System specs:
- Windows 11 Pro 64-bit (Build 26200)
- Gigabyte X470 AORUS Ultra Gaming
- AMD Ryzen 9 5950X (32 threads)
- NVIDIA RTX 5060 Ti, driver 32.0.16.1074
- 48GB RAM
- Monitor: 3440x1440 ultrawide, 165Hz
Issue:
The game launches, EasyAntiCheat initializes, shaders load, but the game crashes and returns to Steam right before reaching the main menu. No error code or message is displayed. 100% reproducible on every launch attempt.
Troubleshooting already completed (no change in behavior):
- Verified game file integrity through Steam
- Fully reinstalled EasyAntiCheat via EAAntiCheat.Installer.exe
- Disabled Core Isolation / Memory Integrity in Windows Security
- Full clean reset: uninstalled the game, deleted %localappdata%\EA SPORTS FC 26, %localappdata%\EADesktop, %appdata%\EA, %appdata%\Electronic Arts, then reinstalled
- Launched EA App in Offline Mode
- Disabled all overlays (Steam overlay, Discord overlay, NVIDIA overlay)
- Ran FC26.exe directly as Administrator
- Updated to latest NVIDIA drivers
- Tested with reduced resolution/refresh rate (1920x1080 @ 60Hz)
- Added -dx11 launch option in Steam
- Checked Windows Event Viewer (System log) — no error-level entries at time of crash, no Kernel-Power ID 41, no BSOD
- No crash dumps generated in C:\ProgramData\eaanticheat\Crashdumps (folder empty)
- No crash dumps found in %localappdata%\EA SPORTS FC 26\twinkle either
- Full DxDiag run — no problems detected on GPU, drivers, or DirectX components
- Followed the official fix posted by EA_Kuba in the forum thread "Anticheat" (same exact symptom as mine): deleted EAAntiCheat.GameServicesLauncher.dll, EAAntiCheat.GameServicesLauncher.exe, and preloader_l.dll, repaired the game to regenerate them, then set all three files to read-only. This did not resolve the issue.
Since the crash happens consistently right before the main menu with zero logged errors anywhere (OS, EAC, or game logs), and the documented official fix for this exact symptom didn't work either, this may require deeper investigation on EA's side — possibly related to the account authentication/handshake phase rather than a local configuration issue.