Give Every Tennis Data Table a Source, Coverage Date and Update Date

A published tennis data table should answer three questions without making the reader hunt through the article:

1. Where did the numbers come from?
2. How recent are the observations?
3. When did this table last change?

Those questions require separate labels. A single “Updated” date cannot communicate all three.

Use this minimum metadata block

Place the following information immediately above or below the table:

Label	What to display	What it tells the reader
Source	Named provider and a link to the specific dataset, statistics page or match record	Who supplied the underlying observations
Data through	Latest match, tournament or ranking date included	Where the sample ends
Table updated	Date this published table was last recalculated or substantively revised	Which edition the reader is viewing
Method	Short definition of the metric, denominator, filters and exclusions	How the displayed values were produced

The source should be a direct link where possible, not merely “ATP,” “WTA” or “official statistics.” If a table combines providers, list each one and identify which columns or rows it supports.

The W3C Data on the Web Best Practices recommends human-readable metadata and identifies the publisher, temporal coverage and date of last modification among useful descriptive fields. Its provenance guidance also calls for information about the data’s origins and any changes made to it.

Do not collapse three different dates into one

Source date is the publication or revision date supplied by the provider. If the provider gives no date, write “accessed 22 September 2026” rather than implying that the source itself was updated that day.

Data-through date describes the observations. For a season table, that might be “matches completed through 20 September 2026.” For career statistics, it might be “through the 2026 Laver Cup.” This is temporal coverage, not a publication date. Schema.org defines temporalCoverage as the period to which a work’s content applies, while dateModified is the date the work itself was most recently changed.

Table-updated date belongs to the publisher’s transformation of the data. Change it when matches are added, a filter changes, a calculation is corrected or revised source data alters the values. Do not reset it merely because unrelated prose on the page was edited.

A page can therefore correctly say:

Source: Named provider, linked to the relevant record Data through: 20 September 2026 Table updated: 22 September 2026 Method: Tour-level completed singles matches; retirements excluded; ace rate calculated as aces divided by service points

The dates in this display illustrate the format; they are not a claim about a particular dataset.

Define the transformation, not only the provider

A source citation does not explain what the publisher did after collecting the figures. For a serve table, the method note should identify at least:

• the unit of analysis: points, service games, matches or seasons;
• the denominator used for every rate;
• tours, draws and competition levels included;
• treatment of retirements, walkovers and incomplete match records;
• surface and date filters;
• rounding rules; and
• whether values were copied, aggregated or independently calculated.

This matters when similarly named columns are not interchangeable. First-serve percentage and unreturned-serve rate, for example, measure different events and can also vary with provider definitions. Our explanation of first-serve percentage versus unreturned rate shows why naming the denominator and provider is essential.

If one table mixes official match totals with a publisher’s own rate calculation, label both layers: “Match totals: [provider]. Rates: Serving Stats calculations from those totals.” That preserves the distinction between the original record and the analysis.

Handle revisions visibly

When a correction changes a value, retain the new table-updated date and add a short dated note explaining what changed. Silent replacement makes it difficult for a reader to reconcile an earlier citation or screenshot.

The W3C recommends a unique version number or date for each dataset and a version history explaining the changes. It says explicit versioning supports repeatability, comparisons and clarity (data-versioning guidance). A full archive may be excessive for a small editorial table, but a compact correction log is not:

• 22 September 2026: Corrected the hard-court filter; three qualifying matches were added.
• 18 September 2026: Initial publication.

Also label provisional data explicitly. “Current through” does not mean “final,” particularly when a tour, tournament or statistics provider may later amend a match record.

Keep the metadata with the table

Put the visible note in the table caption or in an adjacent line, where readers will encounter it with the figures. Do not hide the only source link in a general references section. For long tables, repeat a compact source-and-date line in downloadable files.

Machine-readable metadata can supplement that display, not replace it. Schema.org’s Dataset type includes properties such as datePublished and dateModified, while temporalCoverage describes the period represented by the content (Schema.org dataset properties). The reader-facing labels should still remain next to the figures.

The practical standard is simple: name and link the source, state the observation cutoff, date the table’s own revision, and define the calculation. A reader can then tell whether a tennis statistic is current, comparable and traceable before interpreting it.