for two separate systems. We covered the buyer’s side of this in How to Choose Competition Management Software.

What this means for coaches and judges

For coaches: the per-judge breakdown is gold. If your athlete scored 4.2 / 4.0 / 3.6 / 3.9 / 4.2 on accuracy, you can see that one judge had a strong objection to something specific. That’s training feedback. The averaged total hides it.

For judges: a private, electronic submission removes the social pressure of public scoring. You score what you saw, you submit, you move on. The system does the rest. Judges who’ve worked both ways tend to prefer electronic submission once they’ve tried it — it removes the worst part of the job, which is being the one judge whose score made the difference and having coaches stare at you afterwards.

When consensus scoring is the wrong tool

Consensus is overkill for high-volume sparring. If the rule set defines points-per-action, use point scoring; the panel is the referee plus corner judges, not five independent assessors.

And if you’re running a creative or technical event where the panel doesn’t produce a number at all — just “competitor A looked better than competitor B” — you’re in visual-pick territory, which is the next post.

Key takeaways

• Consensus scoring combines independent panel scores into one result using a defined rule.
• The rule matters: average, drop-high-drop-low, majority, or threshold-plus-tiebreak — pick one and load it before the event starts.
• Poomsae uses drop-high-drop-low per criterion, not on totals.
• Panel sparring needs simultaneous, private decision submission to avoid judge-to-judge influence.
• Paper consensus is slow, lossy and unreviewable. Electronic capture removes all three problems.
• The biggest win for coaches is the per-judge breakdown for training feedback.