Fourth iteration. See v1, v2 and v3 for the save and test conditions, identical throughout.
CHANGES FROM V3 In OOP: the 4-4-2 becomes a 4-3-3:
Midfield: two DMs -> three, a Dropping DM flanked by two Pressing DMs Front: two AM and two Wide Midfielders -> one AM and two Wingers, back to the v1 front three
WHAT I SAW Possession: 52% -> 54% PPDA: 20.0 -> 19.6 Shot accuracy: 45.9% -> 49.1% npxG/90: 2.1 -> 2.0 xGA per game: 1.0 -> 1.0 Season: 71 points and 80/36, against 81 points and 78/30 for v3
The pressing numbers moved the way I wanted and the underlying quality barely moved at all. The ten point drop sits on top of an identical xGA, so I have no idea how much of it is real. That is what I would like the test to settle.
Third iteration. See the v1 & v2 for the full context on the save and the test conditions. They are identical across all three.
The idea here is simple: take what worked in v1 and what worked in v2 and put them together. I am trying to build the best attacking 4-3-3 I can from a footballing point of view, short passing and possession, with the most effective pressing shape I can find behind it. The aim is a short passing tactic that is genuinely competitive on points per season, built from what all the testing on this forum has taught me about the engine.
Only change from v2: Attacking Width back to Much Narrower. Everything else from v2 is unchanged.
Follow-up to my first post, tested at 73.5 PTS, +32 GD, xG diff +26.9.
Same conditions as v1: OGC Nice, fresh save, original database, holiday, no transfers, training and opposition instructions delegated, same set piece routines.
CHANGES
Out of possession, the 5-2-3 becomes a 4-4-2: Keeper: Goalkeeper -> Sweeper Keeper Back line: five with the DM dropping in -> four with the DM staying in midfield Front: one AM -> two AM, a channel midfielder pushing up beside the striker Wide: Wingers -> Wide Midfielders Defensive line: Much Lower -> Higher
In possession: Attacking Width: Much Narrower -> Wider Striker: Deep Lying Forward -> Channel Forward, with Come Deeper and Fewer Forward Runs
WHY
Two things in v1 I could never justify: a high press sitting on top of a Much Lower defensive line, and Much Narrower width in a shape whose inside forwards are told to hold the touchline. Attack narrow, then tell your widest players to stay wide. That is just fighting yourself.
MY READ
The width change works because of the player instructions, not in spite of them. On Much Narrower, telling the forwards to stay wider barely pulled them off the centre. On Wider they actually reach the touchline, the wing backs inherit the half spaces, and the two of them put 37 assists between them.
The 23 extra goals conceded are another matter. xGA went from 1.0 to 1.1 per game, so roughly 34 expected against in v1 and 37 in v2, against 25 and 48 actually conceded. Same quality of chance given up, a 20 goal swing in what went in. On 34 games I would call that variance rather than tactics, but I would rather be told than guess.
What I changed from meta-tactics was what I needed to change to make it look like the football I want to watch:
- shorter passing instead of the More Direct - the striker as a deep lying forward, dropping into midfield rather than pinning the last line - the midfield set up to press rather than sit & roam from positions - individual instructions reworked to dribble more & more pressing
I'd like to put a tactic forward for testing. Quick warning up front: it's not another strikerless low block counter. It's pretty much the opposite, and that's kind of the whole point.
1- What I was chasing
I've been obsessed with what Luis Enrique has been doing at PSG the last couple of seasons. Not the highlight reel stuff, the structure. The way the shape turns into a 3-2-5 the second they have the ball, how the wing backs go and hold the touchline so the wide forwards can drift into the pockets instead of being stranded out there one on one, the way there are always two bodies sitting in front of the back three so the counter press starts the instant the ball goes. It's positional play, but it's aggressive. It doesn't pass sideways waiting for a gap to appear, it passes fast and moves faster.
So that was the target. But I also didn't want to be naive about the patch. 26.3 rewards certain things, and pretending otherwise just means losing 3-1 every week with 65% of the ball. So the build is honestly a compromise. The shape and the ideas come from watching modern football, the dials come from paying attention to what actually works in this match engine. Narrow attacking width, much higher tempo, counter attack transition, work the ball into the box, discourage long shots. Those are meta calls and I'm not pretending they aren't. But they're bolted onto a possession structure with a real number nine instead of a deep block with nobody up front, and that combination is what I'd love to see tested properly.
