Twisting wire on VPI for anti skate?

With no anti skating, all was well except one album that skipped on the first track.

A/S is intended to balance stylus pressures against the opposing groovewalls during play, not to prevent the stylus from racing down the angled slope of most lead-in grooves.

The key to preventing skipping during cueing is to keep the tonearm under control (manually) until the stylus locks into a groove. It would be a major error to use A/S for that purpose, you'd end up with far too much for playing the music.

It's debatable that A/S is needed at all. Many people including me don't use it at all. The test is not how does the arm cue, the test is how does it play. If you can track the toughest passages on tight inner grooves without R channel breakup or fuzziness, you don't actually need A/S at all. You may want some depending on your sonic preferences, but that's a matter for individual experimentation and listening.

The manual does say it very succinctly. Read it more carefully

My method consists of minimum anti-skate. First I set up the cartridge, for SRA, Azimuth and VTF, with no anti-skate. I also take a Digital Multi-Meter reading of the cross talk as a reference after setting azimuth. No twisted wire, fish wire disconnected. So now everything is set, VTF, SRA, Azimuth, and I have a reading of my cross talk with Azimuth set.
I have the VPI Classic 3, Dyna XX-2 MK II, VTF 1.988 grams. Crosstalk is -26.620db on the Left channel and -26.021 on the right. Azimuth is 0.19 degrees tilted to the right channel. SRA is 92 degrees (measure with a digital microscope).
With the SRA set (using the VTA tower), I make sure the fish wire is horizontal to the mechanical anti-skate.
Now I hook up the fist wire, ensuring the attachment point on the tonearm and to the top of the mechanical anti-skate is horizontal. I line it up, so I have one donut on the bottom, so I can put the fish line on the 4th indent from the top (first indent from the bottom). So one donut under the 4th indent from the top, the fish line on the 4th indent from the top.
Then I pull out the Hi-Fi news test record. I use the easiest bias setting test, Side one band 6, 300hz tone L+R, +12 db. When I play that with no anti-skate, left channel is perfect, the right channel will distort occasionally. I hook up the anti-skate and put one donut on the horizontal bar. I slide the donut out, until I don't get any distortion in the right channel on the +12db track. The one donut ends up close to the end of the rod. I don't even play the +14db, +16db and +18 db tracks. Those will lead you to nowhere.
So now the anti-skate is set, minimun with the +12db track. I then just double check crosstalk to ensure it is the same. I just want to be sure I am not twisting the tonearm in some way. This is probably an unnecessary step, since I have never noticed a difference in crosstalk, probably because I am careful about the horizontal setup of the fish wire.

Thanks for the info! And I'll re-read the manual.
It sounds amazing as it is but it can't hurt to explore a bit more and I do have the Hi-Fi News album.

1. Anti-skating does not affect crosstalk. Checking crosstalk because you adjusted A/S is like checking the oil because you adjusted tire pressure. ;)

2. The HFN&RR test record makes a handy dust cover for the platter when the TT's not in use. It serves little other practical purpose. Mine hasn't been out of its jacket in years.

Its so-called anti-bias (anti-skating) tracks are incorrectly designed and their instructions poorly written. Thousands of people have mis-used them to set anti-bias higher than it needs to be, to the detriment of their music and perhaps their cartridges.

***
In the end, one can't fine tune VTF with a scale, VTA/SRA with a microscope or anti-skating with a test record. One must train ears and brain to understand how these parameters affect the behavior of the stylus in the groove whilst playing real music.