Go, Xiangqi and shogi results: compare records without mixing rankings

Begin with the archive’s own legend

The Nihon Ki-in results page separates recent results from forthcoming games and explains its display conventions. Its legend identifies the first player and places the winner on the left. A reader who assumes that the left-hand name always means the first player can therefore reverse the meaning of a record. Before importing any table, copy its legend and identify which symbols describe colour, outcome or player status. Keep scheduled games out of completed-result statistics even when the same page displays both kinds of record.

Identify what each official source publishes

The WXF champions archive separates individual and team categories, while the JSA ranking page distinguishes several performance summaries. An archive list ending in 2022 should not be presented as a complete list through 2026 merely because the website itself is current.

Do not convert rank labels into a common numerical scale

A dan designation, a title, a tournament seed and a numerical rating can all appear beside a player’s name. They have different administrative and sporting meanings. A ninth-dan Go player and a similarly labelled player in another game do not occupy positions on one shared measurement scale. Even within one game, professional status and an event’s current seed can answer different questions. For a comparison table, use separate columns with their original labels. If a field does not apply, leave it explicitly inapplicable rather than assigning a zero that looks like a measured weakness.

Show the denominator behind a win rate

Consider two fictional records: Player A wins eight of ten completed games, while Player B wins eighteen of thirty. Their simple win rates are 80 percent and 60 percent. This arithmetic does not establish that A is the stronger player, because the opponents, time controls and event levels may differ. It also does not describe any actual player in the cited archives. A good statistical summary shows wins, losses, draws or other recorded outcomes and the total eligible games. State how forfeits, replays and unplayed rounds are treated before calculating the rate.

Respect delayed publication and incomplete coverage

The JSA ranking page includes a dated coverage note and a qualification concerning television games that have not yet been broadcast. This is a reminder that an official public table can have an intentional publication boundary. Do not infer an undisclosed result from a gap, and do not assume that a change in the table means a new game was played that day. When comparing snapshots, record their cutoff dates and inclusion rules. If a source later publishes a previously withheld result, explain the change as a coverage update where appropriate rather than as an unexplained statistical jump.

Separate game points, team points and event placing

A team competition may aggregate several boards into a match result, while an individual tournament may rank players through its own score and tie-break system. Record board-level outcomes and team-match outcomes in separate tables connected by an event identifier. Otherwise, a team winning three boards can be mistakenly counted as winning three team matches. In a knockout event, final placing also depends on the bracket and any classification matches. The official regulations decide those relationships. A general data template should make room for the rules rather than forcing every competition into a football-style league table.

A cross-game archive checklist

• Copy each source’s legend, cutoff date and coverage note.
• Keep game, round, match and event identifiers separate.
• Preserve original rank and title labels without numerical conversion.
• Show the denominator and treatment of unplayed results.
• Compare equivalent categories and disclose missing coverage.

For a small research project, begin with one event in one game and manually validate several records. Only then extend the structure to another game. This exposes assumptions that would otherwise become hidden spreadsheet formulas. Publish a short methods note with the finished table, including its exclusions and the date of the last check. Readers can then understand what the comparison demonstrates and, just as usefully, what it cannot demonstrate about the players’ wider careers.

Before merging archives, write a field dictionary that defines “game”, “match”, “round” and “event” for each source. The same English word may be used differently by different publishers. A stable player identifier is also preferable to a name alone, especially where romanization or character variants occur. If an identity cannot be confirmed, retain an unresolved record rather than merging two people prematurely. These modest controls prevent large apparent trends from being created by duplicate names, inconsistent units or an unnoticed change in the archive’s coverage.

Sources and verification

• Go results and forthcoming games — Nihon Ki-in — 2026-10-23
• World Xiangqi Championship champions archive — World Xiangqi Federation — 2026-10-23
• Annual shogi games, wins, win-rate and streak rankings — Japan Shogi Association — 2026-10-23