AI football training plans: how to prompt and how to check

What an AI football training plan is good for, and where it fails

Text assistants generate plausible coaching content because they have read a great deal of it. That makes them useful for some tasks and risky for others.

The pattern is consistent: structure and language are strengths, facts that change and numbers that matter for safety are weaknesses. Treat every number in an AI draft as unverified until you have checked it.

How to write a prompt that produces a usable session

A vague request ("give me a passing session for kids") returns a generic one. The assistant cannot see your players, your pitch or last Saturday's match, so the prompt has to supply them. Include six things: age and level, numbers, space and equipment, duration, the objective linked to your game model, and the constraints.

Three habits improve the output further. First, ask for the reasoning: "explain why each practice fits the objective" exposes weak links. Second, ask for a version for fewer players, because attendance changes. Third, follow up with corrections in the same conversation ("the second practice has eight players waiting; redesign it so everyone is active") rather than starting again.

For a season outline, give the assistant the fixture dates, school holidays, half-term breaks and the number of sessions per week, and ask for themes in four- to six-week blocks. The season planning guide gives a structure to compare it with.

Reviewing the draft: a coach's checklist

Run every AI-generated session through the same questions before it reaches the pitch. It takes five minutes and catches most problems.

1. Objective. Does each practice actually create the situation you want to train, with opposition and decisions, or is it an unopposed pattern?
2. Numbers and space. Do the players add up? Divide the area by the number of outfield players; for possession practices with young players, very small areas per player make the game a scramble.
3. Activity. How many players are waiting at any moment? Queues are the most common flaw in AI drafts.
4. Load. Count the minutes of high-intensity work and the rest. AI drafts often stack three demanding practices with short breaks.
5. Age and rules. Check formats, ball size and match rules against current FA guidance. From 2026/27 the FA pathway starts with 3v3 at under-7 and each format moves up a year; deliberate heading is being phased out of matches from under-7 to under-11.
6. Safety. Goals anchored, distances between areas, surfaces, weather.
7. Transfer. Does the final game look like the match your team plays on Saturday?

Typical errors in AI-generated sessions and how to fix them

A useful follow-up prompt after the review: "List every number in this session (area, players, minutes, rest) in a table, and flag any that you are not certain about." Assistants are not reliable judges of their own uncertainty, but the table makes your own check faster.

Running tournaments is one of the most time-consuming jobs at a grassroots club: registrations, the draw, the schedule and results. DorsalCUP does it for free, and the tournament fixture generator builds a one-day schedule in minutes, no sign-up.

Personal data, injuries and the coach's responsibility

Do not paste players' names, dates of birth, injuries or behaviour notes into a general-purpose assistant. Health information is special category data under UK GDPR, and anything typed into a consumer tool may be stored and, depending on the settings, used to improve the service. Describe the group instead: "two players returning from ankle injuries, cleared by a physio for non-contact work" carries what the plan needs without identifying anyone.

• Use a club or work account with data-sharing for training switched off where the tool allows it.
• Keep injury and medical decisions with qualified people; an assistant cannot assess a growing adolescent's knee.
• Save the final, reviewed session in your own planner, not only in the chat history.
• If the club adopts AI tools formally, write one page of guidance for coaches: what can be entered, what cannot, who reviews.

The ICO's guidance on AI and data protection applies to clubs as to any organisation; it notes that it is under review after the Data (Use and Access) Act 2025, so check the current version. Responsibility does not move to the software: the coach who runs the session answers for its content, its safety and its suitability for the age group.

A simple workflow for the next four weeks

1. Write one reusable prompt template with your team's context (age, numbers, pitch, kit, game model principles). Store it in a note.
2. Each week, add the objective from the last match and ask for a session.
3. Review with the checklist and correct in the same conversation.
4. Run the session; after it, write three lines on what worked and what did not.
5. Feed those lines into next week's prompt.
6. After four weeks, compare: did the AI drafts save time, and did the sessions improve? Keep the tool only if both answers are yes.

Coaches who get most from these tools treat them like a junior assistant coach with wide reading and no pitch experience: fast, tireless, occasionally confidently wrong, and never left in charge.

Frequently asked questions