Even though they are brief, from an editing and style point of view these reports try to make reading easier for the user by highlighting in bold the names of the teams, the football players (goal scorers, replacement players, and goalkeepers when they have been the protagonists), and even the coaches.

The vocabulary used is predictably repetitive. Even so, there is an effort from the programming of the bot in the newsroom to use synonymous expressions that make the report seem less ‘robotized’. In this way, in all of the eighty reports analysed, different expressions are combined in the text when referring to victory, draw and defeat, as well as to situations during the match, especially regarding the evolution and variation in the scoring.

When referring to victory, the verbs “to win”, “to beat” or “to triumph” appear most often, while in draws the expressions “settle for sharing the points”, “tie game”, and “points shared” appear. In the case of a 0-0 result, the most common expressions are “no goals scored”, “no team managed to score”, or “no one was fortunate in scoring a goal”. In this regard, obvious information and redundancies are sometimes reproduced, such as “neither of the teams scored a goal and therefore the scoreboard showed the same score”, or “as no goals were scored, the result showed no alteration”.

Moreover, there are references to the evolution of the results in the course of a match, such as: “they took the lead”, “they took the lead on the scoreboard”, “they managed to turn the scoreboard around”, “they increased the distance on the scoreboard”, “they widened the differences”, “they shortened the distance”, “they expanded their advantage”, “they managed to come back” or “the game started in a positive way for...”. Likewise, comments regarding scoring of the various goals usually begin with, “thanks to the success of...”

In addition, there are alternative ways of referring to the next day’s events (“the next appointment on the calendar”), or to the replacement of players (“the coaches made all possible changes”, or “both teams ran out of changes”, or “both coaches used all the players on the bench”), and a phrase was even used to vary the way to say that half-time had arrived (“the players lined up at the entrance of the changing rooms”).

Apart from the lexical alternatives that the editing department can introduce in order to improve the textual quality of these automated reports, another element that enriches these pieces is without doubt the incorporation of statistical data. These are becoming more and more frequent, not only in subheadings for the purpose of enhancing the headlines (“a win at home will mean a 6th league victory”), but also in the body of the text. The use of statistical data in these reports appears in the lead in order to contextualise the result in a cumulative series throughout the championship (“the locals are coming to the match with four consecutive victories”, or “before the match, the locals had achieved 6 points out of 15 possible, while the visitors came to the competition after winning their last match”).

Therefore, taking into account the objectives set for this research, both the quantitative and qualitative measurement of the naturally-generated language in these reports shows a significant advance in terms of the number of texts produced each day (forty), with each and every one of the teams in the category being included, and full coverage being ensured. In spite of this, all of the doubts about aspects related to the quality of the resulting information have still not been clarified.

Paying special attention to the degree of repetition in language referring to situations of victory, draw or defeat in the course of the matches covered, a clear effort can be observed in the use of synonyms and second references in the texts, as well as the progressive incorporation of statistical data that can improve the quality of the reports.