The brightness of a nova prior to its eruption is both critical for evolution models and hard to measure. To catch the pre-eruption magnitudes (mpre) for the many faint and not-recent novae, I have made nine trips to archives across Europe and America seeking deep and old plates. With this forming most of the cases, I now have a total of 71 classical novae for which the mpre has been measured, although 11 cases have only usefully deep limits. For all cases, I have pulled out measures of the magnitudes in the same band for the last few years. All these post-eruption magnitudes (mpost) are from long after the end of the eruption and the flattening of the light curve. I find that 86% of the novae have mpre = mpost, with an RMS scatter of 0.57 mag due to the usual fluctuations in nova light curves on all time scales. (The remainder, called V1500 Cyg stars, are >2.5 mag brighter than pre-eruption in a very extended fade due to irradiation, with the cause apparently due to the white dwarf having some sustained nuclear burning.) These pre-eruption magnitudes can be used to test two strong predictions from the new version of the Hibernation model (Hillman et al. 2020) for the evolution of cataclysmic variables. The first prediction is that mpost (for 30 years after the nova peak) must be greatly fainter than mpre, with more than a 5.5 mag difference for systems with orbital periods under 4 hours. From my sample of 7 novae with P<2hr, all have the mpost brighter or equal to mpre. From my sample of 17 novae with 2hr<P<4hr, mpre ~ mpost, with the faintest mpost being within 1.3 mag of mpre. So the first prediction of new-Hibernation fails greatly. The second prediction is that all the systems with P<2hr will have absolute magnitude MV=6.5. To test this, only three novae with P<2 hr have usable Gaia EDR3 distances, and these have limits on MV of >10.0, >9.3, and >9.1. So the second new-Hibernation prediction fails greatly.