As well as my recently launched Points Above Replacement metric, I’ve also been working on a model that can rate the performance of head coaches too. Whilst I’m not ready at this point to share the results from that overall model, part of it is ready to use as a first iteration for some historical analysis.
It’s probably fair to say that 90% of a head coach/manager’s hard work takes place before the referee blows the whistle to start proceedings with a kick-off, with the remaining 10% coming on a match day. For a fan though, those percentages are probably inverted when it comes down to what we’re able to talk about.
We don’t know what the training and analysis sessions look like throughout the week - all we can go off is how the players perform in their assigned roles (which is so often up for interpretation too), and how the man in the dugout responds to the game as it happens.
Hindsight is a wonderful thing - it’s very easy to say the starting lineup was wrong after the game, it’s very easy to talk post match about how a substitution was the wrong decision.
My Head Coach Performance Model rates the following decisions relating to a game, attributing a positive or negative impact that is initially driven from the dugout:
Unenforced starting lineup changes
A fit and available player comes in for a player who is relegated to the bench
Starting formation changes
A team sets up in a different formation compared to their previous game
In game substitutions
The personnel changes made during a game (injuries can be excluded if necessary)
In game formation changes
Changes to the broader tactical set-up that take place during the match, perhaps for example to chase a game, or to shut up shop
Within the ratings, we take into account Opta’s Team Strength ranking for both teams, we look at the form of both teams leading up to the game, and also the venue of the game.
Since promotion, Forest have had 6 different men sit in the dugout. Steve Cooper, Nuno Espirito Santo, Ange Postecoglou, Sean Dyche, Vitor Pereira & Oliver Glasner. Let’s see what the model rates as some of the best match-day decisions made by these men since the beginning of the 2022-23 season.
The Best Match Day Decisions
Nuno Espirito Santo | 14 Sep 2024 - Liverpool (A) | Callum Hudson-Odoi for Nicolas Dominguez | 0-0 at 53:03 | 1-0 at FT
This is the substitution I thought would be the winner in these rankings. It’s one of the victories that comes to mind when you think about how the coach has clearly tilted the game in Forest’s favour. 4 games into this season and Nuno took the decision to leave out Hudson-Odoi & Anthony Elanga to really restrict Liverpool.
This gamble paid off in a huge way as he gradually introduced these players, who memorably combined for the winning goal later in the second half.
In the entirety of the game leading up to this substitution, Forest hadn’t taken a single shot - not one attempt at goal. Liverpool had taken 8 shots worth 0.73 xG - the data suggested Nuno had to make a change.
Oliver Glasner | 12 Sep 2026 - Aston Villa (A) | Igor Jesus on for Dan Ndoye | 1-1 at 74:08 | 2-1 at FT
I was a little surprised by this decision being so high, but it does make sense considering this was a late substitution that resulted in the player coming on scoring the decisive goal in a tough away game.
At the time of this substitution, the data suggested Forest’s expected points was 1.02 - the flow of the game and chances created for either side after this increased this figure to 1.98. Forest went on to win this game and it was Igor Jesus who scored the goal.
Forest were clearly in the ascendancy before Glasner made this sub. The model rewards managers for being ‘brave’, and making changes even when the underlying data suggests things are going well - and that’s exactly what Glasner did at Villa Park.
Considering that Aston Villa are deemed a “stronger team” too, and that it was away from home, this decision is rated very highly. Forest were of course a little fortunate in this game as Ian Maatsen did leave the field injured, leaving Villa down to 10 men as Forest pushed hard for the win.
Steve Cooper | 14 Jan 2023 - Leicester (H) | Sam Surridge on for Gustavo Scarpa & switch to 4-3-3 | 0-0 at 54:15 | 2-0 at FT
I must admit I didn’t really think of this game as one that would have a decision rank so highly - but when you think about it, this substitution really swung the game into Forest’s favour. Within a minute of coming on, Surridge was involved in the move that helped Gibbs-White release Brennan Johnson for the opening goal.
In addition, with 6 minutes left on the clock Surridge was involved again in the build up as the same two players combined to complete such an important victory.
After Cooper made this sub and switched the formation from a 4-3-1-2 to a 4-3-3, Leicester didn’t have a single attempt at goal. This was a very good example of how proactivity from the touchline can really help teams - Brendan Rodgers’ Leicester side played 4-2-3-1 throughout, even bringing on players out of position.
Oliver Glasner | 07 Nov 2026 - Man City (H) | Luca Netz & Dan Ndoye on for Daniel Muñoz & James McAtee | 2-1 at 64:15 | 3-1 at FT