The one thing I still can't fully explain is the defensive setup. High press line of engagement sitting on top of a much lower defensive line, with step up more on top of that. It reads like nonsense on paper. It gave me 16 clean sheets and 0.7 conceded a game. I'd genuinely like to know whether that's the tactic or whether I got lucky for 34 games.
2- The save, fully transparent
OGC Nice, first season, fresh save on 26.3. No signings, nobody sold, nobody loaned out. Squad exactly as the game handed it to me. Training, individual training and opposition instructions all delegated to the assistant, so none of that is me micromanaging in the background. Set pieces are short corners and free kicks kept on the ground, two man routines, which produced 13 league goals. Normal career save, no custom test league. The only things I touched were the tactic and the starting eleven.
The bit I'd point at: only PSG finished above us (91 points), and everyone else was 13 points behind. Second best attack in the league (79, with the third best on 60) and second best defence (25, with the third best on 36). So it wasn't a lucky second place scraped on goal difference, the gap to the rest was pretty comfortable.
Other numbers: 16 clean sheets, 0.7 conceded and 1.0 xGA per game, 2.3 goals per game, 2.1 npxG/90, 18.6 shots per game at 45.7% accuracy, 89.2% pass completion, 53% average possession (only 5th in the league, which honestly surprised me). Most shots in the league at 633 and most tackles won at 515, which tells you the press was doing its job. We also went out in the Champions League round of 16 to Dortmund on aggregate.
3- The shape
In possession: ball playing keeper, two ball playing centre backs, two advanced wing backs pushing on either side of a DM who drops in, two channel midfielders in the half spaces, two inside forwards either side of a deep lying forward. Becomes a 3-2-5 with every lane occupied.
Out of possession: back four, two DMs, wingers tracking back, one man left higher. Everything folds in and gets compact.
CHANGES FROM V3
In OOP: the 4-4-2 becomes a 4-3-3:
Midfield: two DMs -> three, a Dropping DM flanked by two Pressing DMs
Front: two AM and two Wide Midfielders -> one AM and two Wingers, back to the v1 front three
WHAT I SAW
Possession: 52% -> 54%
PPDA: 20.0 -> 19.6
Shot accuracy: 45.9% -> 49.1%
npxG/90: 2.1 -> 2.0
xGA per game: 1.0 -> 1.0
Season: 71 points and 80/36, against 81 points and 78/30 for v3
The pressing numbers moved the way I wanted and the underlying quality barely moved at all. The ten point drop sits on top of an identical xGA, so I have no idea how much of it is real. That is what I would like the test to settle.
Thanks for reading.
Third iteration. See the v1 & v2 for the full context on the save and the test conditions. They are identical across all three.
The idea here is simple: take what worked in v1 and what worked in v2 and put them together. I am trying to build the best attacking 4-3-3 I can from a footballing point of view, short passing and possession, with the most effective pressing shape I can find behind it. The aim is a short passing tactic that is genuinely competitive on points per season, built from what all the testing on this forum has taught me about the engine.
Only change from v2: Attacking Width back to Much Narrower.
Everything else from v2 is unchanged.
Follow-up to my first post, tested at 73.5 PTS, +32 GD, xG diff +26.9.
Same conditions as v1: OGC Nice, fresh save, original database, holiday, no transfers, training and opposition instructions delegated, same set piece routines.
CHANGES
Out of possession, the 5-2-3 becomes a 4-4-2:
Keeper: Goalkeeper -> Sweeper Keeper
Back line: five with the DM dropping in -> four with the DM staying in midfield
Front: one AM -> two AM, a channel midfielder pushing up beside the striker
Wide: Wingers -> Wide Midfielders
Defensive line: Much Lower -> Higher
In possession:
Attacking Width: Much Narrower -> Wider
Striker: Deep Lying Forward -> Channel Forward, with Come Deeper and Fewer Forward Runs
WHY
Two things in v1 I could never justify: a high press sitting on top of a Much Lower defensive line, and Much Narrower width in a shape whose inside forwards are told to hold the touchline. Attack narrow, then tell your widest players to stay wide. That is just fighting yourself.
