Android TV updates sometimes revoke a previously granted battery-optimization exemption without asking again, and once that's revoked, the system can kill the app before it fully loads, especially if it was running quietly in the background at the time. This reads to you as a random, unexplained crash, but it is really the OS managing power more aggressively on the new build than it did before the update installed on your box. Re-adding the app to the allowed background list in system settings is the direct, reliable fix for this particular cause.

An Android TV update can change how the system routes video decoding by default without any visible warning, and hardware decode can fail silently for a stream type the update no longer handles the same way it used to before installing. Testing a lower resolution or lower-bitrate channel isolates whether decode is the actual bottleneck here, since a lighter stream working fine while a heavier one consistently fails points squarely at the decode path changing, not your home connection or bandwidth being the cause of what you're seeing on screen.

If multiple apps are misbehaving after the same system update, the cause is more likely the update filling the device's storage or cache partition, which is common on Android TV boxes with limited built-in storage to begin with, leaving no room left for temp buffer files any app needs to run smoothly through playback without stalling out mid-stream. Clearing cache device-wide and checking available storage explains this pattern far better than troubleshooting each individual app one at a time as if they were unrelated problems with separate root causes, when in fact one shared cause explains all of them at once.

Some updated builds don't fully release memory when an app is backgrounded and then resumed later, so switching away and back destabilizes the session even though a completely fresh launch works fine every time on the same device with the same content playing. This is a distinct crash pattern from a plain launch failure, and it's easy to mistake for general instability if you haven't isolated it first with a direct test. Fully closing and relaunching the app, instead of just resuming it from the background, avoids triggering this specific fault entirely.