A previous-wait term gives no clear NICER forecast gain for 21 PSR J0537-6910 glitch intervals

We test whether the previous waiting time adds forecast value after glitch size is known. Two linear models are fitted to 43 RXTE cases of PSR J0537-6910 and applied to 21 NICER cases. RMS error is 20.254 days for size alone and 20.272 days with the extra term. Separate block resampling of training and test cases gives gain ranges that include zero. Date perturbations preserve the shared endpoint between adjacent waits and retain the null result. A known added term is recovered in an algebraic control. This test does not exclude physical memory, missed events, or a nonlinear relation. The source data and prior benchmark were public before the test. We report no new glitch or unexplained signal. Code, frozen inputs, raw hashes, source checks, a figure, and offline steps are included.

astronomy · PSR J0537-6910 · glitch · waiting time · memory · RXTE · NICER · null result