Nacsport AI in the Lecture Theatre: How Dr Mark Quinn Teaches the Next Generation of Analysts at Salford
- Ollie Seymour
- 2 days ago
- 7 min read
What Nacsport AI Is
Nacsport AI is a paid plugin for Scout, Pro and Elite users. It puts a Large Language Model (LLM) directly inside the software, working on your own tagged data. Ask it a question in plain language and it returns an answer drawn from your analysis.
The plugin sits in three environments inside Nacsport:
The Timeline. AI reads the full tagged matrix, giving you league-wide or season-wide insights from a single prompt.
The Clip Filter (Pro and Elite). AI works on the results of your filter, perfect for interrogating a tightly scoped slice of the data.
The Presentations Environment. AI takes its context from your active list, ideal for generating focused output for a specific meeting or debrief.
There is a generous token allowance as standard. A 12-month licence ships with 30 million tokens, more than enough to get started, and top-ups are available in packs whenever you need them.
Two ingredients turn Nacsport AI into a serious analytical asset: well-structured tagged data and well-written prompts. Invest in both and you compress hours of manual analysis review into minutes, with charts, tables and written summaries delivered in a single output.
This case study is about a programme that has built its teaching around exactly that combination.
Nacsport AI in the Classroom
Dr Mark Quinn leads the Sport and Exercise Science undergraduate degree and the MSc in Performance Analysis in Sport at the University of Salford. His route into teaching ran through elite rugby league. He spent seven years as an applied sport scientist at the Wigan Warriors between 2010 and 2017, a period that brought the club significant Super League success.
That background shapes how he runs the programme at Salford. The brief is simple. Students should leave with the same tools and the same workflows used inside professional clubs.
"The benefit of this is that it makes the gap between applied practice and what we do at the university much smaller. Students can practise and simulate what it's like working in the real world, so when they go out to work with clubs, teams and athletes, they've already had that experience of using the software and using the AI."
That principle now extends to AI. Salford has placed Nacsport AI at the centre of the MSc workflow, giving students hands-on experience with the same tool that professional analysts are adopting across the industry.
The Project: Tries from Kicks
The project Mark's cohort worked through is a forensic study of how tries are scored from kicks across a full Super League season. Over five weeks, students tagged every try-scoring kick across the league's 12 clubs. They then used Nacsport AI to query the dataset.
The real story is in the structure behind it.
Building a Tagging Window That AI Can Read
The project starts in the tagging window, and the design of that window does a lot of the heavy lifting. Quinn and the team built a custom Tagging window optimised for AI, organised around five analytical sections:
Field Position & Context: Location, distance from try line, kicking foot, tackle count, time band, play-the-ball speed, state of play.
Kick Design & Execution: Type (grubber, chip, bomb, cross field, other), depth, direction, trajectory, target.
Defensive System Cues: Pressure on the kicker, backfield numbers, fullback position, winger and D-line shape, marker count.
Chase & Contest: Contest type, chase intensity, proximity, chase band, receiver action.
Try Mechanism: Primary cause, defensive error, point of failure, scorer position.
Every descriptor is named with intent. The descriptor isn't just a label. It's a piece of context the AI will read later.
"Within that descriptor we have not only the descriptor but also the theme embedded within it. This helps the AI to understand exactly what it is looking at."
It's a smart design principle and one any Nacsport user can adopt. Naming descriptors with intent turns the tagging window into a richer source of context for the AI, which feeds straight into the quality of the output.
Data Consistency Comes First
A custom tagging window is at its most powerful when the data poured into it is consistent. Salford runs the project the way an analysis department would inside a club.
Students were given a glossary of definitions and video examples for every key performance indicator. Quinn walked the cohort through each one in a group session so everyone tagged to the same standard. Working in small groups during the tagging itself meant students could clarify any ambiguous moment in real time, the same way analysts do when they share an office with the coaching staff.
"It's one of the main things we focus on at the university here at Salford: validity, reliability and consistency with the performance analysis work that we do."
Consistent tagging is what allows Nacsport AI to find genuine patterns at scale. The cleaner the data going in, the sharper the insights coming out.
The RCTF Prompt Framework
With the tagging finished, students moved to the AI section inside Nacsport. This is where the second layer of structure comes in.
