How to seed a tournament bracket

Why seeding matters

In a fully random draw, the top two competitors in a field might meet in the first round, eliminating one of them before the semi-finals. The event loses its best match-up to chance, and participants who have trained hard may feel the result does not reflect their true level. Seeding prevents this by reserving the best match-ups for the later stages.

For spectators and participants alike, a seeded bracket is more credible: the person who wins was not just lucky with the draw. Seeding is particularly important in larger events and at events where participants have travelled to compete.

Standard seed positions in a single-elimination bracket

Seeds are placed at fixed positions in the bracket according to a standard pattern. The core principle: seeds 1 and 2 go on opposite sides of the bracket so they can only meet in the final. Seeds 3 and 4 go in the two remaining halves so the top four can only meet in the semi-finals.

Seeds 3 and 4 are typically placed randomly into the two available positions (upper half bottom, lower half top) rather than in fixed positions, to add some unpredictability to which semi-final pairing occurs. Seeds 5–8 are also usually placed randomly within their respective quarters.

All non-seeded participants are then drawn randomly to fill the remaining bracket slots.

How many participants to seed

The number of seeded participants is typically proportional to the bracket size. Common practice:

• 8-slot bracket: seed the top 2 (sometimes top 4 for larger, well-ranked fields)
• 16-slot bracket: seed the top 4 (sometimes top 8 at high-level events)
• 32-slot bracket: seed the top 8 (top 16 at major championships)
• 64-slot bracket: seed the top 8–16

Over-seeding creates problems: if you seed 16 participants in a 16-person bracket, you have fully predetermined the draw, which defeats the purpose of a draw entirely. Leave room for randomness in the lower seeds.

Seeding criteria

The most credible seeding is based on objective ranking data. When that is not available, the following sources are used in practice:

Whatever criteria you use, publish them before the draw so participants can see how seeds were assigned. Disputes about seeding almost always arise from opaque processes, not from the seeds themselves.

Byes and seeding

When the participant count does not fill the bracket exactly, byes are added. Byes should be distributed to the top seeds — they have earned the advantage of a first-round rest through their ranking. Giving byes to lower-ranked or unseeded competitors is inconsistent with the logic of seeding.

With 12 participants in a 16-slot bracket, 4 byes are needed. Assign them to seeds 1, 2, 3, and 4. This means the top four seeds each advance directly to the second round and are guaranteed at least a quarter-final appearance — a fair reflection of their seeded status.

If there are more byes than seeds, distribute the remaining byes randomly among unseeded participants after all seeds have received theirs.

Seeding in round robin pools

In pool-based round robin events, seeding ensures the strongest competitors are spread evenly across different pools rather than concentrating in one. A pool with three of the top five competitors will eliminate two of them in the pool phase, while another pool produces a weak finalist.

The standard approach is to seed by snake order across pools. With four pools and eight seeded participants: seed 1 goes to pool A, seed 2 to pool B, seed 3 to pool C, seed 4 to pool D, seed 5 to pool D, seed 6 to pool C, seed 7 to pool B, seed 8 to pool A. This distributes top competitors evenly and places two seeds in each pool at opposite ends of the ranking.

For combat sports events, it is also common to separate competitors from the same club or country into different pools to avoid early intra-club or intra-national match-ups.

Handling late withdrawals

When a seeded participant withdraws after the draw is published, the usual approach is to promote the next-ranked unseeded participant into their slot rather than re-drawing the entire bracket. This minimises disruption while preserving the overall seeded structure.

If a seeded participant withdraws before the draw, simply re-run the draw with updated seeds. Have a seeding reserve list ready for events where withdrawals are expected — participants ranked just below the seed threshold who can step up cleanly.

Common mistakes

• Seeding without publishing criteria. Participants who disagree with their seed have no way to understand the decision. Always state the basis for seeding before the draw.
• Placing seeds 3 and 4 in the same half. This means seeds 1 and 3 (or 2 and 4) may meet in the semi-final. Seeds 3 and 4 must go into different halves of the bracket.
• Giving byes to unseeded participants while seeds play round one. Byes should go to the top seeds, not distributed randomly.
• Over-seeding. Seeding more than half the field removes most of the randomness from the draw without enough ranking data to justify it.
• Changing seeds after publication. Once the draw is published, only clear errors (data entry mistakes, verified withdrawals) justify changes. Adjusting seeds based on coach complaints after the fact undermines trust in the process.