The way cannon shells are doing damage is fundamentally wrong. They should not be ripping wings off. They should not be blowing tails off, or larger control surfaces. What they should be doing is cutting control cables, starting fuel and oil leaks, and killing your pilot.
High explosive shells
I want to start the conversation with this video for a very simple reason. The shells used in the test data have 58% more muzzle energy, and 3 times as much TNT as the shells in the video.
Let’s start with some terminology. An “A” kill is damage serious enough for the aircraft to be disabled within 5 minutes(Think fires, serious engine damage, pilot death). A “B” kill is damage that will cause the aircraft to crash between 5 minutes and two hours(Think fuel and oil leaks). The only thing that matters to Warthunder is “A” kills. This post relies heavily on data taken from the optimum caliber program and is by proxy, heavy with terminology derived from it.
These firings were conducted against a P-47 at its most vulnerable angle, where the pilot is not protected by armor, and where most of the internal systems are. The chance for a singular 20MM HEI shell to kill a P-47 is 6.1%, with nearly half of that being the death of the pilot. Structure accounts for a measly 1.9% of the total chance for an “A” kill. The report defines structure as:
To end this conversation on explosive damage, let’s look at some real-world examples. In fairness, these are examples of survivorship bias, but it’s important to look at them nonetheless.
Entry hole(MG-151 mineshell):
Exit holes(MG-151 mineshell):
A6M zero 20mm entry hole:
Exit:
Note that the control cables are completely undamaged, even though the cannon shell fuzed right on top of them.
For reference, one 30mm Mk 108 damage:
So, are smaller-caliber cannons overperforming? Yes, absolutely. They are doing far too much damage to the structure of the aircraft. That Bearcat at the beginning should not have lost its wingtip. I highly doubt it would have even fully lost its aileron or wing control, given the fact that the Bearcat does not use cables, but push-pull rods. So, what should high explosive cannon shell damage be doing? It should mostly be killing pilots/ crew, causing excessive drag, and making control surfaces inoperable. So why change the way cannon damage works now? The severe damage mechanic effectively eliminates kill stealing, which is the main reason that cannon damage was so high to begin with.
I would like to thank @KillaKiwi for the Bearcat footage and @DerGrafVonZahl for giving me data from the optimum caliber program.