MY READ
The width change works because of the player instructions, not in spite of them. On Much Narrower, telling the forwards to stay wider barely pulled them off the centre. On Wider they actually reach the touchline, the wing backs inherit the half spaces, and the two of them put 37 assists between them.
The 23 extra goals conceded are another matter. xGA went from 1.0 to 1.1 per game, so roughly 34 expected against in v1 and 37 in v2, against 25 and 48 actually conceded. Same quality of chance given up, a 20 goal swing in what went in. On 34 games I would call that variance rather than tactics, but I would rather be told than guess.
Thanks for reading.
I want to watch:
- shorter passing instead of the More Direct
- the striker as a deep lying forward, dropping into midfield rather than
pinning the last line
- the midfield set up to press rather than sit & roam from positions
- individual instructions reworked to dribble more & more pressing
I'd like to put a tactic forward for testing. Quick warning up front: it's not another strikerless low block counter. It's pretty much the opposite, and that's kind of the whole point.
1- What I was chasing
I've been obsessed with what Luis Enrique has been doing at PSG the last couple of seasons. Not the highlight reel stuff, the structure. The way the shape turns into a 3-2-5 the second they have the ball, how the wing backs go and hold the touchline so the wide forwards can drift into the pockets instead of being stranded out there one on one, the way there are always two bodies sitting in front of the back three so the counter press starts the instant the ball goes. It's positional play, but it's aggressive. It doesn't pass sideways waiting for a gap to appear, it passes fast and moves faster.
So that was the target. But I also didn't want to be naive about the patch. 26.3 rewards certain things, and pretending otherwise just means losing 3-1 every week with 65% of the ball. So the build is honestly a compromise. The shape and the ideas come from watching modern football, the dials come from paying attention to what actually works in this match engine. Narrow attacking width, much higher tempo, counter attack transition, work the ball into the box, discourage long shots. Those are meta calls and I'm not pretending they aren't. But they're bolted onto a possession structure with a real number nine instead of a deep block with nobody up front, and that combination is what I'd love to see tested properly.
The one thing I still can't fully explain is the defensive setup. High press line of engagement sitting on top of a much lower defensive line, with step up more on top of that. It reads like nonsense on paper. It gave me 16 clean sheets and 0.7 conceded a game. I'd genuinely like to know whether that's the tactic or whether I got lucky for 34 games.
2- The save, fully transparent
OGC Nice, first season, fresh save on 26.3. No signings, nobody sold, nobody loaned out. Squad exactly as the game handed it to me. Training, individual training and opposition instructions all delegated to the assistant, so none of that is me micromanaging in the background. Set pieces are short corners and free kicks kept on the ground, two man routines, which produced 13 league goals. Normal career save, no custom test league. The only things I touched were the tactic and the starting eleven.
How it went (Ligue 1 2025/26)
2nd place, 78 points, 24 wins 6 draws 4 losses. 79 scored, 25 conceded, +54.
The bit I'd point at: only PSG finished above us (91 points), and everyone else was 13 points behind. Second best attack in the league (79, with the third best on 60) and second best defence (25, with the third best on 36). So it wasn't a lucky second place scraped on goal difference, the gap to the rest was pretty comfortable.
Other numbers: 16 clean sheets, 0.7 conceded and 1.0 xGA per game, 2.3 goals per game, 2.1 npxG/90, 18.6 shots per game at 45.7% accuracy, 89.2% pass completion, 53% average possession (only 5th in the league, which honestly surprised me). Most shots in the league at 633 and most tackles won at 515, which tells you the press was doing its job. We also went out in the Champions League round of 16 to Dortmund on aggregate.
3- The shape
In possession: ball playing keeper, two ball playing centre backs, two advanced wing backs pushing on either side of a DM who drops in, two channel midfielders in the half spaces, two inside forwards either side of a deep lying forward. Becomes a 3-2-5 with every lane occupied.
Out of possession: back four, two DMs, wingers tracking back, one man left higher. Everything folds in and gets compact.
Thanks for reading