Source Integrity in Sports Reporting: How to Move from Empty Input to Verified News

This report is not about a specific match, player or tournament. It is about the process by which a sports news story is produced — and about the situation in which that process breaks down. The input supplied here carries one clear signal: the Stage-1 analysis returned completely empty. The article title is unavailable, the source is unavailable, the article type and author stance are unrecorded, the one-sentence summary is blank, and most importantly, the Information Points list is entirely empty. Only one signal survives: the domain label 'basketball'. That label tells us the subject area, but it carries no fact, statistic, quote or claim. If a writer were to attempt a 'deep analysis' from this position, they would have to invent scorelines, injury reports, coach statements and timelines. Such an article would look complete on the surface and be entirely false underneath. Readers would treat it as fact, share it, and form opinions on it. That is not journalism; it is the manufacture of misinformation. The first and most honest conclusion of this report is therefore: no genuine sports news article can be written from the available input. But stopping at 'it cannot be written' is also insufficient. From an engineering standpoint, a useful answer must say what to do next, in what order, and through which verification gates. This report sets out that framework. Section one: why an empty input cannot produce a report A news article is a chain of facts. Its structure comes from language, but its substance comes from fact. Without fact, language is only style. A sports report requires at minimum: who (player/team/coach/institution), what (event or announcement), when (date and time zone), where (venue or competition), how (circumstance or cause), and source (where this information came from). If any one of these is missing, the sentence is incomplete; if all are missing, the sentence cannot be formed at all. The Stage-1 output sits exactly in that position. Who, what and when are all absent. Analysis then has no basis, because the work of analysis is to find relationships between facts, not to invent them. One confusion must be cleared up here. Some assume a domain label is sufficient signal — that if we know the subject is basketball, general knowledge will carry the writing. This fails for two reasons. First, basketball is an enormous field with hundreds of leagues, thousands of teams and decades of history; saying 'basketball' is as vague as saying 'sport'. Second, an article written from general knowledge is not news — it is background. News is defined as new information. Without new information, a news article is merely general discussion. Section two: what a minimum fact unit looks like A usable report needs at least five to ten atomic facts. An atomic fact is a sentence that cannot be broken down further, containing one subject, one action or claim, and one traceable source. Consider the structure, with facts replaced by brackets: [Source: official league press release, date] — [Team A] beat [Team B] [score] to win the [competition] title. [Source: club medical report, date] — [Player X] is out for the remainder of the season with a knee injury. [Source: post-match press conference, date] — [Coach Y] said the defensive structure had to change in the second quarter. [Source: official statistics portal, date] — [Player Z] averaged [number] points per game. Note that each line requires three things: subject, claim, source marker. The third is the most neglected and the most important. A sentence without a source marker is not information; it belongs in the category of rumour. If an automated pipeline returns this list empty, it means either the source document was never reached, or text extraction failed, or the language model stage returned an empty array. In all three cases the remedy is diagnosis, not writing. Section three: source tiers and reliability Sports journalism sources fall into three tiers. Tier one — direct sources: official press releases, league or federation websites, club announcements, post-match press conferences, direct player statements, official statistics portals. These are the strongest basis because they can be traced and involve little mediation. Tier two — indirect sources: news agency reports, restatements by other outlets, analyst interviews, posts from official social media handles. These are useful but carry transmission error. They should be verified against a tier-one source. Tier three — unspecified claims: phrases such as 'according to sources', 'it is said', 'it is believed', where no name or date is given. These may be used only if explicitly labelled unconfirmed and no conclusion is built on them. The rule that protects reader trust is simple: the bigger the claim, the higher the source tier required. Section four: time sensitivity Sports journalism works with perishable facts. A player's injury status, a team's ranking, the opening and closing of a transfer window, a coaching appointment, a competition calendar — all can change in a week or a day. Every fact therefore needs a timestamp. 