There's a surprising comment in the later edition of Magriel's Backgammon where he says he later came to view making the opponent's bar point as possibly even more important than making the famous "golden point." That sounds like an overcorrection to me, but it does underscore how valuable it is to make that point in the early game, especially when the opponent has improved her offensive position.
In undeveloped or moderately developed positions, the 18pt anchor is almost always more valuable than the 20pt anchor. I came to this conclusion on my own. Among other observations...
When I wrote the tutorial, I looked for positions reached by short, well-played sequences for which the top plays within the same Nactation family were very close in equity. In the early game, the option of making the 18pt kept winning by a clear to large margin, but with some trial and error, I found one for which making the 20pt is only slightly worse. See Nactation tutorial, position #34.
As an aside, I'll add, based on my experience, that most players hit too often and don't make (or keep) an advanced anchor often enough. In the aforementioned #34 position (also diagrammed below), I incorporated that knowledge for any readers who might be paying attention.
When the opponent has a strong board, especially if you're on the bar or very likely to be hit, the 20pt anchor can be better than the 18pt anchor. Match score is also a factor.
I don't know how thoroughly or sloppily the comment in the later edition of Backgammon was vetted. If they had asked me, I would have recommended to the revisionist(s) that they should:
(a) qualify their statement along the lines of my notes above, or
(b) determine their assertion scientifically and elucidate the methodology for making that determination, or
(c) omit the comment. :)
Nack
                       |
| ||||
| XGID=-aa---E-C--aeD---bAeA----A:0:0:1:32:0:0:3:0:10 | |||||
| 62S-41X-65H-62H-32...[@ #10 H46] | |||||