Mark teaches a prompt framework called RCTF, which stands for Role, Context, Task, Format. It is the spine of every prompt the students write.
Role: Who is the AI meant to be? A rugby league performance analyst, in this case.
Context: What's the dataset? Kicks tagged across the Super League season.
Task: What's the question? Generate a league-wide summary of all 12 teams.
Format: How should the output look? A written introduction, a data table, a bar chart and a written summary.
For multi-step jobs, students chain the steps inside the prompt itself. Step one does X. Step two does Y. The AI follows the structure cleanly and returns exactly what's asked of it.
The dictionary inside the Nacsport AI section adds the final piece. Students populate it with the key performance indicators they've used. That gives the AI the vocabulary it needs to read the data accurately, particularly useful when working with sport-specific acronyms and terminology.
What the AI Produced
The output from the RCTF prompt is where the project earns its keep. From one structured query, students received a four-section report covering an introduction, a data table, a graphical analysis and a written summary of key trends across all 12 Super League teams.
Some of the findings the AI surfaced:
The grubber kick produced 54 of the 139 tries, around 39 per cent of all tries scored from kicks across the season.
Warrington Wolves and Wigan Warriors finished joint-top with 18 tries from kicks each. The two clubs got there in completely different ways.
Warrington relied heavily on the grubber, scoring 11 of their 18 kick-tries that way. Wigan went the opposite direction, scoring 10 of their 18 from chip kicks, an outlier across the league.
Hull FC and Castleford Tigers were the most effective cross-field exponents, with 6 and 5 tries respectively.
Wakefield Trinity led the league in tries from bomb kicks, with 5.
The students had this information in front of them in minutes, with the chart and table already formatted.
"It does it really quickly. Something that would normally take the analyst quite a lot of time to go and look into, you can get the answer straight away, and then you can start to do further analysis, check those responses, build playlists and presentations on the back of using the AI."
The Analyst Stays at the Core
Quinn is careful about how the AI's role is framed for his students. Nacsport AI is built to support the analyst, not replace them. It's a tool that lets the analyst ask more questions, faster, and dig deeper into the data than time would normally allow.
"The analyst is still at the core and drives the work. It's the analyst who asks the AI the questions, it's the analyst who does all the tagging and builds the presentations. What the AI can do is uncover things that you might not automatically see in the data, and you can query many questions very quickly and get a response to help you with your workflow."
That framing matters for a sport science degree. Students are taught to design the question, design the dataset, design the prompt and then interpret the output critically. AI becomes an accelerator for their thinking, not a substitute for it.
Why Salford Teaches AI in Year One
Some universities have taken a cautious line on AI in assessment. Salford has gone the other way and embedded it into modules from the first year.
Quinn's reasoning is practical. Students are going to use AI regardless. The job of the university is to teach them the responsible way to do it.
"What we want to do at the university is show them the responsible ways to use AI, what it can be good for, some of the limitations, and how they can integrate it into their own workflows within performance analysis."
That covers prompt structure, workflow integration and critical evaluation of AI output. By the time students leave Salford with an MSc, they've already refined their prompts through real projects and built genuine fluency with the tool.
A Real Analysis Lab
The infrastructure behind the programme is built to mirror an applied environment. The MSc cohort works in a dedicated computer lab with 20 Nacsport Scout licences. Every student tags, analyses and queries on the same software running inside professional Rugby League clubs.
Paired with the AI plugin, the lab is a working analysis department. Students aren't being taught about Nacsport in the abstract. They're producing reports inside it.
What Other Universities and Analysts Can Take From This
This workflow is replicable. The components are available to any Nacsport user with the AI plugin, and the methodology travels across sports.
The transferable lessons:
Design the tagging window with AI in mind. Names and themes inside descriptors give the AI rich context to work with.
Invest in data consistency. Glossaries, video examples, group calibration sessions and small-group tagging give Nacsport AI the clean foundation it needs to deliver sharp insights.
Use a prompt framework. RCTF gives the AI a clear job description and produces structured, presentation-ready output every time.
Populate the AI dictionary. Teaching the AI what your KPIs mean is one of the quickest ways to lift the quality of every output that follows.
Build the data discipline first. Layer Nacsport AI on top. If you would like to learn more about Nacsport AI you here, or you can contact us directly and we can help you get started.
Comments