The answer to many questions are unique to your gun and ammo and may be different from what others may report. Your get a lot of opinions and "a friend of my buddy" answers from folks that are really trying to help but may be different from what you find with your equipment if you do the testing yourself.
Amen to that!
I did a classic varied-jump series a few weeks ago with the newish 183gn SMK in my 284 (1:8 gain-twist 30-inch Bartlein; load: Lapua, Murom KVB-7, 53.8gn Viht N165.)
I usually shoot 180gn SMKs set around 15-20 thou' 'in' or 175gn SMKs jumped in this rifle, same brass, primer and powder, but wanted to see how the very long 183 'uber-VLD' worked out as I have the barrel twist to allow its use. So I'd previously had a go with varying charges and the bullet well 'in'. ES values were fantastic, all single figures for 4-round batches, one charge weight with an ES of 1fps - ie three shots with identical MV and one only 1 fps different. (With some of the chronographs I've owned over the years, I'd have discounted these ranges as a chronograph accuracy error, but this was from a Labradar.)
Groups though were not as satisfactory. All had at least a trace of vertical, the worst a four shot vertical string. So the 183s went back on the shelf and relegated to another 'go' when time and inclination allowed. Then I started to see AS Forum threads saying they like to be jumped, super-VLD long nose shape or not (Litz measured Rt/R ratio value of 0.37, one of the lowest values = most aggressive secant ogives listed in his latest book on bullet shapes and performance,
The Performance of Rifle Bullets 3rd edition.)
So, taking the charge weight giving the least amount of vertical in the previous series, I loaded 3 shots at the original 'in' COAL simply to foul the barrel and provide a baseline MV, then 5 rounds each at: barely touching, 10 out, 20 out, 40 out
MV for the first three 'in' was 2,738 and I didn't bother about ES, then:
just touching: 2,741 Av / 17 ES
10 out: ......... 2,740 Av / ES 9
20 out: ......... 2,738 Av / ES 11
40 out: ......... 2,728 Av / ES 12
So, until the final 40 thou' jump, remarkably consistent MVs and no real range of ES values although no 1 fps result now! The slight drop for a large jump confirmed the previous studies I've seen on this, principally one by Lapua or Viht many years ago with a 7.62 NATO military loading showing seating the bullet successively deeper in steps produced neither pressure nor velocity change until a certain point was reached where the jump was such that the bullet gets 'a run at the lands' and both readings fell. (This of course is the opposite of what many believe to be the case.)
(As to groups, the 40 thou' jump did the trick and I now need to see how the combination performs at long range in a match. Using it in a 600 yard BR match produced promising results, but the weather was such that it wasn't a good day to learn a great deal - or shoot small groups. However, as Alex says under 20 fps and that's good enough as long as the groups are OK)
Two other things worth mentioning. The 100 yard groups not only saw no or little MV change but barring the 'just touching' setting which produced a 0.9-inch C to C vertical string, every other group centre was in a near horizontal line. Finally, the BR match used Shot Marker electronic targets that also measure retained velocity and they produced SDs that would pretty well match those with 10-20 fps ES values at the muzzle, so one of the four strings aside, it seems spreads didn't grow by much over 600 yards with this bullet at ~2,730 fps MV.
BUT ... this is one make / model of bullet in one load combination from my Barnard P action and Bartlein barrel. As chkunz says, it just cannot be extrapolated to other bullets, barrels or loads.