This project is my attempt to recreate *Xabi Alonso's Chelsea 3-4-2-1 in Football Manager 26*.

The objective wasn't simply to build the strongest possible 3-4-2-1 inside the FM26 match engine.

As always with these projects, I wanted to start from the football idea, translate the behaviours into Football Manager, and then see how far the system could go across an entire season.

And this one produced a very interesting result.

🏆 Season Results

🏆 *Premier League Champions*

🏆 *UEFA Champions League Winners*

❌ *FA Cup Semi Final*

❌ *Carabao Cup Third Round*

Premier League

*81 points*

25 Wins

6 Draws

7 Losses

*90 Goals Scored*

*39 Goals Conceded*

*+51 Goal Difference*

Underlying League Numbers

*88.08 xG*

*35.18 xGA*

*78.76 xPTS*

The league wasn't completely dominant.

In fact, Chelsea finished level with Tottenham on *81 points*, and the final part of the season became much more complicated than the first part.

There were rotations, some inconsistent performances and several moments where the system didn't simply destroy the opposition.

And I actually think that is an important part of this test.

🏆 Champions League Final

The season ended with a *2-1 win against Real Madrid after extra time* in the Champions League Final.

Real Madrid 1-2 Chelsea

⚽ Vinícius Júnior 32'

⚽ Cole Palmer 63'

⚽ Trevoh Chalobah 105'

The final itself was also a good example of the kind of football this tactic can produce.

Chelsea finished with:

*22 shots*

*13 shots on target*

*2.75 xG*

while Real Madrid produced:

*25 shots*

*9 shots on target*

*2.67 xG*

It wasn't a match where we simply controlled everything through possession.

It was competitive, aggressive and relatively balanced.

But we created enough to win it.

📊 Full Season Record

Across all competitions:

*59 Matches*

✅ *40 Wins*

🤝 *8 Draws*

❌ *11 Losses*

⚽ *135 Goals Scored*

🛡️ *59 Goals Conceded*

📈 *+76 Goal Difference*

🔥 *67% Win Rate*

And we finished the season with the two trophies that mattered most:

*Premier League + Champions League.*

⚽ The Tactical Structure

The base structure is a *3-4-2-1*.

In possession we have:

three central defenders

two aggressive wing-backs

a central midfield pair

two attacking midfielders operating behind the striker

and a central forward occupying the highest line.

The idea isn't simply to have eleven players sitting inside a static formation.

The roles are selected to generate different behaviours depending on the phase.

The wing-backs are extremely important because they provide the width while the two attacking midfielders can operate much closer to the central areas.

In midfield, the pairing is also deliberately different.

One midfielder has more responsibility in progressing and connecting possession, while the other provides a more stable reference.

Ahead of them, the two attacking midfielders aren't completely symmetrical either.

That creates different passing angles and movements around the striker instead of simply having two identical players occupying the same spaces.

🔄 In Possession / Out of Possession

One thing I continue to explore heavily in FM26 is the separation between *In Possession and Out of Possession roles*.

I don't want to think only in terms of:

"This player is a CM."

or:

"This player is an AM."

What matters more is what I want that player to do during the different phases.

That is why you can see several role changes between IP and OOP.

The midfielders, wing-backs and defenders don't necessarily perform exactly the same job when the team loses the ball.

For me, this is one of the most interesting tools FM26 gives us when trying to recreate real tactical behaviours.

🧠 Control Possession, But Not Possession for the Sake of Possession

The tactic uses a *Control Possession base with a Positive mentality*.

But the objective was never to chase a possession percentage just because a recreation is associated with positional football.

Chelsea averaged around *51% possession in the Premier League*.

That placed us around the middle of the league possession rankings.

And I'm perfectly fine with that.

Possession alone doesn't tell me whether the structure is working.

I care much more about whether the team can progress the ball, create chances, occupy the correct spaces and remain structurally competitive when possession is lost.

The final league numbers were:

*90 goals*

*88.08 xG*

*35.18 xGA*

Those numbers tell me much more about how the tactic functioned across the season.

📉 About the Numbers, Imperfection and Recreations

I also want to make something very clear.

*I don't think these numbers are exceptional.*

And I don't think this tactic is perfect.

There is sometimes a strange expectation around Football Manager tactics where a system is considered bad if it doesn't produce 100+ Premier League points or completely absurd attacking numbers.

I don't agree with that.

