Simulation Flow Unleashed: A Reworked Simulation Lag Fix

🧪 Version 2.2 Beta is now available

Version 2.2 Beta keeps the 33/50/33 scheduling profile and autonomy tuning used by the previous Beta. This update focuses on safer startup and clearer error reporting rather than changing normal gameplay performance.

SFU now confirms that loading has finished for the current lot before it begins advancing its additional autonomy timeline. During the initial startup period, an additional autonomy pass is allowed only after the main simulation update completes in under 21 ms. After six stable checks, SFU returns to the normal threshold selected in the CFG.

This may make Sims briefly feel less responsive immediately after entering a lot than in an older version. That short delay is intentional: it reduces the risk of additional autonomy work beginning while the lot and its gameplay systems are still being initialized.

The previous Stable version remains available. Stable and Beta versions must not be installed together.

If the Beta feels too active on your computer, open its CFG with a text editor such as Notepad and change:

update_threshold_ms = 40

This more conservative setting may improve visual or menu smoothness on some systems, although Sims may sometimes take slightly longer to begin autonomous actions.

👉 Feedback about responsiveness, stuttering, loading behavior, or unusual interactions is very welcome.

🎯 WHAT THIS MOD TRIES TO IMPROVE
When I play The Sims 4, I often notice situations such as:
‧ A Sim finishes one action but remains in place before starting the next one.
‧ Several Sims repeatedly compete for the same object, fail, and then try again.
‧ Food, trash, or other objects are left behind before the related action is completed.
‧ Sims become less responsive in crowded situations.

I cannot say that all of these situations have the same cause. However, they made me pay more attention to how the game handles autonomy and how mods such as Simulation Lag Fix attempt to address simulation lag.

Simulation Flow Unleashed is a version I created after studying and modifying Simulation Lag Fix. Based on my own experience playing and testing the game, I revisited some of its timing behavior and added several adjustable or optional tuning changes.

My goal is simple: to see whether these changes can reduce situations in which a Sim is slow to start the next action, autonomous actions repeatedly fail in crowded situations, or objects are left behind before an action is complete.

This mod is not intended to replace original versions of Simulation Lag Fix. It is another adjustment direction for players who may prefer a different balance in their own saves and hardware setups.

⚙️ WHY I MADE THIS REWORK
The July 21, 2026 update expanded Memory Boost to all platforms. According to EA, Memory Boost is intended to improve memory management, reduce lag, improve frame rates and responsiveness, and reduce memory-related problems during longer play sessions.

I had already started using Memory Boost during its earlier access period. I noticed that loading felt faster and that the game seemed less prone to stuttering.

However, although I noticed some improvement in Sim autonomy, I did not feel that it fully addressed the problems I was observing. I therefore continued using community mods commonly used for this purpose, including Simulation Lag Fix.

While using and studying Simulation Lag Fix, I noticed that some of its settings reflected trade-offs made when hardware and engine resources were more limited. This led me to reconsider whether some of its more conservative limits were still the most suitable approach for every player.

Simulation Flow Unleashed began as an attempt to explore that question. It keeps the central idea of processing autonomy through a separate timeline, while revisiting the time-slice handling, the condition that allows the additional autonomy pass to run, and several related autonomy tuning values.

📂 WHICH VERSION SHOULD I USE?

Stable version
The stable version is the previous public release. It uses the established SFU scheduling profile and includes a separate, optional tuning package.
Choose this version if you prefer the previous behavior or want to continue using the optional autonomy tuning changes described below.

Version 2.2 Beta
The 2.2 Beta uses:
- the newer 33/50/33 scheduling profile;
- independently adjustable main, threshold, and autonomy values;
- a CFG that reloads when a game session or lot is loaded;
- autonomy tuning retained from the previous Beta;
- a current-lot loading gate and safer startup behavior;
- expanded compatibility and runtime reporting.

Choose it if you prefer the Beta’s more responsive scheduling approach and are comfortable testing its newer safety architecture.

Only install one version of Simulation Flow Unleashed at a time. Do not install either version together with another Simulation Lag Fix.

⚖️ HOW THE STABLE VERSION DIFFERS

The stable version keeps the basic architecture of the original Simulation Lag Fix approach:

- The normal simulation continues to run as usual.
- Autonomy is processed through a separate autonomy timeline.
- An additional autonomy pass may run after the normal simulation update.

