Risks associated with remaining on a single frame for extended periods

In my game, players can engage in monster hunting even while inactive, which results in some users keeping the game running for extended periods. During this idle time, there is no frame movement, and players stay on a frame with around 10,000 lines of events for several to over ten hours. Subsequently, although minor processing lags are acknowledged, they do not seem to hinder gameplay significantly.

I am using list objects to manage loot, which is saved in arrays. However, abnormalities in the acquisition process seem to emerge as more time elapses. Despite meticulous reviews of the related events, no flaws have been found, and in fact, no anomalies occur during tests of around 5-6 hours. Is it possible that keeping the game running on a single frame for long durations, like 10 hours, could be the cause of these abnormal processes? Unnecessary objects are being properly disposed of, and there is no issue of reaching the object count limit; the number of objects remains consistent regardless of the hours passed since launch. Given the minor processing lags, memory issues also seem unlikely.

Would incorporating processes like periodically moving across other frames be necessary to ensure safety?

Quote from Fernando

By only considering some value types, you can reach around 414 days. I will take this task for weekend to see if anything can be of a problem, of course i will appreciate it if you can add the extension list used in your app.

Thank you for your assistance. The extension I am using is as shown in the image. screen_shot.mfx is something created by a Japanese Clicker for posting screenshots to Twitter.

Following further investigation, suspicion has arisen that the patternization of random numbers may be a factor. Reports of abnormalities regarding loot acquisition from users include issues such as dropping equipment with the exact same performance despite having extremely high randomness in their status, or instances where the drop rate is extremely biased. Upon conducting a probability test, a regularity was discovered. This pattern appears to differ with each startup. While I was aware that the denominator cannot exceed 65535, I was not aware that there was a regularity. Could the seed value be the cause? If so, is it possible to change the seed value? Additionally, for some reason, in this test, 1/16384 is not counted even after conducting a million draws.

The issue might not necessarily stem from long-duration gameplay itself, but rather the increased number of attempts through prolonged play could expose this pattern, potentially raising suspicions among users.

I have uploaded the mfa to Google Drive, so I would appreciate it if you could check it.

Please login to see this link.

Thank you in advance.

Quote from Fernando

Yes, random seed is a short.

You may need to create your own extension for this.

Changing this will alter all old application.

It would be very helpful if this could be achieved through an extension. I kindly request that you consider its implementation.