New data analyst

Forums New posts
Assessing whether a data analyst’s "models worked" is tricky because, unlike a striker’s goals, the results are often hidden in the spreadsheet. However, looking at Reading FC’s performance and internal shifts during Sean Richardson’s tenure (Sept 2025 – April 2026), there are strong indicators that his work was highly effective.
Here is the breakdown of the "Richardson Effect" at Reading:

1. The "Small Squad" Theory
One of Richardson’s most specific models focused on squad architecture. He crunched the numbers to prove that in League One, clubs with smaller, more consistent squads actually achieved higher points-per-game than those who rotated heavily.
The Result: Reading moved toward a leaner squad during the winter window. By January 2026, they had the second-best points-per-game in the league (2.0) since the appointment of manager Leam Richardson.

2. Recruitment Efficiency on a Budget
Richardson worked under the "Reading Model"—a strategy designed to find value despite the club's recent financial recovery.
The Result: He specifically analyzed the "spread of goals." His data suggested that relying on a single "super striker" was a high-risk strategy. Instead, he modeled a recruitment plan for "secondary scorers."
Outcome: Reading saw significant contributions from across the pitch, with captain Lewis Wing and Jack Marriott both hitting peak numbers simultaneously, validating the "diversified threat" model.

3. The "Festive Period" Peak
Data analysts are often judged on their "load management" models—predicting when players need rest to avoid the "red zone" of injury.
The Result: Over the 2025/26 festive period, Reading picked up 13 out of a possible 15 points, the highest in the Football League. This suggests that the physical and tactical "peaking" models Richardson was involved with were working perfectly when the schedule was most grueling.

4. Validation from Leadership
The strongest evidence that his models worked is the "Stuart Fenton Factor." Fenton, Reading’s Head of AI, is a pioneer in the field. The fact that Richardson was his "right-hand man" in building the "Reading Model" (an AI-driven system for match analysis) speaks volumes.
The Verdict: Hibernian didn't just hire him for his potential; they hired him because Reading’s data department became the envy of the EFL in 2026. The "models" transitioned Reading from a club in financial and tactical chaos to one of the most statistically efficient teams in League One.

Seems like a decent appointment from the club. (Sorry for the spiel I canny seem to upload screenshots)
 
. The "Small Squad" Theory
One of Richardson’s most specific models focused on squad architecture. He crunched the numbers to prove that in League One, clubs with smaller, more consistent squads actually achieved higher points-per-game than those who rotated heavily.////

To an extent that held true for The Tornadoes
The problems arose after Brownlie got his leg break and a couple of suspensions to key players came in (Edwards a huge unwarranted ban for one )
John Hazel and Johnny Hamilton,both forwards,were the main "reserves" so to speak, and although we had like others, a reserve team at the time who played in Reserve League, the quailty there didn't yet match "the" Tornadoes .
Adding quality to the squad was a debate amongst fans.
Turnbull bought Joe Harper, which was fine.
Selling Jimmy O' to cater for Joe wasn't. That wasn't Joes fault .
I think by then Jock Stein had really got into Eddies head (he hated Jock) and Jock won hands down.
So yes small squads have proven some success
Remarkable when you think that tackles that get punished,with cards, these days,were the norm back then.
The Tornadoes certainly ate their Weetabix !!
 
Last edited:
Emerald Harp said:
. The "Small Squad" Theory
One of Richardson’s most specific models focused on squad architecture. He crunched the numbers to prove that in League One, clubs with smaller, more consistent squads actually achieved higher points-per-game than those who rotated heavily.////

To an extent that held true for The Tornadoes
The problems arose after Brownlie got his leg break and a couple of suspensions to key players came in (Edwards a huge unwarranted ban for one )
John Hazel and Johnny Hamilton,both forwards,were the main "reserves" so to speak, and although we had like others, a reserve team at the time who played in Reserve League, the quailty there didn't yet match "the" Tornadoes .
Adding quality to the squad was a debate amongst fans.
Turnbull bought Joe Harper, which was fine.
Selling Jimmy O' to cater for Joe wasn't. That wasn't Joes fault .
I think by then Jock Stein had really got into Eddies head (he hated Jock) and Jock won hands down.
So yes small squads have proven some success
Remarkable when you think that tackles that get punished,with cards, these days,were the norm back then.
The Tornadoes certainly ate their Weetabix !!
Click to expand...
Des Bremner came in when Brownlie was injured, he wasn't too bad. I think the legendary Tornadoes team we can all recite only played together 22 times. Reserves like Tony Higgins made their firstteam debuts before the classic XI came together and in my opinion along with Hazel, Hamilton, McArthur etc can be considered Tornadoes just as much as the first choice side, for two or three years they were all part of a side you went along to watch thinking I wonder how many we will win by this week.
 