The main differences from the commonly maintained Simulation Lag Fix are:

- The threshold used to allow the additional autonomy pass.
- The handling of the normal simulation time slice.
- Optional autonomy scoring, delay, lockout, weighting, and randomization adjustments supplied through a separate package.
- A configuration file that allows players to adjust the autonomy-pass threshold.

This is a trade-off, not a universal performance improvement. When the additional autonomy pass runs, it performs additional CPU work. If your computer has limited processing capacity, or if the game is already under heavy simulation load, this may contribute to occasional stuttering.

However, this trade-off is not unique to Simulation Flow Unleashed. It is part of the general approach of allowing additional autonomy processing.

🔧 STABLE VERSION: MAIN CHANGES

1. No forced 17ms time-slice cap

The original Simulation Lag Fix used the following

time_service.TimeService.MAX_TIME_SLICE_MILLISECONDS = 17

This forced the game's normal simulation time-slice limit to 17ms. Such a limit can help control the duration of a single update and reduce the chance that the simulation process will affect visual updates by running for too long. However, it may also divide some simulation and autonomy-related work into very short portions. The stable version of Simulation Flow Unleashed does not forcibly override this setting. In my test environment, the game's original time-slice limit is 33ms.

After using this setting for an extended period, I observed fewer situations that may be related to very short time slices, such as:
- Sims leaving before an action was fully completed.
- Objects being left on tables or other surfaces.
- Objects failing to complete follow-up processing.

These observations do not prove that every such problem is caused by the time-slice limit. The Sims 4 simulation includes many interacting systems. This change simply gives the normal timeline more opportunity to complete work before yielding.

2. Adjustable autonomy-pass threshold
The original code was designed to check whether the normal simulation update finished within approximately 17ms before allowing the additional autonomy timeline to run:

max_time_ms = self.MAX_TIME_SLICE_MILLISECONDS

if (time4 - time3) < 0.017:
    time1 = time.monotonic()
    result = self.autonomy_timeline.simulate(
        services.game_clock_service().now(),
        max_time_ms=max_time_ms
    )
    time2 = time.monotonic()

The stable version makes this threshold adjustable through its configuration file. Its default value is 21ms.
A lower threshold is more conservative because the additional autonomy timeline is allowed to run only after a relatively short normal update. A higher threshold allows the autonomy timeline to run after a wider range of normal updates.

This value is not the time limit of the autonomy timeline itself. It determines whether the autonomy timeline is allowed to run after the normal simulation update has finished.

🔧 VERSION 2.2 BETA: MAIN CHANGES

1. Separate main and autonomy time slices
The main simulation timeline and the additional autonomy timeline now use separate time-slice values.

The current Beta profile is:
• Main time slice: 33ms
• Update threshold: 50ms
• Autonomy time slice: 33ms

The main time slice controls the normal simulation timeline. The update threshold determines whether an additional autonomy pass may begin after the main update. The autonomy time slice is the target budget used when advancing the separate autonomy timeline.

The 50 ms threshold does not give the autonomy timeline 50 ms of processing time. These values are also cooperative scheduling budgets rather than hard CPU limits: an internal game function that cannot yield during its work may occasionally take longer than the configured value.

In my testing, 33/50/33 with the included autonomy tuning generally produced more successful autonomy results and gave Sims more frequent opportunities to begin their next actions.

2. Reloadable CFG settings
The three values can be adjusted through:

[AP]SimulationFlowUnleashed_v2.2_beta.cfg

The Beta reads the CFG again whenever a new game session or lot is loaded. You may return to the main menu, edit the CFG, and reload your save without completely restarting the game.

The .cfg and .ts4script files are both required and must remain in the same folder.

3. Safer startup after loading
The 2.2 Beta confirms that the loading screen has finished for the **current lot** before it begins advancing the additional autonomy timeline.

During startup, SFU temporarily uses a conservative 21 ms rule. An additional autonomy pass is permitted only after the main simulation update completes below 21 ms. SFU accumulates six stable checks before switching to the normal CFG threshold; if the main simulation becomes busy during this period, the stability count is reset.

