as I understand it the bullet has to "stick" by semi-lodging in the throat of the barrel (beginning of bore/rifling) but too slow to catch the "wave" of propulsion due to gas expansion, some of the pressure "blows by" and it's enough that the bullet WILL NOT get enough to further it down the bore completely. once it's past this point of no return, if it has enough energy behind it to keep the pressure building past the strength of the steel and weapon design BANG!!!
going by my way of thinking, it's easier to do with bottleneck cartridges because as the mouth expands and lets the bullet jump the freebore into the rifling there is a "compressed" column of propulsion behind the bullet pushing the center of the base. if the gas can "blow by" the projectile then the pressure is pushing lateral/outward instead of down the barrel, it's creating pressure on both ends of the round (how that's measured I don't know, ask a physicist to detail those effects), and the extra surface area of the shoulder allows that "shape-charge funnel" to direct the remaining energy from the still burning gas onto the back of the projectile to pressure spike a second and FINAL time upon lodging inside the barrel. this is exacerbated by low loading density, the powder can burn low in the case pushing low on the bullet, instead of centered as it is supposed to. low density means inconsistent "push" follow-through to direct the bullet perfectly forward and not slightly angled, not enough power is not enough to overcome bore friction, and the second spike is the unburnt and still burning powder pushing not only on the back of the bullet but ALSO the inside of the barrel but not in a column straight out, but every direction because it's going for the path of least resistance. that path is everywhere because the chamber is bulging, the steel is NOT strong enough to contain the pressure but a fraction of a moment (long enough if the pressure is given an easy way out the bore, as it is with a properly loaded cartridge).
also, the more powder you pack in, the more potential energy you get- especially with a ball powder that isn't compressed tight enough to get a good-consistant-constant burn. PROGRESSIVE BURN IT THE KEY! with a really slow powder, the initial pressurization due to the primer going off can cause the bullet to walk out of the crimp and into the rifling, which lowers the pressure while the powder burns slower than anticipated. it WILL built that pressure up WAY higher than you meant though, but it won't build to the intended peak and then continue to burn (progressive burn) to keep the pressure somewhat constant while the bullet travels down the bore- it will build the pressure slow and then when the gas pressure has no where to go it stops going up bit by bit and BANG!!! Detonation!!! PS- this is a known issue for H110 in .357 magnun, and it needs no bottleneck because a revolver cylinder gap and forcing cone provide that "shape charge funnel" for it to happen. I'll tell you from personal experience that going from top to bottom, 1.2grs powder charge deviation of h110 in starline brass, winchester spm primers, berry's plated double struck 125gr hp and/or hornady xtp 125gr hp I had noticed the recoil being strong but acceptable to fierce and with companion muzzleblast like a shotgun in the movies (huge orange flash out the barrel visible in strong sunlight, lots of muzzle travel up and down despite the compensator). if you use slow burning ball powders watch yourself in general! extruded and flakes "bulk up" to make sure there is air that can be a temporary filler to "cushion" the pressure increase (it burns up with the powder). can anyone put that into laymans' terms?
