A five-judge panel is only fairer than a single judge if you’ve decided, before the event, exactly how five numbers become one. Most rulebooks stop at “the judges will score the item” and never say what happens next — whether the marks are simply averaged, whether an outlier gets dropped, or whether judges rank the entries instead of scoring them at all. Left undecided, the method gets picked after the fact by whoever is adding up the paper sheets, which is exactly when a coordinator ends up accused of choosing whichever method produced the answer they wanted. This guide covers the three real methods competitions use to combine a panel’s scores into one result, when each one needs more judges than organisers expect, how eTalenter’s own tabulation handles a multi-judge panel, and a printable panel score sheet you can run on paper if you need to.
Why “the judges will score it” is not a method
A rubric tells each judge how to arrive at their own number. It says nothing about how three, five or seven separate numbers for the same performance turn into the single score that decides a rank. That second step is a genuinely different decision — and it is the one most sample rulebooks skip entirely, because it only becomes visible the moment two judges disagree sharply on the same performance, which happens more often than organisers expect.
Three ways a panel’s scores become one result
| Method | How it works | Needs | Best for | Main risk |
|---|---|---|---|---|
| Simple average | Add every judge’s mark for the item and divide by the number of judges | Works with as few as 2 judges | Most stage items in a school, church or club festival — the default for a reason | One unusually high or low mark still pulls the average toward it |
| Trimmed average (drop the highest and lowest mark, average the rest) | Remove the single highest and single lowest score before averaging what’s left | Needs 5 or more judges to mean anything — with only 3, dropping 2 leaves one judge deciding the result alone | High-stakes finals or invitational panels, the same principle sports like diving and gymnastics use | Can discard a genuinely correct extreme mark along with a careless one — it controls for outliers, it doesn’t identify them |
| Rank-position (ordinal placing) | Each judge ranks the entries 1st, 2nd, 3rd… instead of awarding marks; the panel’s positions are combined across judges | Works at any panel size; needs judges comfortable ranking rather than scoring | Circuits without one shared numeric rubric — the placing model used by events like Highland dancing championships and Welsh eisteddfod-style competitions | Shows who finished ahead of whom, not by how much — two very different performances can produce the same rank gap as two nearly identical ones |
None of these is universally “more correct” — they solve different problems. A school kalolsavam with a five-mark rubric across a dozen judged items a day usually wants the simple average: fast, easy for judges to understand, and defensible without needing a large panel. A single flagship item at a district or state final, judged by a bigger invited panel, is where a trimmed average earns its extra complexity. And any competition built around placings rather than percentages — where “how it works in eTalenter” means choosing grades/ranks/points, not a raw percentage rubric — is already closer to the ordinal method than to averaging, whether or not the organiser thinks of it that way.
Decide the method before the rubric goes out, not after
Publish your combination method with the rest of the rules — the same document that carries time limits, deductions and the tie-break ladder (see our guide to breaking a tie for what happens once two combined results land equal). Settle four things in advance:
- Which method, for which items. It doesn’t have to be the same across the whole festival — a written test can use a simple average of markers while a single flagship stage item uses a bigger judging panel and a trimmed average.
- The minimum panel size. Two judges is enough for a simple average but leaves no way to spot a genuine outlier. Three is the practical minimum most festivals run to. Five or more is what a trimmed average actually needs to work as intended.
- What happens if a judge is absent on the day. Decide whether the item proceeds with fewer judges, whether a reserve judge is briefed in advance, or whether the item is rescheduled — and say so before it happens, not while a stage is waiting.
- Who can see the individual judge marks versus only the combined score. Some festivals publish only the average or the grade; others publish a per-judge breakdown. Either is defensible, but decide it once, for every item, rather than case by case after a complaint.
How eTalenter tabulates a multi-judge panel
Each judge scores an item from their own device against the criteria you set up under Scoring Setup, and eTalenter tabulates and averages the panel automatically instead of leaving a coordinator to total paper sheets by hand. The results report goes a step further than the plain average: it also flags when one judge’s individual ranking of a participant sits noticeably apart from the rest of the panel’s — the signal a coordinator actually needs to review a score before results go out, not after they’re already published. That review step is where a trimmed-average-style judgement call gets made in practice: look at what the outlier judge scored, decide whether it reflects something the rest of the panel missed or a genuine slip, and only then move to review and publish. Building a rubric that gives every judge the same anchors in the first place — covered in our guide to designing a fair scoring rubric — is what keeps this flag rare instead of routine.
Printable judge panel score sheet
Useful as a paper backup, or to walk a new coordinator through what the software is doing behind the average it shows on screen.
| Chest no. / participant | Judge 1 | Judge 2 | Judge 3 | Judge 4 | Judge 5 | Total | Average | Rank |
|---|---|---|---|---|---|---|---|---|
If you’re running a trimmed average with 5 judges, cross out the highest and lowest mark in each row in a different colour before totalling the rest — a visible record of which two marks were dropped is worth keeping alongside the sheet, for the same reason a tie-break record sheet is worth keeping: it is the record that answers a question asked a week later, not just the one asked on the day.
Briefing judges reduces disagreement before it happens
No combination method fixes a panel that was never calibrated against each other in the first place. A short briefing before judging starts — walking through the rubric together, agreeing what a mid-band mark actually looks like, and ideally scoring one sample performance together and comparing notes — closes most of the gap a trimmed average or an outlier flag exists to catch. Treat the tabulation method as the safety net, not the substitute for judges who understood the rubric the same way before they ever picked up a scorecard.
Frequently asked questions
What is the best way to average judges’ scores in a competition?
For most school, church and club-level festivals, a simple arithmetic average across all seated judges is the right default — it is easy for judges and participants to understand and works with as few as two or three judges. Move to a trimmed average (dropping the single highest and lowest mark) only once you can field five or more judges on the same item, since dropping two marks from a three-judge panel leaves a single judge deciding the result.
How many judges do you need to drop the highest and lowest score?
Five is the practical minimum. With five judges, dropping the top and bottom mark still leaves three scores to average, which is enough for the method to control for a genuine outlier rather than just reduce the panel to one deciding vote. With three or four judges, a plain average or a published outlier-review step (rather than an automatic drop) is the fairer choice.
Can judges rank competitors instead of giving marks?
Yes — this is the ordinal or placing method, and it’s the standard approach in competitions that decide results by 1st/2nd/3rd rather than by percentage, such as Highland dancing and eisteddfod-style circuits. Each judge ranks the entries in order of preference, and the panel’s positions are combined across judges rather than any raw marks being averaged. It suits panels judging on overall impression rather than a shared numeric rubric, but it doesn’t show how close the placings were — only their order.
Set up multi-judge scoring on eTalenter
Configure your rubric once under Scoring Setup, seat as many judges as your item needs, and let eTalenter tabulate, average and flag a contentious score automatically — so the method you published is the method that actually ran. Explore eTalenter’s kalolsavam & competition management software or our judge scoring & tabulation hub, message us on WhatsApp, or email support@etalenter.com and we’ll help you set up your judging panel.
Leave a Reply
You must be logged in to post a comment.