This is intended to reduce the risk of:

- additional autonomy work beginning while the lot is still initializing;
- sudden CPU workload increases during loading;
- autonomous interactions being scheduled before required gameplay systems are ready;
- timing-related errors being caused or exposed by premature scheduling.

The protection may cause a short period of reduced responsiveness immediately after entering a lot. Once startup stabilizes, the Beta returns to the normal **33/50/33** behavior or the values selected in the CFG.

 4. Expanded compatibility safeguard

Before installing its changes, SFU checks the required game methods and installs all four hooks as one group:

-AutonomyService._register
-TimeService.start
-TimeService.update
-Zone.on_loading_screen_animation_finished

If any required part cannot be used safely, SFU does not intentionally leave a partial installation behind. It remains disabled, and the game continues using its original simulation system.

This reduces the risk from game updates or incompatible script wrappers, but it cannot guarantee compatibility with every future patch or third-party script mod.

5. Clearer reports
SFU creates a report only when the player may need to take action.
SimulationFlowUnleashed_DISABLED.txt

SFU could not start safely and is **not active**. The game is using its original simulation system. Remove this SFU version and check for a compatible update. The compatibility failure does not modify your save.

SimulationFlowUnleashed_Config_Warning.txt
The CFG could not be read or contains an invalid value. SFU remains active using built-in defaults or the last valid settings. Correct or replace the CFG, return to the main menu, and reload the save.

SimulationFlowUnleashed_Runtime_Error.txt
An exception was observed while SFU was active. SFU appearing in the traceback does not automatically mean that SFU caused the error; it may have been advancing work belonging to the game, an interaction, or another mod.

When reporting the problem, please include the SFU report, the game's LastException file if one was created, and a short description of what the Sim was doing.

📂 INSTALLATION
1. Remove all files belonging to the SFU version you are replacing.
2. Place the selected version's `.ts4script` and CFG in the same Mods folder.
3. Keep both files installed; do not delete the CFG after setup.
4. Install only the tuning package intended for the selected version. The tuning package is optional and does not create the separate timeline by itself.
5. Make sure Script Mods are enabled.
6. Never install the Stable and Beta versions together.

⚠️ KNOWN LIMITATIONS

- SFU cannot eliminate every cause of simulation lag, idle Sims, failed interactions, routing problems, or unfinished object handling.
- Crowded social situations may still cause Sims to compete repeatedly for popular objects.
- Some third-party interactions contain their own target-selection or execution problems. SFU may change when the error becomes visible without being its underlying cause.
- Results vary depending on hardware, save contents, game settings, lot population, and installed mods.
- Every additional autonomy pass uses CPU time. A more responsive profile is not automatically the profile that performs the least work.
- On systems already under heavy simulation load, the more active Beta profile may contribute to occasional stuttering.
- If the main simulation regularly exceeds the configured threshold, the additional autonomy timeline will run less often.
- Very busy lots, many simultaneous events, or the end of high-speed mode may temporarily leave more work waiting in the main simulation schedule.

🔧 AUTONOMY TUNING CONFLICTS

1. Simulation Lag Fix
   Because this mod is based on Simulation Lag Fix, it cannot be installed together with another version of Simulation Lag Fix.
   Please keep only one version installed.

2. Autonomy tuning conflicts

The included autonomy tuning modifies autonomy.autonomy_modes and should not be used together with other mods that change the same tuning, including:

‧ Zero's Faster Autonomy and Less Idle
‧ tyjokr's Less Frequent Idles
‧ Coolspear's Quicker Autonomy and Simulation Lag Improvement

1. Other tested combinations
   In my testing, this mod worked alongside:
   ‧ Tyjokr's Queue Guard
   ‧ TURBODRIVER's Simulation Timeline Unclogger
   I recommend using both of these mods alongside Simulation Flow Unleashed. Because of the way The Sims 4's single-threaded simulation engine handles object competition and other simulation tasks, some problems are difficult to avoid completely. Queue Guard and Simulation Timeline Unclogger can therefore serve as important last lines of defense, helping to reduce problems that may still occur even when Simulation Flow Unleashed is installed.
   
   

🤝 ACKNOWLEDGEMENTS
This mod is based on the Simulation Lag Fix source code provided by Xerox.
The original Simulation Lag Fix concept was created by simmythesim. Xerox later continued its maintenance, and SrslySims has also contributed to maintaining the version widely used by the community today.

