What goes in a brief, what gets left out, and how each number is calculated. If something here is wrong, the briefs are wrong too.
From 25 August 2026, matches are chosen by hand. The editor names the matches; this page researches exactly those. The odds filter described below governed editions 2 through 9 and no longer selects anything — it was retired because it was picking matches on the wrong criterion, forcing full research into any first-rounder that happened to be priced near even while leaving out matches worth reading about. The recent-data requirement still applies, and a match priced as a formality is still flagged as one on its card.
What follows is kept as a record of how the archive up to 24 August was assembled.
Each of those editions covered at most ten main-draw singles matches. Two filters ran before any research, so effort only went into matches that would actually appear.
A match is excluded if either player is priced longer than 4.00 decimal, which is +300 in American odds. In practice that removes anything where the favourite is above roughly 75%. The reasoning is that form, workload and injury notes change very little about a match that is already decided — and a brief padded with foregone conclusions is harder to read.
This does exclude matches you might want to read about. A marquee name at 6.11 is gone, and every excluded match is listed at the foot of the brief with its price, so nothing disappears silently.
Both players must have at least three completed singles matches in the last 90 days with sourceable set scores, and must have played within the last 30 days. A card whose form section is guesswork is worse than no card.
Survivors are ranked by how close the market puts them to even, and the ten closest are shown. Item 01 is the tightest match of the day. If fewer than ten survive, the brief is shorter.
Implied probability is 1 ÷ decimal odds for each player, normalised so the pair sums to 100%. That normalisation strips the bookmaker's margin, so the figures show what the market thinks rather than what it pays. Where books disagree materially on who the favourite is, the brief says so on the card.
From 21 August every card carries an Elo rating for each player, taken from Tennis Abstract. Elo is the rating system built for chess: every match moves both players' numbers, by a lot when the loser was highly rated and by very little when they were not. A rating is therefore a running summary of who a player has beaten, rather than how far they got in a draw — which is what separates it from a ranking, where reaching the third round of a big event can be worth more than winning a small one.
Two ratings convert directly into a win probability: 1 ÷ (1 + 10^((Rb − Ra) ÷ 400)). A 100-point gap implies about 64% for the higher-rated player, 200 points about 76%, 400 points about 91%. Because that figure comes from results alone and owes nothing to a betting market, it can be set against the market line and the two compared — which is the whole reason for carrying it. The caret above each probability bar marks where Elo puts the match; the line beneath gives both figures and the size of the disagreement.
The surface-specific rating is used for the comparison, since Tennis Abstract publishes separate hard, clay and grass numbers, each blended with the player's overall rating. Cards show the overall figure and the surface figure, with the surface being played highlighted.
Elo is not a forecast this site endorses. It is a second opinion, computed a different way from a different input, printed so that the market's number is not the only one on the page.
Two windows, trailing from the day of play: last 7 days and last 30 days. Each shows matches played, with total sets in brackets.
Earlier editions reported court time in minutes instead. That was abandoned because match durations are not published for every event, which meant one player's figure could be minutes and their opponent's sets — an unhelpful comparison. Sets are derivable from any scoreline.
All-time singles win-loss, split by surface, taken from a single source for every player on a page so the figures are comparable with each other. Two caveats that matter:
The row for the surface being played is highlighted, because the contrast is often the story — a player with a clay record on a hard court is a different proposition from what the ranking suggests.
Only sourced, dated items appear. Match logs are scanned for retirements, walkovers and gaps of four weeks or more, and each match gets a targeted search. Where a gap has no reported cause, the brief says so rather than inventing one. "No injury concerns found" means nothing surfaced — not that a clean bill of health was issued.
One line per tournament: sky, temperature, precipitation chance and wind for the playing day, from the national weather service for that country. Forecast points are often a nearby city rather than the venue itself, which the brief notes where it matters.
Slate, odds and head-to-head records come from TennisExplorer, with prices checked against a second source where one exists. Match logs come from TennisExplorer player pages cross-checked against draw progression, plus the ATP and WTA official results centres and tournament recaps. Elo ratings come from Tennis Abstract's published reports, dated on each card. Order of play is taken from the tournament's official schedule; at combined events, one feed carries both draws.
A standing warning, since it shapes the briefs: TennisExplorer is unreliable in specific, repeatable ways. It serves stale pages, splices prior-year rows into an in-progress tournament, and prints scores winner-first in a way that makes losses look like wins. Every win-loss in a brief is verified against draw progression rather than read off the score column. It has, in a single week, shown Nosková losing a Wimbledon final she won and Eala losing two finals she won.
These are pre-match briefs, compiled and published before play. Every edition is scored in the next day's, against what actually happened. Nothing here is betting advice; odds appear because they are the most honest available summary of what the market expects, not as a recommendation.