Today has been interesting as I set about testing the hardness of a few pellet brands. It reminded me that around 1965 I spent a happy summer holiday job with Imperial Metal Industries, Kynoch, Birmingham and spent some time testing the hardness of titanium billets I was told were for Concord. I think the machine was Brinnel or Rockwell? It forced a ball bearing into the metal at a set pressure and then you had to measure the indentation using a microscope.
But back to the job in hand. I have no such machine so had to devise one. I toyed with setting up a 'penetrating' machine to force a pointed whatever into the pellet and measure how far it penetrated for a given force. I dismissed that idea as a bit too complex. I toyed with a sort of 'swinging' hammer system to thwack the pellet, but gave up on that as well because it would be difficult to regulate. What I ended up doing was taking a 3/8" metal drift and dropping it down a tube onto the pellet head and measuring the new height of the pellet. Bear with me ... I have not forgotten the point raised by Jim regarding using pellets for this type of test.
But I did test pellets, just for fun. I dropped the drift from a scientifically determined height of 40cm onto the pellet head with the pellet standing on a heavy steel sheet (part of a vice). The 40cm was carefully selected because: a. it worked on a test sample and b. it was the length of pipe I had in my box. Using this method I tested 5 pellets of each of a range of pellets in both .22 and .177 although I did not have all pellets to form a complete set.
After this, and considering Jim's comments about the unreliability of pellet testing like this because the results could be / would be compromised by skirt thickness, I used the tested pellets to make some 5mm diameter lead rods. For each pellet brand I pooled the lead from .177 and .22, melted it down and poured it into moulds; making 2 x 8.5mm long rods for each pellet brand. I then used the rods as before except that the drift was now placed 83cm above the rod (again scientifically determined by what I had available).
Results in all cases were pretty consistent within a pellet type/brand and I give here a summary of the summary data.
Lets look at pellet data first, even though it is an unreliable indicator:
For .22 pellets, the height reduction varied from 29% (Superdome, classed hard) to 29.7 (Marksman), 35.9 (later Wasps), 36.8 (FX), 38.9 (AA Field) to 40 (Accupel, soft). I had no .22 JSB.
Looking at the .177 versions showed: Ely Wasp (41.4, hard), 46 (Superdome), 46.4 (Marksman), 48.9 (AA Field), 49.2 (FX) and 49.5 (JSB Exact). I had no .177 Accupel
So ... reasonable agreement when you consider that I think the data ranking could be variable by +/- 1 position.

But onto the rods. This trial showedagain % height reduction of the rod)
Accupel 20 classed as hard followed by Eley Wasp 25.2, FX 27.06, Superdomes 28.35, JSB Exact 29.82, AA Field 33.35 classed soft. Again, a truer picture using proper equipment may well change the positions of adjacent pellet makes but I tend to think the data are not too bad.

What is interesting is the differences between the pellet tests and the rod tests.In the pellet test, for instance, Accupel was deemed the softest in .22 (I had no .177) yet in the more definitive rod test it came out the hardest. AA Field, on the other hand came out as soft in the rod test and quite soft in the pellet test.
A casual observation that could well be wrong/coincidence was that the Marksman pellets took longer to melt than other types.
So there we have it, not a truly comprehensive array of pellet types but interesting data all the same.

Overall, a fun morning.
Cheers, Phil