Special thanks to the original author for the original concept, as well as to the mod authors who continued maintaining it. I did not use the earliest version myself. It was through later maintained versions of Simulation Lag Fix that I discovered how useful this mechanism could be. That led me to study it further and create this reworked version.

Some autonomy weighting ideas were inspired by Zero's Faster Autonomy and Less Idle, and the FULL_AUTONOMY_STATISTIC_SCORE_VARIANCE tuning is based on Coolspear's Quicker Sim Autonomy & Simulation Lag Improvements (33ms version), with credit to Coolspear for the original tuning approach. Its use in SFU was tested independently in combination with SFU's scheduling changes.

FAQ

1. Does SFU work with MCCC's Game Time Speed?
   No incompatibility has been found so far. I tested SFU with MCCC Game Time Speed set to 50, and its separate autonomy timeline continued to run normally.  For more details, please see my test article on Patreon.
2. Does SFU fix Social Bunny simulation lag?

No. Social Bunny uses a separate background system that SFU does not modify. TURBODRIVER's Social Bunny Simulation Lag Fix addresses that system and can be used alongside SFU.

Thank you for reading, and I hope this mod is useful in your game. :)

🍵 Enjoyed AveryPing's Sims 4 mods?

👉 Click here to treat me to a cup of nice tea! 🍵

-----------------------------------------------------------

🧪 Version 2.2 Beta 現已推出

Version 2.2 Beta 保留上一個 Beta 版所使用的 33/50/33 排程配置及自主行為調校。本次更新的重點是更安全的啟動機制與更清楚的錯誤報告，而不是改變一般遊玩時的效能表現。

SFU 現在會先確認目前地段已完成讀取，才開始推進額外的自主行為時間軸。在最初的啟動期間，只有當主模擬更新在 21 ms 以內完成時，才允許執行一次額外的自主行為處理。通過六次穩定檢查後，SFU便會回到CFG中選定的正常門檻。

因此，剛進入地段時，模擬市民可能會在短時間內顯得比舊版稍不靈敏。這段短暫延遲是刻意設計的，目的是降低地段及其遊戲系統仍在初始化時，就開始進行額外自主行為運算的風險。

先前的Stable版仍然保留。Stable版與Beta版不可同時安裝。

如果你覺得Beta版在你的電腦上運作得太積極，可以使用記事本等文字編輯器開啟其CFG，並將以下設定改為：

update_threshold_ms = 40

這項較保守的設定可能會改善部分電腦上的畫面或選單流暢度，但模擬市民有時也可能需要稍長一點的時間才會開始自主行動。

👉 非常歡迎你分享對反應速度、卡頓、讀取行為或異常互動的使用回饋。

🎯 這個模組想改善什麼？

遊玩《模擬市民4》時，我經常會注意到以下情況：

• 模擬市民完成一個動作後，會站在原地一段時間才開始下一個動作。
• 多名模擬市民反覆爭用同一個物件，失敗後又再次嘗試。
• 食物、垃圾或其他物件在相關動作完成前就被留在原處。
• 在擁擠的情況下，模擬市民的反應變得較不靈敏。

我無法斷言這些情況全都有相同的原因。不過，它們讓我開始更加關注遊戲處理自主行為的方式，以及Simulation Lag Fix這類模組如何嘗試改善模擬延遲。

Simulation Flow Unleashed是我研究並修改Simulation Lag Fix後製作的版本。根據我自己遊玩及測試遊戲的經驗，我重新檢視了其中部分時間安排，並加入數項可調整或選用的調校。

我的目標很單純：看看這些變更是否能減少模擬市民遲遲不開始下一個動作、在擁擠環境中反覆無法成功執行自主行為，或物件在動作完成前就被留在原處等情況。

這個模組並不是為了取代原有的Simulation Lag Fix版本，而是為偏好不同平衡方式的玩家，針對各自的存檔與硬體環境提供另一種調整方向。

⚙️ 為什麼我要重新製作這個模組？

2026年7月21日的更新將Memory Boost擴大至所有平台。根據EA的說明，Memory Boost旨在改善記憶體管理、減少延遲、提高畫面更新率與反應速度，並降低長時間遊玩後出現的記憶體相關問題。