Emerald Harp said:
. The "Small Squad" Theory
One of Richardson’s most specific models focused on squad architecture. He crunched the numbers to prove that in League One, clubs with smaller, more consistent squads actually achieved higher points-per-game than those who rotated heavily.////

To an extent that held true for The Tornadoes
The problems arose after Brownlie got his leg break and a couple of suspensions to key players came in (Edwards a huge unwarranted ban for one )
John Hazel and Johnny Hamilton,both forwards,were the main "reserves" so to speak, and although we had like others, a reserve team at the time who played in Reserve League, the quailty there didn't yet match "the" Tornadoes .
Adding quality to the squad was a debate amongst fans.
Turnbull bought Joe Harper, which was fine.
Selling Jimmy O' to cater for Joe wasn't. That wasn't Joes fault .
I think by then Jock Stein had really got into Eddies head (he hated Jock) and Jock won hands down.
So yes small squads have proven some success
Remarkable when you think that tackles that get punished,with cards, these days,were the norm back then.
The Tornadoes certainly ate their Weetabix !!
Click to expand...
The way we've been recruiting of late by signing players who can play multiple positions this appointment makes perfect sense. I was always under the impression that we sold O'Rourke way too soon. I think turnbulls biggest downfall was breaking up that squad way too early, they certainly should have won more than they should but they were unlucky in the sense that they were up against arguably one of the best teams to ever play in Scotland in the Lisbon lions.
 
Zab said:
Ridiculous eh! Can only hope the next generational manager to replace Pep is completely anti data, and everyone copies that.
Click to expand...
Aye hopefully but I doubt it, that man has completely ruined the beautiful game. If it wasn't for Hibs and Coupons I would have stopped watching football years ago!
 
  • Like
Reactions: Zab
Interesting.
I'm entirely in favour of data analysis.
One problem with data analysis in football is that it's relatively new. This means most data analysts will be winging it to an extent.
Everyone in professional baseball now understands that the more obvious statistical traits Bill James used to identify under-appreciated players are/were... ehhh... "obvious" and that they ought to have been identified sooner and used before Billy Beane adopted them at Oakland. These had long been measured by widely published figures & stats - and completely ignored. See "Moneyball."
"Why do you like him?"
"Come on, guys, or do I have to point at Pete again?"
"He gets on base!"
Even in baseball, a lot of the Bill James theories and analyses that cannot be measured by a number in an officially recognised statistical category remain contentious.
Football doesn't have many concrete stats. Almost universal opinion is that it has very few others that can easily identify value.
The geeks behind Jamestown Analytics might dispute that, but nobody outside the Jamestown team has access to their methods.
The "small squad" theory makes sense on most levels. In football, as in everything else, being familiar with your surroundings and comfortable with them tends to lead to better performance than if you're operating in unfamiliar circumstances.
Aston Villa winning the English League with 14 players isn't going to be repeated. It was mainly 12: Deacey & Geddis had a handful of games; Williams & Gibson split the left-back spot; the other 10, including Des Bremner, played around 40 games each.
Sports science & nutrition can improve cardiovascular fitness and muscles, but it has no effectiveness on tendons, ligaments & cartilages, which are more likely to go "pop" when a bigger 2020s player turns more quickly than his smaller 1980s counterpart.
Getting someone who can do data analysis well could make a huge difference, but there's nobody qualified to confidently assess the merits of data analysts, so it's something of a pish in the dark.
 
Last edited:
Glenmore Road said:
Interesting.
I'm entirely in favour of data analysis.
One problem with data analysis in football is that it's relatively new. This means most data analysts will be winging it to an extent.
Everyone in professional baseball now understands that the more obvious statistical traits Bill James used to identify under-appreciated players are/were... ehhh... "obvious" and that they ought to have been identified sooner and used before Billy Beane adopted them at Oakland. These had long been measured by widely published figures & stats - and completely ignored. See "Moneyball."
"Why do you like him?"
"Come on, guys, or do I have to point at Pete again?"
"He gets on base!"
Even in baseball, a lot of the Bill James theories and analyses that cannot be measured by a number in an officially recognised statistical category remain contentious.
Football doesn't have many concrete stats. Almost universal opinion is that it has very few others that can easily identify value.
The geeks behind Jamestown Analytics might dispute that, but nobody outside the Jamestown team has access to their methods.
The "small squad" theory makes sense on most levels. In football, as in everything else, being familiar with your surroundings and comfortable with them tends to lead to better performance than if you're operating in unfamiliar circumstances.
Aston Villa winning the English League with 14 players isn't going to be repeated. It was mainly 12: Deacey & Geddis had a handful of games; Williams & Gibson split the left-back spot; the other 10, including Des Bremner, played around 40 games each.
Sports science & nutrition can improve cardiovascular fitness and muscles, but it has no effectiveness on tendons, ligaments & cartilages, which are much likely to go "pop" when a bigger 2020s player turns more quickly than his smaller 1980s counterpart.
Getting someone who can do data analysis well could make a huge difference, but there's nobody qualified to confidently assess the merits of data analysts, so it's something of a pish in the dark.
Click to expand...
Analytics to managers is like VAR to referees in my opinion. Managers are actually allowing these stats to dictate how they manage games just like referees are allowing VAR to referee the games for them. With Gray you see him making utterly bizarre substitutions in games and I think that's down to the computer/analytics telling him he needs to make subs because certain players are tiring instead of managing the game himself tactically, it's cost us quite a few points this season. I understand why it's there and why they use it but I think it's regressing our managers massively and they rely too much on it.
 
You must log in or register to reply here.
  • Menu
  • Home
  • Hibs Chat
  • New Posts
  • Just Back
  • Chatroom