Pairing & season format

When two or more teams have the same number of points, we need a consistent way to order them — for standings, for division cutoffs (e.g. who gets 8th and Division A vs 9th and Division B), and for display. We use a fixed tiebreaker order. We compare teams on the first criterion; if tied, we move to the second; and so on until we have an order.

1. Points. First we look at raw points: wins × 3, plus bye × 3 if you had a bye. This is already the main ranking, but when points are equal we need more.

2. Resistance (opponent win %). We look at the average win rate of the opponents you actually played. Only opponents you played count — bye rounds do not add a "ghost" opponent. So if you had a bye in Round 3, that round simply doesn't appear in your resistance calculation; we don't give you a 0% or 50% fake opponent. That way having a bye doesn't unfairly hurt or help your tiebreaker. Resistance rewards playing stronger schedules: if you're 3–1 and your opponents have a high combined win rate, you're ranked above another 3–1 team whose opponents were weaker.

3. Soul differential. If points and resistance are still tied, we use souls for minus souls against (across your Phase 1 matches). Higher differential ranks higher.

4. Total souls. If we're still tied (rare), we use total souls scored. Higher total ranks higher.

We don't use head-to-head as a tiebreaker for Phase 1 standings (it can create circular ties and is harder to explain). The order above is simple and deterministic.

Division boundaries: Sometimes two teams are tied exactly at a cutoff (e.g. 8th vs 9th for the last Division A spot). Per-season policy may allow a playoff match in the break week between Phase 1 and Phase 2: one game (or short series) to decide who gets the higher division. Winner gets Division A; loser gets Division B. This is optional and announced at the start of the season if we're using it. Platform note: OCElock does not auto-create or schedule that tiebreaker match yet — staff run it out-of-band if the season rules call for it, then use Admin → Rounds division preview / adjustments before confirming transition.

Who gets the bye: When the number of teams is odd, exactly one team gets a bye that round. We don't pick at random. The bye goes to the lowest-ranked team that hasn't had a bye yet. Rank is determined by the same order we use for standings: points first, then resistance (opponent win %), then soul differential, then total souls. So the team at the bottom of the current standings, among those who haven't had a bye, gets the bye. If everyone has already had a bye (e.g. 5 rounds with 6 teams — everyone gets one bye), we use fewest byes so far and then lowest rank. That way we spread byes as fairly as possible and don't give the same team two byes while someone else has had none.

Points: A bye is worth 3 points, the same as a win. So in the standings it's like you "won" that round. You're not penalised in points for having sat out.

Resistance (tiebreakers): Your "strength of schedule" for tiebreakers is based only on opponents you actually played. Bye rounds do not add a ghost opponent. We don't give you a 0% or 50% fake opponent for that round — we simply don't count that round in your resistance. So having had a bye doesn't dilute your resistance (e.g. by adding a weak fake opponent) and doesn't skew it either. When we're deciding 8th vs 9th for Division A, you're not disadvantaged for having had a bye; your resistance is computed only from the teams you actually faced.

A forfeit is when one team doesn't show or can't field a full team by the agreed match time. The other captain (the one whose team showed) marks the match as a forfeit in the app. That's not a "bye" — it's a win for the team that showed and a loss for the team that forfeited.

Team that showed: They get a win and 3 points, exactly as if they had won a played match. For tiebreakers, the forfeiting team is not added to their resistance. We don't treat the forfeiting team as a 0% opponent (which would unfairly inflate the showing team's strength of schedule). So the showing team gets the W and the points, but that match doesn't count in their "opponent win %" — it's as if that round didn't exist for resistance purposes. That keeps things fair: you don't get a free strength bump from someone no-showing.

Forfeiting team: They get a loss and 0 points. So the result is a normal win/loss for both sides in terms of points and record; the only special rule is how we handle resistance for the team that showed. Full forfeit rules — including disputes, repeated forfeits, and how we handle multiple no-shows — are in Match rules.

If a team withdraws from the season during Phase 1 (drops out), we don't rewrite history. Here's exactly what happens.

Past results: All completed matches involving the dropped team stand. If they beat you in Week 2, that's still a loss for you. If you beat them in Week 1, that's still a win. We don't retroactively remove their results or pretend those matches didn't happen. Standings and records for everyone else stay as they are for the rounds already played.

Future rounds: For every round in which the dropped team was scheduled to play but no longer plays (because they've withdrawn), their opponent gets a bye. So that opponent gets 3 points and one bye for that round — same as a normal bye. We don't leave them without a match or give them a "free win" without points; we treat it as a bye so the bye rules (points, resistance exclusion) apply consistently.

Resistance: Teams who already played the dropped team keep that opponent in their resistance. The dropped team's final record (at the time they left, or however we close it out) counts like any other opponent when we compute "opponent win %." We don't remove them from your strength-of-schedule just because they dropped. So your resistance stays consistent with the matches that were actually played.

We designed the pairing and scoring rules to avoid several common problems. Here’s what we avoid and how.

Uneven pairing. If we just "paired the first two same-record teams" and then the next two, we might leave awkward leftovers (e.g. one 2–1 stuck with a 0–3). We use global cost minimisation (Blossom-style): the whole round is optimised in one go. So the set of pairings is as good as possible across the entire field — same record where possible, and when we have to pair across records, the algorithm chooses who gets the pairdown in a way that minimises total cost (and spreads pairdowns when we track mispairing count).

Rematches. We don't want you playing the same team twice in Phase 1. Pairings that already happened are either disallowed or given very high cost, so the algorithm only chooses a rematch when there is no other valid way to pair everyone (e.g. small field, many rounds). So you only play the same team again if it's unavoidable.

Repeated pairdowns. If we didn't track "how often has this team been paired with a different record," we might give one team two or three pairdowns in a row while others get none. We track how many times each team has already been paired with a different-record opponent and add a penalty so the algorithm prefers to give them a same-record opponent next. That spreads the pairdowns across teams instead of stacking them on one.

Unfair bye. The bye isn't assigned by a separate rule that might conflict with pairing. The bye is a node in the same graph and the same optimisation. The cost of giving the bye to a team is set so that the lowest-ranked team that hasn't had a bye yet gets it. So the bye assignment is fair and consistent with the rest of the round.

Bye hurting tiebreakers. Some leagues count a bye as a "ghost" opponent (e.g. 0% or 50%), which can skew strength-of-schedule. We don't. Resistance uses only real opponents; bye rounds do not count in your opponent win %. So you're not penalised at division cutoffs for having had a bye — your resistance is based only on the teams you actually played.