我在Memory Boost較早期開放使用時就已經開始採用。我注意到讀取似乎變快了，遊戲也比較不容易出現卡頓。

然而，雖然我確實看到模擬市民的自主行為有所改善，卻不認為它已完全解決我觀察到的問題。因此，我仍繼續使用社群中常用來處理這類問題的模組，包括Simulation Lag Fix。

在使用及研究Simulation Lag Fix的過程中，我注意到其中部分設定反映了硬體與引擎資源更有限時所做的取捨。這讓我開始重新思考，其中一些較保守的限制是否仍然適合每一位玩家。

Simulation Flow Unleashed就是從探索這個問題開始的。它保留了透過獨立時間軸處理自主行為的核心概念，同時重新檢視時間切片的處理方式、允許額外自主行為處理執行的條件，以及數項相關的自主行為調校數值。

📂 我應該使用哪一個版本？

Stable版

Stable版是先前的公開版本。它使用既有的SFU排程配置，並附有獨立且可選用的調校套件。

如果你偏好先前版本的運作方式，或想繼續使用下方所述的選用自主行為調校，請選擇這個版本。

Version 2.2 Beta

2.2 Beta採用：

• 較新的 33/50/33 排程配置；
• 可分別調整的主模擬、執行門檻與自主行為數值；
• 在開始新的遊戲工作階段或載入地段時重新讀取的CFG；
• 保留自上一個Beta版的自主行為調校；
• 目前地段的讀取門檻及更安全的啟動方式；
• 更完整的相容性與執行期報告。

如果你偏好Beta版反應較靈敏的排程方式，也願意測試其較新的安全架構，請選擇這個版本。

同一時間只能安裝一個Simulation Flow Unleashed版本，也不要將其中任何一版與其他Simulation Lag Fix同時安裝。

⚖️ STABLE版有何不同？

Stable版保留原本Simulation Lag Fix作法的基本架構：

• 一般模擬流程照常運作。
• 自主行為透過獨立的自主行為時間軸處理。
• 一般模擬更新完成後，可能會再執行一次額外的自主行為處理。

與目前一般維護中的Simulation Lag Fix相比，主要差異如下：

• 允許額外自主行為處理執行的門檻。
• 一般模擬時間切片的處理方式。
• 透過獨立套件提供的選用自主行為評分、延遲、鎖定時間、權重及隨機化調整。
• 可讓玩家調整自主行為處理門檻的設定檔。

這是一項取捨，而不是在所有情況下都能提升效能。當額外的自主行為處理開始執行時，就會產生額外的CPU工作。如果你的電腦處理能力有限，或遊戲本身已承受很高的模擬負載，這可能會造成偶發的卡頓。

不過，這項取捨並不是Simulation Flow Unleashed獨有的問題，而是允許額外自主行為處理這類作法本身的一部分。

🔧 STABLE版：主要變更

1. 不強制套用17 ms時間切片上限

原本的Simulation Lag Fix使用以下設定：

time_service.TimeService.MAX_TIME_SLICE_MILLISECONDS = 17

這會將遊戲的一般模擬時間切片上限強制設為17 ms。這類限制能控制單次更新的持續時間，並降低模擬處理運作過久而影響畫面更新的可能性；但它也可能將部分模擬及自主行為相關工作切割成非常短的片段。Simulation Flow Unleashed的Stable版不會強制覆寫這項設定。在我的測試環境中，遊戲原本的時間切片上限是33 ms。

長時間使用這項設定後，我觀察到一些可能與時間切片過短有關的情況有所減少，例如：

• 模擬市民在動作完全完成前就離開。
• 物件被留在桌面或其他表面上。
• 物件未能完成後續處理。

這些觀察並不能證明所有這類問題都是時間切片上限造成的。《模擬市民4》的模擬流程包含許多彼此交互作用的系統。這項變更只是讓一般時間軸在交出執行權前，擁有更多完成工作的機會。

2. 可調整的自主行為處理門檻

原始程式碼會檢查一般模擬更新是否在約17 ms內完成，然後才允許額外的自主行為時間軸執行：

max_time_ms = self.MAX_TIME_SLICE_MILLISECONDS

if (time4 - time3) < 0.017:
time1 = time.monotonic()
result = self.autonomy_timeline.simulate(
services.game_clock_service().now(),
max_time_ms=max_time_ms
)
time2 = time.monotonic()

