[r-t] Shipway Minor

[Matthew Frye]

On 16 Jun 2014, at 22:49, Alexander Holroyd <holroyd at math.ubc.ca> wrote:
> Well, don't forget that the "division" (or lead) of shipway consists of two eights, so the Stedman Triples type calls already occur in two places in the lead (as they do in Stedman).  So I don't know what the "obvious calls" are, and the extremes are actually division-end calls (if you start at a symmetry point).  So my extent actually has 5 different types of call, although that doesn't bother me…

I make no apology for considering the "division end" to not be at the same point as we (conventionally) start the method. I also consider both quick and slow eights to be separate divisions (and the method to be made of strictly alternating between the two), so standard Stedman bobs are in my view all at division ends.

> My own criteria for what makes a reasonable extent don't have that much to do with the number and placement of calls.  To me it is more important that some decent sized chunks of the plain course get rung, and that when calls do occur they don't dirupt the method so much that it no longer feels like ringing it.  (In Shipway I would even be willing, just, to accept 56 places made in multiple possible places during the eight - I started out my current investigation with a 5 part extent like that).
> 
> I'd be interested to hear other people views on what constitutes a reasonable extent in a problem method.  I suspect there are substantial differences.

"Reasonable" is obviously very variable between different methods, not least depending on what seems to be possible. In this case, the existence of the structure provided by eights seems to be something I would ideally like to be kept intact if at all possible. This extent gets very close to this, akin to an extent of Grandsire Triples with only 2 singles in a single SBBS section. If this turns out to be the best it is possible to do I wouldn't be overly disappointed, but I would certainly have a preference for an extent without those mid-division calls if it is possible.

> I suspect a MEB is possible with only one type of call, even _any_ call that doesn't obviously fail for some reason.  In fact there might even be a general theorem to that effect.


Sketch of proof if the call you want to use can access every possible course (usually true for a single): select Q-sets to include all the courses. Add in extra plain whole courses to get the same number of every course. Requires calls to be single change. Can be extended to some methods/calls where you can't access every possible course, e.g. bobs only S Minor with correct nature of rows through a lead. Obviously an Alliance method requires a variable hunt bob, although this is probably captured under the "access every course" requirement which could probably be better stated to require courses with different hunt bells for Alliance, but not for e.g. a standard Surprise.

Not terribly useful for Shipway, as it'll give you at least a 16 extent block, possibly more.

MF