False 9 Football - My little football sim project

[Madmac79]

Hey.

I've been working on the "ugly" side of things so don't have much to show in terms of new gameplay:

By ugly, I mean:
- I am converting a lot of visual coding (blueprints) to C++. I have had to do this for multiple reasons, blueprints are quicker to make but can only get you so far and they get messy, dont support modding very well and lots of things i cant do (cleanly) within them
- I am building the AI Layer in C++ with lots of modifiers depending on situations and tactics, its taking a lot of my time
- I have built a fake player generation system in C++
- I have built a career/tournament fixture generation and simulation system in C++ with working simulated league tables
- I have been building up the menu system
- Learnt a lot about modules within Unreal Engine so I can release "demos" without including big features like Career or Modding

What I am deeply concerned about is my ability to make animations or at least good ones. However, I get this comes with time and counts as polish. So jank is expected!

Love to see how you are focusing on these surely boring things, that probably won't steal the eye on words, but are indeed the very heart of every football game!

Keep on with your vision! Always waiting patiently, but eagerly for your updates

 

[Paul-v]

Having just gone through this, my recommendation for the AI is a finite state machine with several layers, a utility (points) evaluator when decisions are not binary, as well as an urgency evaluator (walk, jog or sprint), and a contextual risk tolerance modifier (depending on score, time, field position and perhaps tactics). While it sounds complicated, you only need to make one decision evaluation system as it applies to all players.

I have urgency and a points evaluator now created without any real dependency on the app / editor being open, using the Unreal Testing Framework. It's pretty cool because I can create scenarios and suggest what the outcome should be. This should save me from breaking AI with changes in the future.


I am now moving on to the Decision Layer for Attacking decisions.

This can be a little tedious but I need to lay a good framework before going back to the pretty stuff.

 

[sleevel]

An automated test harness is an excellent way to tune and assure performance of your AI. You can recreate any of the FIFA minigames and put your AI through it to make sure it’s successful.

In your defensive decision making, where does does revert to formation position factor in? Unless this is part of MarkOpponent, it would appear to be missing. I would also discourage you from making defense and offense so binary. Even if the other team has the ball, I might still be putting myself in an advantageous position in case of a turnover. That’s how high pressing counterattacks are really successful.

I would suggest making your player states public and available to fellow teammates. For example, if two defenders switch which player they’re marking, if all players move up together in unison, or if I can aggressively go for the ball because my teammate behind me is covering. This does mimic reality, as a well drilled team knows what the other player is going to do in different situations.

Enjoy the programming - it’s worth it.

 

[NeoRetro]

@Paul-v don't hesit to post stuffs like you've answered to @PES4eva : that's exactly what we, well i think most of us love to read.
Those technical part you're workin' on are really on top and very detailled, and even if it's perphas "too technical" some would say, that's perhaps most of us love the most even than some vids... Well vids even in pre-alpha-alpha / or squeletton let's say, of world football building is something we discover and we could give you some small advice from what is important, what's less, etc. etc.
But stay on your own vision to what you want, don't lost focus on all ideas we gives you. Still our game, our vison of what a FG should be to be good.

Continue you job as long as you got fun to do it like if you're writing a book : with enjoyement and without any sort of pressure.
I've been impressed by the guys who was trying to create the "gameplay football" i tried as a pre-alpha release, looking a lot regarding the flow and basis + freedom on the pitch, but your one man army is on another level.
I personnally don't care if you don't reach it or not, or making it simply as en engine basis not playable, but it's super interesting to read the advancement of what you did. The latest list of what you were working on popped my eyes on i was like "he's already working on THIS??"

Congrats, courage, our hearts is with you from what you're trying to achieve : no matter what it will be : to us it's like reading a book of "how to developp a football game" and you know we're passionned and details are the things we really all care. Repeat mode: don't hesitate to share : it's like visiting a museum with someone explaining us the story of all type of arts we have no clue explaining to us the story behind any sort of statue, painting or whatever else (as we're don't have your programmer skills)
Just thanks, it makes us dreamin' a bit in those sad time without a proper builded football game, while we're, for some of us, playing real old games where you could get and most of all, "feel" a casual mostly realistic football as we should have in 2026.

 

[Sany023]

Mate Paul-v, and may be later, during some time of development, while polishing code, searching wrong and true decisions for small and big tasks you are too make a game, that will be suitable for adding into PlayStation base of games in their PSN magazine as other our evoweb user did it with his cool game Arise ?

P.S. Very interesting read about this project and some news about development issues and finding solutions ...