Stable版讓這項門檻可以透過設定檔調整，預設值為21 ms。

較低的門檻比較保守，因為只有在一般更新相對較快完成後，額外的自主行為時間軸才會獲准執行。較高的門檻則會允許自主行為時間軸在更多不同耗時的一般更新後執行。

這個數值不是自主行為時間軸本身的時間上限，而是決定一般模擬更新完成後，是否允許自主行為時間軸執行。

🔧 VERSION 2.2 BETA：主要變更

1. 分離主模擬與自主行為時間切片

主模擬時間軸與額外的自主行為時間軸現在使用不同的時間切片數值。

目前的Beta配置為：

• 主模擬時間切片：33 ms
• 更新門檻：50 ms
• 自主行為時間切片：33 ms

主模擬時間切片控制一般模擬時間軸；更新門檻決定主模擬更新後，是否可以開始一次額外的自主行為處理；自主行為時間切片則是在推進獨立自主行為時間軸時使用的目標預算。

50 ms門檻並不是給予自主行為時間軸50 ms的處理時間。這些數值也是協作式排程預算，而不是嚴格的CPU時間上限；如果某個遊戲內部函式在工作途中無法交出執行權，實際耗時偶爾可能超過設定值。

在我的測試中，搭配隨附自主行為調校的33/50/33，整體而言產生了更多成功的自主行為結果，也讓模擬市民獲得更多開始下一個動作的機會。

2. 可重新讀取的CFG設定

這三項數值可透過以下檔案調整：

[AP]SimulationFlowUnleashed_v2.2_beta.cfg

每當開始新的遊戲工作階段或載入地段時，Beta版都會重新讀取CFG。你可以回到主選單、修改CFG，然後重新載入存檔，無須完全關閉並重新啟動遊戲。

.cfg與.ts4script兩個檔案都是必要檔案，而且必須放在同一個資料夾內。

3. 地段讀取完成後的安全啟動機制

2.2 Beta會先確認目前地段的讀取畫面已結束，才開始推進額外的自主行為時間軸。

在啟動期間，SFU會暫時採用較保守的21 ms規則。只有當主模擬更新在21 ms以內完成時，才允許執行一次額外的自主行為處理。SFU會累積六次穩定檢查後，才切換至CFG中的正常門檻；如果主模擬在這段期間變得繁忙，穩定計數就會重新開始。

這項設計是為了降低以下風險：

• 地段仍在初始化時就開始進行額外的自主行為運算；
• 讀取期間的CPU工作量突然增加；
• 必要的遊戲系統尚未準備完成，就開始排程自主互動；
• 過早排程造成或暴露與時序有關的錯誤。

這項保護可能會使剛進入地段後出現一小段反應較慢的時間。啟動狀態穩定後，Beta版就會回到正常的 33/50/33 運作方式，或採用CFG中選定的數值。

4. 更完整的相容性防護

安裝變更前，SFU會檢查所需的遊戲方法，並將四個hook作為一組安裝：

• AutonomyService._register
• TimeService.start
• TimeService.update
• Zone.on_loading_screen_animation_finished

