RRS A8.1 compares the full best-to-worst list, including back-of-fleet scores. Pat and Quinn both net 12 in a 7-boat fleet with one discard. Their kept (non-discarded) scores are Pat [6, 6] and Quinn [5, 7]; at the first point of difference Quinn's 5 beats Pat's 6, so Quinn ranks ahead — matching HalSail.
This reproduces the DBSC Thursday Blue regression. The pre-2025 place-count tie-break only counted places up to (races + 1) = 4, so every distinguishing score here (5, 6, 7) fell outside its window; it found no difference and fell through to a last-race countback that ranked Pat ahead. Wrong.
Scoring configuration: 1 discard from race 3. DNF/OCS scoring: A5.2 (series entries).
7 boats, N+1 = 8 (DNC). One discard.
Pat: 6th, 6th, DNC → {6,6,8} discard 8 → net 12, kept best-to-worst [6, 6]
Quinn: 5th, 7th, DNC → {5,7,8} discard 8 → net 12, kept best-to-worst [5, 7]
Pat and Quinn tie at net 12.
A8.1 — kept [6,6] vs [5,7]: Quinn's best 5 beats Pat's best 6 → Quinn wins.
The old place-count loop only looked at places 1..(races+1)=1..4, so the
5/6/7 scores were invisible and it wrongly fell through to a countback.
Fillers finish ahead and never tie the contenders:
F1 1/1/1 → net 2; F2 2/2/2 → net 4; F3 3/3/3 → net 6;
F4 4/4/4 → net 8; F5 7/5/5 → {7,5,5} discard 7 → net 10
Pat and Quinn both DNC (8 pts); fillers finish 1-5