For these 8 projects, hereâ€™s the number of prime numbers discovered in the past 5 years  0 for 321, 0 for CUL, 2 for ESP, 5 for GCW, 1 for PSP, 1 for SOB, 1 for TRP and 1 for WOO.
Now that LLR2 is running a â€œshortâ€ double check, using about 1% of the CPU time that the â€œfullâ€ LLR test requires, what are the pros and cons of running LLR2 double checks tasks first, then if LLR2 verifies a prime number, run the â€œfullâ€ LLR task?
If this LLR2 strategy were to work, PrimeGrid could determine if many more numbers are prime in each project:
 Does LLR2 ever not verify a prime number that LLR has discovered?
 I would assume that LLR2 may determine that erroneous numbers are prime numbers, but LLR would verify if itâ€™s a prime number?
 Is there any knowledge of the probability of LLR2 missing a prime number?
Philosophical Question  Is it better to run many more tasks to try to find a prime number, and possibly missing a prime number OR continue the current more thorough methodology?