如果其中任何必要部分無法安全使用，SFU不會刻意留下不完整的安裝。SFU會維持停用狀態，遊戲則繼續使用原本的模擬系統。

這能降低遊戲更新或不相容腳本wrapper所帶來的風險，但無法保證能與未來每一次更新或所有第三方腳本模組相容。

5. 更清楚的報告

SFU只會在玩家可能需要採取行動時建立報告。

SimulationFlowUnleashed_DISABLED.txt

SFU無法安全啟動，目前並未生效。遊戲正在使用原本的模擬系統。請移除此SFU版本，並確認是否已有相容的更新。這次相容性檢查失敗不會修改你的存檔。

SimulationFlowUnleashed_Config_Warning.txt

CFG無法讀取或含有無效數值。SFU仍會繼續運作，並使用內建預設值或最後一次有效的設定。請修正或替換CFG、返回主選單，再重新載入存檔。

SimulationFlowUnleashed_Runtime_Error.txt

SFU運作期間觀察到例外。SFU出現在traceback中，並不會自動代表錯誤是SFU造成的；SFU當時可能只是在推進屬於遊戲、某項互動或其他模組的工作。

回報問題時，請附上SFU報告、遊戲產生的LastException檔案（如果有），以及模擬市民當時正在做什麼的簡短說明。

📂 安裝方式

1. 移除你要替換之SFU版本的所有檔案。

2. 將所選版本的.ts4script及CFG放在同一個Mods資料夾內。

3. 兩個檔案都必須保留；設定完成後請勿刪除CFG。

4. 只安裝與所選版本相對應的調校套件。調校套件是選用檔案，單獨安裝並不會建立獨立時間軸。

5. 確認遊戲已啟用腳本模組。

6. 切勿同時安裝Stable版與Beta版。

⚠️ 已知限制

• SFU無法消除所有模擬延遲、模擬市民發呆、互動失敗、尋路問題或物件處理未完成的成因。
• 在擁擠的社交場合中，模擬市民仍可能反覆爭用熱門物件。
• 部分第三方互動本身就有目標選擇或執行方面的問題。SFU可能改變錯誤顯現的時機，但不一定是問題的根本原因。
• 實際結果會隨硬體、存檔內容、遊戲設定、地段人數及已安裝模組而異。
• 每次額外的自主行為處理都會使用CPU時間。反應較靈敏的配置，不一定就是執行工作最少的配置。
• 在模擬負載已經很高的電腦上，運作較積極的Beta配置可能造成偶發卡頓。
• 如果主模擬經常超過設定門檻，額外的自主行為時間軸就會較少執行。
• 非常繁忙的地段、大量同時發生的事件，或高速模式結束時，可能會暫時讓更多工作留在主模擬排程中等待。

🔧 自主行為調校衝突

1. Simulation Lag Fix

由於本模組以Simulation Lag Fix為基礎，因此無法與其他Simulation Lag Fix版本同時安裝。請只保留一個版本。

2. 自主行為調校衝突

隨附的自主行為調校會修改autonomy.autonomy_modes，因此不應與其他修改相同調校的模組同時使用，包括：

‧ Zero's Faster Autonomy and Less Idle
‧ tyjokr's Less Frequent Idles
‧ Coolspear's Quicker Autonomy and Simulation Lag Improvement

3. 其他已測試的搭配

在我的測試中，本模組可與以下模組一同使用：

‧ Tyjokr's Queue Guard
‧ TURBODRIVER's Simulation Timeline Unclogger

我建議將這兩個模組與Simulation Flow Unleashed搭配使用。由於《模擬市民4》的單執行緒模擬引擎在處理物件爭用及其他模擬工作時的運作方式，有些問題很難完全避免。因此，Queue Guard與Simulation Timeline Unclogger可以作為重要的最後一道防線，協助減少即使已安裝Simulation Flow Unleashed仍可能發生的問題。

🤝 致謝

本模組以Xerox提供的Simulation Lag Fix原始碼為基礎。

Simulation Lag Fix的原始概念由simmythesim提出，之後由Xerox繼續維護，SrslySims也曾協助維護目前社群廣泛使用的版本。

特別感謝最初提出概念的原作者，以及後續持續維護這個模組的作者們。我自己並未使用最早期的版本，而是在接觸後續維護的Simulation Lag Fix後，才發現這套機制非常實用，進而促使我深入研究並製作這個重新調整的版本。

部分自主行為權重概念受到Zero's Faster Autonomy and Less Idle啟發；FULL_AUTONOMY_STATISTIC_SCORE_VARIANCE調校則以Coolspear's Quicker Sim Autonomy & Simulation Lag Improvements（33 ms版本）為基礎，原始調校方式的功勞歸於Coolspear。SFU所採用的方式，則已另外搭配SFU的排程變更進行獨立測試。

FAQ

1. SFU可以與MCCC的Game Time Speed一起使用嗎？

目前尚未發現不相容。我曾將MCCC Game Time Speed設為50並搭配SFU測試，其獨立自主行為時間軸仍能正常運作。更多資訊請參閱我在Patreon上的測試文章。

2. SFU能修正Social Bunny造成的模擬延遲嗎？

不能。Social Bunny使用SFU不會修改的另一套背景系統。TURBODRIVER's Social Bunny Simulation Lag Fix是針對該系統設計，並且可以與SFU一同使用。

感謝你閱讀這些說明，希望這個模組能對你的遊戲有所幫助。:)