Endgame and All-Last Strategy
23.1 All-Last (オーラス)
The final hand of the match is called all-last (オーラス, oorasu). This is the most strategically intense hand because it is your last opportunity to influence your placement. Every decision in all-last should be made with explicit reference to point gaps and placement possibilities. The normal rules of thumb for push/fold, riichi/dama, and hand value are all modified by placement considerations.
23.2 Point Gap Calculations
Before the all-last hand begins, calculate the point differences between yourself and the players directly above and below you. Then determine: What kind of hand do I need to move up? What kind of deal-in would drop me down? These calculations should guide every decision.
Example scenario: You are in 3rd place with 22,000 points. 2nd place has 28,000 (gap: 6,000). 4th place has 15,000 (gap: 7,000). You need to win a hand worth approximately 6,100+ points (to pass 2nd by ron from 2nd place, accounting for the +100 tiebreaker going to the earlier seat) to move to 2nd. A mangan tsumo (2,000/4,000) would give you +8,000, moving you to 30,000 and past 2nd place (who would drop to 26,000). Planning: pursue a mangan-class hand or higher if possible.
23.3 All-Last Strategic Modes
Comfortable first place: Fold almost everything. You win by not losing. Even reaching noten at a draw is fine if your lead survives the noten penalty.
Close race for first: Push for a hand that secures first. May riichi for extra value or stay dama for safety, depending on the exact gap.
Third trying for second: Target specific point thresholds. Consider direct-hit ron against the player you need to pass. Open hands for speed if the value threshold is low.
Fourth place desperate: Maximum aggression. Consider normally suboptimal plays: opening for cheap hands if a small win moves you to 3rd, or pushing toward expensive hands regardless of risk.
A common all-last mistake is "auto-pilot"—playing the final hand with the same strategy as any other hand. The all-last requires conscious calculation of target scores and placement implications for every decision. Failing to adjust your strategy for all-last is one of the biggest sources of avoidable placement loss.
QUIZ — Question 23.1
All-last. You are 1st with 35,000. 2nd has 33,000. You are dealt a mediocre hand (4-shanten). What is your priority?
Answer: A. With a razor-thin 2,000-point lead and a 4-shanten hand, the probability of winning is low and the risk of dealing in (dropping to 2nd or worse) is significant. Play pure defense. Even reaching noten at a draw only costs 1,000-3,000 depending on how many others are tenpai—which might drop you to 2nd, but dealing into a mangan definitely would. Minimize risk.
23.4 All-Last Calculation Examples from Real Games
Consider this all-last scenario from a typical online game: You: 28,000 (2nd). Player A: 33,000 (1st). Player B: 22,000 (3rd). Player C: 17,000 (4th). You are non-dealer.
Target calculation: To overtake 1st by direct ron from Player A, you need a hand worth more than (33,000 - 28,000) / 2 = 2,500. Any ron from Player A worth 2,600+ puts you ahead. A 2-han 30-fu hand (2,000 ron) is NOT enough. A 2-han 40-fu hand (2,600 ron) IS enough. So your target is the minimum 2-han 40-fu hand — or 3-han 30-fu (4,000) for safety.
Tsumo calculation: A tsumo win does not take directly from 1st. Non-dealer tsumo at 3-han 30-fu pays 1,000/2,000. You gain 4,000, Player A loses 2,000. New scores: You 32,000, Player A 31,000. You overtake! But if you tsumo at only 2-han 30-fu (500/1,000), you gain 2,000 and Player A loses 1,000. New: You 30,000, Player A 32,000. NOT enough.
These precise calculations are what strong players perform at the start of every all-last hand. The ability to quickly compute target scores is trained through practice and is emphasized in Katsumata Takeru's (勝又健志) endgame strategy teachings and M-League commentary.
Source notes: Content validated against standard Japanese riichi mahjong references and strategy literature. Strategic concepts reflect consensus from Japanese professional commentary and analytical sources.