However, since the station of Aldwych closed and became a [Ghost Station]?, the nature of the loop has changed, and the circumstances in which it applies have become significantly reduced. Indeed, when the Holland Park 2000 ruleset originally came out, the chances of this loop actually occurring were reduced to such a trivial level that people actually talked about the "abolition" of the loop (as only an actual temporal distortion, going far enough back to the time when Aldwych was open, would have actually sufficed to allow the loop to be possible. This would, of course, have been extremely rare, as temporal distortions measurable in years or even months are highly unstable at the best of times.) Naturally, this effective removal in fact removed one of the possible counterbalances and escapes from a [Dollis Hill Loop]?, and indeed after the disastrous second set of amendments even threatened to reintroduce the [Mornington Crescent Loop]?.
The introduction of split-weighting for ghost stations (accounting for the station's importance and its effect on the network when it was a "live" station, as well as its current ghost status) in the third set of HP2K amendments was, largely, an attempt to at least partially restore the A-A loop to its former glory. The results have been dubbed a "qualified success" (i.e. it works about three times out of ten, maybe four if you don't look too closely.)