GUIDE + RESEARCH

Kachi-koshi and Make-koshi: Sumo’s Winning and Losing Records Explained

Last verified: 3 Sep 2026Linked sources support only the corresponding claims in this article. Times, prices and venue operations can change, so check the current linked guidance before travelling.

Kachi-koshi means finishing a tournament with more wins than losses. Make-koshi means finishing with more losses than wins. A makuuchi or juryo wrestler normally competes once on each of the 15 days, so 8–7 is the smallest winning record and 7–8 is a losing record. Wrestlers in makushita and the divisions below normally have seven bouts, making 4–3 their smallest kachi-koshi and 3–4 a make-koshi.

English broadcasts often translate these terms as “winning record” and “losing record,” which is accurate but does not quite capture the suspense. Kachi-koshi is not simply a label added after Day 15. It can arrive in a single, visible moment: a wrestler on seven wins takes the eighth, bows, and suddenly the previous two weeks feel as if they have clicked into place.

You do not need to memorize the full ranking system to feel that moment. Choose one wrestler, check how many bouts he is scheduled to have, and know which next win crosses the line.

Eight is the usual threshold for makuuchi and juryo

Wrestlers in the two salaried divisions, makuuchi and juryo, normally have one bout on every tournament day. With 15 decisions and no unusual result, an 8–7 record contains more wins than losses; 7–8 contains more losses than wins. The Japan Sumo Association illustrates those exact records in its official explanation of a win-loss pattern called *nukenuke*.

This creates two especially dramatic situations late in a tournament. A wrestler at seven wins needs one more for kachi-koshi. A wrestler at seven losses must avoid an eighth loss to keep that possibility alive. On senshuraku, the final day, a 7–7 matchup can carry an entire basho in one bout even when neither wrestler is in the championship race.

The “eight” is not a mystical rank requirement. It is the smallest majority in a normal 15-bout schedule.

Four is the usual threshold below juryo

Makushita, sandanme, jonidan and jonokuchi wrestlers normally compete in seven bouts spread across the 15 days. Four wins therefore make a majority, while four losses make a losing record. An official Association article records that Hakuho’s first ranked tournament, in jonokuchi, ended 3–4: a make-koshi despite the extraordinary career that followed.

That example is useful for two reasons. First, it confirms how a seven-bout record is read. Second, it stops us treating one early make-koshi as a verdict on a career.

Lower-division schedules also explain why many cells in an official results row show no bout. A blank day or the symbol used for “rest/no bout” is not automatically a defeat. The wrestler may simply be between his scheduled contests.

Read the symbols before doing arithmetic

On the Association’s results table, a circle indicates a win and a filled circle a loss; square symbols distinguish a win or loss by default, and the Japanese character *ya* marks rest or no scheduled bout. Rare symbols also exist for a draw and an injury draw.

If a wrestler withdrew, returned, or received a default result, counting calendar days can produce the wrong answer. Use the official total and its day-by-day symbols. In the lower divisions, never count every non-bout day as a loss.

There can also be an exceptional eighth bout below juryo because of scheduling needs. That exception is exactly why “usually seven” is safer than claiming every lower-division wrestler always has precisely seven under every circumstance.

A winning record does not promise a fixed promotion

Kachi-koshi generally points upward on the next banzuke and make-koshi generally points downward, but the terms do not specify an exact number of ranks. Current position, the size of the record, openings elsewhere and other wrestlers’ results all matter to the ranking process. Ozeki have the separate kadoban and demotion framework; yokozuna cannot be explained by an ordinary one-rank movement chart.

So separate two statements. “He secured kachi-koshi” is a result that can be verified now. “He will rise exactly five places” is a prediction until the next official banzuke appears.

This distinction lets you enjoy fan forecasts without accidentally presenting them as Association decisions.

Kachi-koshi is not the same as a championship

A winning record compares one wrestler’s wins with his own losses. A championship compares the best records across an entire division. Many makuuchi wrestlers can finish 8–7 or better, but only one becomes the top-division champion. If the best records are tied after regulation bouts, a playoff determines the champion.

The reverse perspective matters too. A wrestler can be far from the title race yet have a deeply meaningful final day because his eighth win is still at stake. Sumo contains several simultaneous stories, not one leaderboard with attention reserved only for first place.

Find the bout that can cross the line

After the middle weekend, open the official results and choose a wrestler with six or seven wins—or, in makushita and below, three wins. Check the division and the number of completed bouts. Then watch for the contest that can secure the majority.

Listen after the decision. The arena may react differently to what looks like an ordinary win because many people already know the record. An expression may look still, but it does not establish the wrestler’s feelings; the record supplies the 14 days of arithmetic.

That is the detail I love: kachi-koshi turns a results table into anticipation. One white mark does not merely increase a total. It closes a small circle around an entire tournament, and once you have waited for that mark with one wrestler, the next day’s torikumi becomes much harder to ignore.

Sources

  • Japan Sumo Association, quiz Nos.1386, 1232, 17 and 1306, and the official results-table key (checked September 3, 2026)

Source links used for verification

Each link was used for the corresponding claim in this article; inclusion is not a blanket endorsement of the whole page.

Article verified: