Ibrox Noise and our very own Tom Anderson have been working together on a discovery Tom has been looking into regarding Rangers’ analytics and data forensics system.
Regulars to Ibrox Noise will know we’ve delved into the system Tony Bloom has, Jamestown, at Hearts, Como, USG and Brighton, and we’ve been looking into how Rangers’ own counterpart, Kitman Labs compares.
Using collaborative research and investigation, Tom dug into Rangers’ use of Kitman Labs, and how Paul Quinn, an Everton fan and major investigator in the Analysis series, dug himself into Starlizard, (that drives Jamestown), as well. Crucially, the ‘core differences’ between Kitman Labs and Jamestown.
We present his findings:
“From a review of information that is available on the Web, assuming that this is up to date, Rangers FC uses an integral data analytics ecosystem that is anchored by Kitman Labs to unify performance, medicine, sports science and coaching operations.
Kitman Labs is based at Mento Park, California and has operations in Glasgow. It is a Global Enterprise SaaS Platform that provides a single centralised operating system for performance intelligence in Sports. (SaaS platforms are cloud-based software delivery systems where applications are hosted remotely and accessed by users over the internet through a Web Browser or mobile app).
According to available information, Rangers use the system for performance, medicine, performance optimisation and coaching development from academy to first team for both men’s and women’s teams.
The System was brought to the club by Creag Robertson, (now Head of Football Operations at Al-Qadsiah Saudi Club.
Kitman Labs is also used by Bayer Leverkusen, Fulham, Chelsea, Ipswich Town, and Liverpool and represents the industry’s largest network of elite and youth organisations, (2,000+), and is the industry’s largest dataset of talent, performance and medical labs.
It is considered to be one of the world’s leading sports and performance intelligence companies worldwide. The technology is designed to optimise human performance, enhance overall health wellness and longevity, reduce industry risk and drive operational efficiency.
It holds medical history all in one place, holds individually player data, mental status, injury recovery, and training feedback.
It’s used for injury lifecycle management
It allows the user to build a complete picture of every player across any metrics critical to performance: new, where been, and how they stack up against others.
Plus it holds centralised training data that can be configured and used by coaching teams. It allows for the building and tracking of plans for training sessions aligned to the principles of playing styles, and for planning workload management
It also allows the club at all levels to understand the impact of injuries to individual players and team performance and should help to highlight alternative players and formations.
For scouting new players, Rangers uses a multi-platform data operation combining a number of analytics providers rather than relying on a single system.
These include:
StatsBomb, (part of Hudl), since 2022
Comparisonator
Wyscout and Opta (Video Scouting Libraries)
Transfer Labs (Analytics FC) that provides predictive performance metrics, player profile scores etc for statistical recruitment matching.
While the risk of using Data Analytics can’t measure a player’s mental resilience, leadership and how they will handle playing for Rangers, (or a major club), or playing in Scotland, they provide significant information that quality analysts working with a knowledgeable coaching team can utilise to work positively with the recruitment team.
The analysis is only as good as the data quality and the expertise of the coaching and scouting departments, however given the integrity of the data we can assume that the onus falls on the scouting, analysis and coaching team working in tandem to secure players with strong odds of success.
Spreadsheets
It follows therefore that while ‘spreadsheet players’ look elite on paper, that they could lack the physical or technical reality required.
If we look at the Jamestown Analytics Model as a comparator, it uses the same analytics Model as is used to beat Global Sports Betting Markets, (Starlizard), but has been repurposed to identify undervalued football talent.
The system rejects traditional football scouting metrics in favour of probability and clinical data.
It eliminates human error such as scout or coach sentimentality, recency bias, and focuses solely on probability of success metrics and targets tactical set ups and coaching staff appointments to find elite talent from obscure leagues.
Jamestown Analytics
Jamestown Analytics is in essence an external recruitment strategy weapon designed to find undervalued players.
This review below of Tony Bloom and Starlizard provides a good insight into why this is one of the best systems currently in use and of course Heart of Midlothian has sole rights to the technology in Scotland under the agreement made when Bloom invested £10m for a 23% stake in the club.
Core Differences Between Kitman Labs and Jamestown Analytics at a Glance
| Feature | Jamestown Analytics | Kitman Labs |
| Primary Purpose | Player recruitment, scouting, and match outcome probability. | Injury prevention, medical tracking, and academy development. |
| Data Source | Global external match data (video, event tracking, league metrics). | Internal player data (GPS trackers, medical scans, wellness surveys). |
| Key Users | Sporting Directors, Recruitment Heads, and Club Owners. | Sport Scientists, Medical Staff, and Academy Directors. |
| Commercial Model | Exclusive, highly guarded club-by-club partnerships. | Broad industry integration across hundreds of global teams. |
The choice between Jamestown Analytics and a conglomeration of StatsBomb, Comparisonator, Wyscout, Opta, and Transfer Lab comes down to Bespoke Consultancy vs. Full Platform Control.
Jamestown Analytics represents an exclusive, highly accurate consultancy built out of Tony Bloom’s Starlizard betting infrastructure that delivers answers (vetted player shortlists).
Conglomeration
In contrast, the conglomeration gives you the most complete, raw, and visual data stack in world football, allowing an in-house analytics department to build anything they want.
Opta vs. StatsBomb vs. Wyscout – Which Football Data Provider is Best?
According to various sources on the Web, every professional club eventually asks the same question, which football data subscription do we actually need?
The three names that seem to appear first are generally the same, Opta (Stats Perform), StatsBomb (Now part of Hudl)), and Wyscout, (Hudl). They are the reference points and which one you choose shapes how the analysts, scouts and coaches work every day.
The general view by experts is that they are not interchangeable, so for most top teams it will not be which one, it will be which combination gives us the best results, or works best for our analysts and coaches etc., and there is also the cost versus value question.
Most top clubs in the EPL subscribe to more than one provider so that they can triangulate metrics. They will use a provider -agnostic analytics layer like GAMECODE.Ai as a layer that turns the stats into metrics owned by the club.
(I haven’t been able to establish from publicly available records whether Rangers uses GAMECODE Ai. What we do know is that they work with various top clubs across major leagues like the EPL, Bundesliga, La Liga and the MLS. We also know that Rangers heavily incorporates data analytics and artificial intelligence into their operations through partnerships with various performance platforms).
By supporting a wide range of algorithms and models from top developers GAMECODE Ai ensures compatibility with leading sports data tools, and empowers the club, (scouts, analysts and coaches), to access the best insights regardless of preferred technology.
If you’re still here at this point, well done!
Tom’s analysis of this research is as follows:
From the brief review undertaken, (and I have gone into a bit more research looking at the success rates that are available from the systems data available, generally short tables, pie charts etc.), there are a number of concerns emanating from this.
By chopping and changing analysts, coaches and to an extent, scouts, (the latter, principally at Director and Head of Scouting level), going back as far as when we implemented the main data analytics platform, we have most likely lost a lot of the expertise associated with interpreting and using the data. This is most relevant given the changes in coaching teams.
While we have changed the coaching teams repeatedly, we have retained a number of old-style scouts where this system and our use of analytics may be alien to these scouts. Old habits don’t go away easily!
We have lost some good guys along the way that have moved on for promotion or become disillusioned with the club. This is particularly relevant where we are looking at our academy and wanting to develop our own players.
Appointing wrong managers, (Head Coaches), is one of the major problems assuming that we are committed to and believe in the systems deployed. Logically we should have looked at the clubs using the same technology, assuming we were not going to spend top dollar to bring in a seasoned coach with a proven track record. The alternative would be to appoint a manager/coach that has the top-quality contacts that allows us to confidently override the system and the analysts.
Overriding
Allowing coaches to override the system where the odds based on the algorithms and analysis are stacked against the player being a success. (This has happened on a number of occasions recently where coaching teams and managers have overridden the system to bring in players that they wanted and none of them have been a success)
Unfortunately, where the manager / coach has overridden the system to reject a player we have no way of knowing what the outcome would have been had they signed / not been sold.
The last concern for this commentary is related to the data – are we using correct data and are we interpreting it correctly. As far as actual data is concerned this should be fairly accurate as it is being gathered electronically and a lot of it, from a scouting perspective is gathered externally from a multitude of sources. The only question in respect of data input therefor is in respect of what is being input at Rangers. This is particularly relevant to training and actual game play. Individual player scores for games played is subjective and consequently questionable but sufficient analytical data should be available to guide a quality manager and their coaching team on who fits for system A, system B, etc., and what is needed against Team A, Team B, etc.
This ignores whether we are predominantly looking at players for onward sale at a profit, which is the prime objective of the Tony Bloom System, or whether the prime objective of the analytics is to win trophies, which will most definitely be that prime objective of the top teams in the EPL and Europe.
Build a squad
My view on this one is that we should be able to build a squad that has a number of players that are bought for profit while having a small nucleus of players that have a dual role of helping to develop these players while adding the quality and leadership / established professionalism providing we get it right with our managerial and coaching appointments. In my opinion this is a major contributor to our problems but as can hopefully be seen from the commentary it is just one of a number of issues.
There is a lot more technical data available about the technology that we deploy, I’ve also included a recent article on Bloom and his technology, (no point in recreating the wheel). I’m very much on the side of using Data, (probably from my background in Banking, Insurance and Credit Risk Management / Country Risk Profiling), and I know a little bit about SaaS from living close to their Headquarters in Marlow.
While the underlying data is totally different, the underlying premise of odds is constant. This is perhaps where Bloom’s system works best as it relies less on human interaction and generally works on probability odds. This is very similar to the way that lending and insurance risk works. It takes mass amounts of related data and then gives a predictive score. The user then determines where their cut off is for what is acceptable and what is not. In the middle is a grey area.
Manual interaction
This area requires further use of neural networks or manual interaction and is almost likely to be ignored by Bloom’s analysts in terms of his Jamestown Technology System. When we are using the systems we use we are almost definitely depending more on analyst interpretation and Coaching Team / Scouting Team interaction and understanding alongside the analysts.
One concern here is that as we move from priority one for a position to lower-level priorities the failure risk grows, and they become closer to this grey area. I’m concerned that while we might have good analysts, we don’t have a quality scouting department or Coaching team that are capable of working together to quickly establish and secure the priority targets.
There is also the likelihood that we are rejecting better players to fit with positions that we need to fill and there has historically been too much Manager override of analytical data that has resulted in the acquisition of poor players and coaching teams.
Still with us?
So the basics of Tom’s research into Quinn’s deep analysis of Starlizard and his review of Kitman Labs and the Multi-platform data operation used by us can be summed up thus:
Good tech, bad people. Sort of.
While Jamestown is an outstanding data analysis system, it’s been built, developed and run by the same people who then operate it at club level. It was also designed as a successful gambling/risk/odds prediction tool which was repurposed for sporting benefit.
Meanwhile Kitman has been built as an excellent tool itself but is then provided alongside the other ‘Scouting Systems’ that we refer to above, to the end user who has to figure it out themselves. And that’s a big failing at Rangers end, which isn’t strictly anyone’s fault. It’s the nature of the beast.
Both systems great, but one is being operated by people who understand it. The other isn’t so much. As well as Rangers’ constant recycling of staff which means just when one Chief Scout/Coaching Team understands the system, they end up moving on anyway.
Overall Ibrox Noise Conclusion:
Until Rangers are stabilised in the scouting and recruitment in terms of quality staff and has a stable quality coaching team it doesn’t matter what system we have. The more turnover, the more unsettled, and the less well our tools get used.