I've also done a little more digging today... The crash followed the usual symptoms of some graphical glitches followed by a total freeze.
Looking through the rest of the log, I saw some systemic memory failures, which was interesting because the only application I had running was SWGOH. Looking at the stack trace, it is another example of a null pointer issue. This is often a failure mode in C/C++ applications when they fail to aloocate memory (and then don't check if the allocation failed). So my working hypothesis is that we still have memory issues.
Fortunately, this time I was running the adb shell at the same time, so can compare the memory usage of SWGOH at the time of the crash and after normal operation.
Crash:
PID USER PR NI CPU% S #THR VSS RSS PCY Name
30303 u0_a211 16 -4 3% S 102 3080860K 850216K fg com.ea.game.starwarscapital_row
Normal operation:
PID USER PR NI CPU% S #THR VSS RSS PCY Name
31616 u0_a211 16 -4 9% S 102 2137360K 638212K fg com.ea.game.starwarscapital_row
The key thing to note here is that the crash is trying to use nearly 1GB more memory (VSS) than in normal operation. I suspect that there is something causing the application to leak memory over time.
My guess is that the devices seeing more crashes are ones that have smaller RAM, or for some reason are hitting the condition that causes the memory leak more often.
Would anyone in Support like to review and comment on my analysis?