Could this tactic probably score more goals?

Yes.

Could I change certain roles or instructions and make the system more efficient inside the FM26 match engine?

Probably.

But at that point I also have to ask myself what I'm actually trying to create.

This is a *recreation*.

And recreations have limits.

There were very good periods during this season, but there were also ups and downs.

The final part of the Premier League season was particularly difficult.

I had to rotate the squad, performances became less consistent and we dropped points that a heavily optimised meta tactic might potentially avoid.

That's fine.

We still won the *Premier League*.

And we still won the *Champions League*.

For me, imperfection isn't automatically a problem when the tactical idea is working.

Recreation vs Adaptation vs Exploiting the Engine

This is something I think is extremely important when talking about Football Manager recreations.

*Every recreation requires adaptation.*

Football Manager is not real football.

The roles aren't real football roles.

The instructions aren't real football instructions.

And the match engine doesn't reproduce every movement or interaction we see on a real pitch.

So when I recreate a real tactical system, I obviously have to translate those ideas using the tools Football Manager gives me.

But:

*adaptation does not mean exploiting the engine.*

Those are two completely different things.

I can adapt a real tactical idea so that it functions inside FM26 without deliberately replacing every behaviour with whatever happens to be strongest in the current match engine.

If I discover that another role produces more goals, that doesn't automatically mean I should use it.

If another structure gives me another ten points, that doesn't automatically make it a better recreation.

There is always a balance.

Recreation

Trying to reproduce the tactical principles and behaviours.

Adaptation

Translating those principles into something Football Manager can actually reproduce.

Meta optimisation

Changing the system primarily because certain roles, instructions or structures perform better inside the match engine.

There is nothing wrong with meta tactics.

I make and test those too.

But they are different projects.

With this tactic I wanted to remain on the recreation side of that line.

Why 81 Points Doesn't Bother Me

Could I look at *81 Premier League points* and say the tactic isn't good enough?

Of course.

But then I would be ignoring the context.

This tactic finished the season as:

🏆 *Premier League Champion*

🏆 *Champions League Champion*

while still remaining relatively close to the tactical structure I wanted to reproduce.

For me, that matters.

A recreation doesn't need to produce 110 points to prove that it works.

Sometimes a tactic being slightly imperfect actually makes the test more interesting.

You see where the structure works.

You see where Football Manager creates problems.

You see which ideas translate naturally into the engine and which ones require compromises.

That information is much more useful to me than simply forcing every recreation toward the same meta behaviours.

🔍 What I Would Improve

I don't consider this the final possible version of the system.

There are still areas I could experiment with.

The midfield balance could potentially be adjusted.

The interaction between the two attacking midfielders and the striker could probably be pushed further.

The wing-back behaviour could also be modified depending on the players available.

And some of the instability we experienced during the final part of the season suggests there is still room to improve the defensive transition.

But I don't want to change things simply because they would improve the numbers.

Any future modification would still have to make sense inside the tactical idea.

That's the important part.

Final Result

🏆 *Premier League Champions*

🏆 *Champions League Winners*

📈 *81 Premier League Points*

⚽ *90 League Goals*

🛡️ *39 League Goals Conceded*

📊 *88.08 xG*

🛡️ *35.18 xGA*

🎮 *59 Matches*

✅ *40 Wins*

⚽ *135 Total Goals*

🔥 *67% Win Rate*

Is it a perfect tactic?

No.

Could it probably be pushed further inside the FM26 match engine?

Yes.

But that wasn't the only objective.

The goal was to recreate the principles of the system, adapt them to Football Manager and see whether the structure could actually survive an entire season.

It did.

The season had highs and lows.

The final part of the Premier League became difficult.

The numbers aren't the absurd numbers you sometimes see from heavily optimised meta systems.

And that's okay.

We still finished the season as *Premier League and Champions League winners*.

For me, that's also part of doing recreations.

*Not every successful tactic needs 100+ points.*

Sometimes an imperfect recreation that maintains its tactical identity and still wins major trophies tells you much more than another tactic designed purely around what the match engine does best.

# 🎮 Console Players

A *Team Instructions / Player Instructions video is included* so console players can manually recreate the tactic without needing to import the tactic file.

More experiments will follow.

I'm still exploring how far these FM26 In Possession / Out of Possession mechanics can be pushed when recreating real tactical systems, and there are already several ideas from this structure that I want to test in future projects.