'The player is injured' is insufficient; 'according to the club's dated medical report, the player was injured at that time' is required. The difference looks subtle but is decisive: the first sentence becomes a permanent claim, the second remains a time-bounded fact and the reader knows when to check it. A systematic report should also flag which conclusions expire quickly. An analysis based on mid-season rankings will be stale before the playoffs; a conclusion drawn from a single match's statistics will need updating after the next match. Such a warning is a kind of contract with the reader: this is the limit of my knowledge. Section five: the entity graph A sports report builds a web of people and institutions: players, teams, coaches, assistant staff, medical teams, league administration, referees, agents, sponsors, supporter groups. Only some of these appear in any given article, and their weight determines where the article's centre lies. Building an entity graph reveals hidden relationships — who has an interest in whom, through whom information is arriving, and which source is independent. If three officials from the same club say the same thing, that is not three sources; it is one source three times. This distinction is among the most neglected and most misleading in sports journalism. With an empty input, the graph cannot be built at all, because nodes come from facts, not imagination. Section six: bias and blind spots Every report carries a narrative frame: the winner's story, the rising player's story, the team in crisis, the management's failure. The frame is necessary because news needs meaning, but it is also dangerous because it makes choices — what to show and what to omit. A responsible report should therefore state what information is unavailable. If a team's loss is attributed to coaching strategy, the effect of a key player's absence should also be recorded; if a player is being praised, the size of the sample should also be noted. The question of what was omitted matters no less than what was written. Section seven: a practical report structure When facts are available, the report can be built in this order. A headline that leads with the most important fact and contains no unverified adjectives. A lead answering four questions: what happened, who was involved, when and where, and why it matters. A second paragraph reinforcing the fact with the strongest source. Middle paragraphs giving context — prior performance, competition standing, historical comparison — where the greatest risk lies: presenting context and opinion as fact. A closing paragraph on what comes next: the next match, an expected announcement, an unresolved question. When facts are not available, only placeholders can fill this structure, and they must be clearly marked as such. For example: '[Team A] beat [Team B] [score] in a match played on [date] at [venue]. This information is based on [source name and date].' Such a sentence is not publishable, but it is producible: filling it takes a reporter fifteen minutes, whereas starting from a blank page takes hours. Section eight: diagnosing the pipeline If the empty output came from an automated system, the order of inspection should be as follows. First, fetching: did the source URL return 403, 404 or an empty response, and was the page JavaScript-rendered with no content in static HTML? Second, extraction: does the CSS selector or parser fail to match the page structure, or was content inside an iframe or hidden div? Third, the language model stage: was the input truncated beyond the token limit, was the output schema strict enough that the model satisfied it by returning an empty array, or did the prompt instruct 'return empty if no information' and thereby block valid content? Fourth, validation: did the schema accept an empty array as valid? A good schema should require at least one fact so that empty results cannot pass silently. Fixing any one of these points is far more productive than running downstream analysis on an empty result. An article built on empty input does not save time; it multiplies correction work, because someone must later determine which sentence was real and which was fabricated. Section nine: minimum acceptable input If the goal is simply to produce an article, either of two inputs will suffice. Option one: the original text, either pasted in full or as an accessible link or file, so extraction can be re-run. Option two: complete the Stage-1 fields with at least five to ten atomic facts, each with subject, claim and source marker, plus title, source name, article type (news/analysis/opinion/feature) and author stance. With either in hand, the work becomes clear: argument structure breakdown, cross-verification of information points, entity graph, time-sensitivity determination, source quality grading, and bias identification. Section ten: conclusion Sports journalism operates under heavy speed pressure. A match ends and a report is expected within minutes. Under that pressure the easiest route seems to be filling what is missing with general knowledge or inference. That route is the greatest single threat to credibility. A wrong scoreline, an invented quote, an unconfirmed injury report — the impact extends far beyond one article. A reader deceived once begins to doubt everything. That doubt matters democratically, because accountability of sports institutions is built through the same channel. The final recommendation of this report is therefore simple. When the input is empty, do not write the article — fix the input first. When facts are available, ask of every sentence: where did this come from, when is it from, and who can verify it? A sentence that cannot answer those three questions is not news. A good sports report is not the one that reads most dramatically. A good sports report is one whose every claim a reader can check for themselves. That is the only standard that holds over time.