This article describes and analyzes the structure used in the Full Tapped online league, organized through their Discord server. The goal is simple: understand the incentives the system creates and how it affects match quality as the event progresses.
Structure overview
1) Group stage: round robin
- 80 players are split into 8 groups of 10, assigned randomly.
- Inside each group, players play a round-robin: each player plays 9 matches.
- Weekly scheduling: each week has one pairing, and players agree on a day and time, then announce it on Discord.
- Matches are best of three.
- No draws (there is no time limit).
- Relevant rule: during the group stage, players are allowed to change decks between rounds.
2) Advancement to playoffs
- The top 4 players of each group advance.
- That produces a 32 player single elimination bracket.
3) Round of 32
- Each group winner (1st) plays a random 4th place finisher from a different group.
- Each 2nd place finisher plays a random 3rd place finisher from a different group.
- Position advantage from the group stage:
- In 1st vs 4th, the 1st place player chooses play or draw in game 1.
- In 2nd vs 3rd, the 2nd place player chooses play or draw in game 1.
4) From the round of 16 onward
- Pairings are fully random, ignoring both group and group stage placing.
- Play or draw is also random, with no advantage for the better group stage record.
- Matches remain best of three.
Issue 1: round-robin increases the frequency of mismatched incentives as the group stage progresses
The key difference between a Swiss stage and a round-robin group is that Swiss adapts to performance. As rounds go on, you tend to face opponents with similar records.
In a round-robin group, pairings do not adapt to records. As the group stage advances, record dispersion inside each group grows, and so does the chance of matches where:
- one player urgently needs points,
- while the other has little or nothing meaningful at stake (already eliminated, or almost eliminated, or already “basically qualified” and only playing for a minor advantage for the round of 32).
These are exactly the matchups that create incentive problems. Even if the match is played, it is easier for the player with little at stake to play with lower intensity, and it is easier for “soft” forms of intentionally losing to exist.
By contrast, Swiss tends to reduce the frequency of those mismatches, and as the event progresses it tends to concentrate the highest tension matches on the top tables.
Issue 2: allowing deck changes during the group stage amplifies the effect
Allowing deck changes in the group stage can be good for participation: players who are doing poorly can keep playing with more motivation, choosing decks they feel like playing rather than always choosing what they believe maximises win rate.
The competitive cost is clear:
- As rounds go on, a larger share of the group is likely to stop optimising purely for wins.
- That adds a “schedule variance” component: it matters when you play someone and what approach they have that week.
- It also makes “intentional losing without conceding” easier: you can simply pick a clearly worse deck, or a deck that is especially bad into that specific opponent.
None of this requires bad faith. It comes from the incentives the structure creates.
Issue 3: group stage performance matters very little for winning the tournament
In this structure, the only advantage of finishing 1st or 2nd is in the round of 32 (a slightly easier opponent in theory, and the play or draw choice in game 1). From the round of 16 onward, group stage results no longer matter at all: pairings are random and play or draw is random.
So a long group stage (9 rounds) primarily acts as a filter to build the top 32, but it has limited impact on actual championship odds after that.
How much does finishing 1st, 2nd, 3rd, or 4th matter? (A simple numeric illustration)
To make this intuitive, assume:
- From the round of 16 onward, each round is a 50% win chance (a neutral approximation).
- The only difference is the round of 32.
- Example round of 32 advantages:
- 1st place wins their round of 32 match 54% of the time.
- 2nd place wins 52.5%.
- Symmetrically, 3rd would be 47.5% and 4th would be 46%.
There are 4 rounds after the round of 32, so:
- 1st: 0.54 × 0.5⁴ = 3.375%
- 2nd: 0.525 × 0.5⁴ = 3.28125%
- 3rd: 0.475 × 0.5⁴ = 2.96875%
- 4th: 0.46 × 0.5⁴ = 2.875%
Useful takeaway: the most meaningful step is between 2nd and 3rd, because it flips whether you enter the round of 32 matchup “with” or “without” that small advantage. Even then, the difference in total championship probability is still small: 3.28125% vs 2.96875%, a gap of 0.3125 percentage points.
The goal of these numbers is not to predict real outcomes, but to show the scale of the effect: if results stop mattering from the round of 16 onward, most of the reward for doing better in groups is compressed into a single early elimination round.
Issue 4: random groups create uneven difficulty
Because groups are assigned randomly, “groups of death” can happen. With a top 4 cut:
- a strong group can eliminate a player who, with a different group draw, would have advanced comfortably,
- and in a weaker group, players can advance with an easier path.
In Swiss, this effect is reduced because you are not locked into a fixed pool: pairings happen across the full field and keep adjusting as records diverge, which lowers the impact of bad luck in an initial group assignment.
This becomes more important when, from the round of 16 onward, group stage performance stops having practical relevance.
Conclusion
The combination of:
- round-robin groups,
- deck switching during the group stage,
- and the total loss of practical relevance of group stage results from the round of 16 onward,
tends to produce two clear effects:
- As the group stage progresses, you get more pairings where one player has strong incentives and the other does not, which lowers average intensity and makes the group stage a weaker measure of relative strength.
- Even though there is an advantage in the round of 32 for finishing higher, that advantage disappears completely from the round of 16 onward, so group stage performance has limited impact on the final outcome.
By contrast, Swiss tends to reduce incentive mismatches and, as rounds progress, it tends to concentrate the most meaningful matches on the top tables.