 

[Gapster22]

I'm a lurk type but after reading OP and sleevel's thoughts and ideas it motivated me to write my own thoughs I've been having about such a game.

About the quick strat select:
Pes has something very basic d-pad options that the user can switch between mid-game (including the corner formation stuff). My idea was to expand it further, by having a system with specific strats with each having a counter-strat in practice and to have them defined in a way for people to easily add new ones.
On top of it I was also thinking of having a system where each individual manager would have their preferred go-tos (for certain strats). The managers would have a rating between 1 and 5 in terms of capability and have their own brain that would "adapt" to your playstyle (within on their capability and with varying degrees of likelihood of change and frequency while also considering their roster's capabilities). This would be true for both formation related stuff and on-the-fly strats (like when to double pressure, covering etc.).
What this all accomplishes is it makes the game more akin to real games where you can (on higher difficulties and coach lvls) have a bit of a rock paper scissors dynamic.
I was even thinking of having better coaches be "aware" of their players' stats and to be able to prioritize strats they have knowledge of (basic default list + coach fav ones) depending on the degree of ability players in their roster have for employing them (all of this would be a feature in case personalities and other flexible/trainable player-related factors were added as a system)
That said having looked a bit into a very good AI brain mod for SPT made me realise the insanity this would be to execute in a football game, let alone make it feel seamless and not eat away performance.

Other match mechanics I was thinking of were random supporter and player events (flares, automatic and manual player aggressive reactions, non-aggressive player-refree interactions...) and similar but making it somewhat modular for modding so people could expand it further.
One other idea was about players on the pitch having a hitbox with each limb having its own hp in order to have a bit better injury system than pes.

Playing controls wise (and this will sound a bit crazy) I was thinking of an advanced fuma where besides the default you could manually tap/hold the L2/R2 button to force using left/right leg and have the right analog for determining angle of touch/kick of the ball, type of kick (normal center, swerve, knuckle, lob etc. to depend on both power as usual as well as the ball surface you hit with the foot, kind of like in set pieces in some games). As for ball control, pushing it, feints and so on my idea revolved around employing double tap and flick controls with analog and sometimes L2/R2. I was considering also having L2+R2 specifically for aerial ball control, stopping, heading. Having the manual kick angle thing would also make so low and high passes including through ball could easily only use 1 and the same button (+leg one optional). this would free up to 2 buttons for other functions like a priority control (super cancel), calling a teammate near to support (this one could be on d-pad instead), goalkeeper stuff etc.
If pes taught me something it's that having a super cancel is very important, especially in co-op. It allows for more accurate and faster reactions, positioning, repositioning, which can help with covering, blocking other players, more accurate marking, can help with gk when assuming control mid-defense to off-time the opponent's momentum, can help shake off pressure and markings while attacking, can be used to bait and complement feint dribbles and so on.

The rest (off the pitch stuff, career mode etc.) I don't think could ever be anywhere nearly as much of a nightmare as the on the pitch stuff. Especially considering ball physics being decided by very manual alterations at the start of its momentum when kicking.

Anyway thanks for listening to my ted talk. No but seriously whoever reads all of this ludacris stuff, props to you, hopefully I didn't waste your time. I swear I wasn't trying of turning football into crusader kings or something...I think.
With that said, huge props to anyone attempting developing a football game and I wish you all the luck.

 

[Paul-v]

Anyway thanks for listening to my ted talk.

I always appreciate a good talk with people on their ideas on things. So its much appreciated. Keep it coming!

That said having looked a bit into a very good AI brain mod for SPT made me realise the insanity this would be to execute in a football game, let alone make it feel seamless and not eat away performance.

This is very off topic but I WAS heavily involved with SPT for a few years. I built SIT which was the precursor to the existence of what is now Fika. People hate my guts over there though, i gave up with that drama lol

Mate Paul-v, and may be later, during some time of development, while polishing code, searching wrong and true decisions for small and big tasks you are too make a game, that will be suitable for adding into PlayStation base of games in their PSN magazine as other our evoweb user did it with his cool game Arise ?

P.S. Very interesting read about this project and some news about development issues and finding solutions ...

Thats a cool project, ill check it out Thanks for the tips.

In your defensive decision making, where does does revert to formation position factor in? Unless this is part of MarkOpponent, it would appear to be missing. I would also discourage you from making defense and offense so binary. Even if the other team has the ball, I might still be putting myself in an advantageous position in case of a turnover. That’s how high pressing counterattacks are really successful.

There are 3 ball possession states, Attacking, Defending and Transition. The revert to formation is the fallback if no other options are viable and cannot get past the defined threshold.