NavList:

Two recent posts have advocated the use of three-star fixes in aircraft over two-star fixes.  I think it depends upon which type of aircraft you’re flying in and how many crew members are prepared to dedicate themselves to the task. 
The RAF Vulcan force used five or even seven-shot two-star fixes utilising the entire five-member crew.  The left-hand seat Pilot flew the aircraft using whichever method they felt best for the prevailing conditions, but almost invariably with the compass systems in DG mode.  The right-hand seat pilot noted the speed change, because the RHS ASI was the most sensitive.  The Nav-Radar shot the astro.  The Nav-Plotter monitored the heading change because they had the most accurate heading readout, and the AEO called the timing.

The choice of two stars might have been helped by the fact that the blister on the top of the fuselage for the pilots’ heads extend backwards over the cabin and housed the multi-seat dinghy so that the Vulcan needed two sextant mountings, one either side of the fuselage.  The nice people in Ops would let you draw a second sextant pre-flight, so you could have one sextant already set up on your favourite stars throughout the trip.  All the Nav-radar had to do was step across the cabin between stars and remember to call out the Ho before winding the sextant up again.  In the tropopause with five people working at a sandwich fix there was only need for one assumed position, and I can’t remember ever having to use MOO and MOB.  If the entire sequence was early or late, you simply moved the assumed position east or west (MOB by drawing).
 
In general, the Nav-radar chose the stars and calculated their altitudes and azimuths using the AA and AP3270, and the these were double checked by the Nav-plotter if he had time.  The plotter filled in the altitudes, speed changes, and heading changes on the astro sheet during the shots, put the GPI 6 to fix at mid-time, and calculated the acceleration error post shots using his Mears Slide.  The Plotter then put the fix and the GPI position on the chart while the Radar double-checked the plotters arithmetic on the astro sheet.  I never spotted a single mistake.

Next came the MPP.  Officially the line between the two positions was to be divided in the ratio of the CEPs to be expected from the astro and the Nav/Bombing System.  More often I suspect, the Radar and Plotter would l look into each other’s eyes while the Plotter’s pencil hovered above the line.  The Radar knew how his shooting had gone.  The Plotter would know how well the system was working.  With a nod, the man from Del Monte would say yes, and defying all the laws of statistics his ‘Ouija’ pencil would come down in the correct place.  

The main problem with five, and sometimes seven-shot sandwich fixing was that it could take up to 15 minutes after mid-time to get a fix on the chart, and in that time, you might have flown 120nm.  Another problem, which I don’t recall anyone ever mentioning at the time, was that you needed to check that neither star was in upper or lower transit. The advantage was the celestial gymnastics and chartwork were relatively simple and you got the benefit of a lot of averaging for your efforts. 

A tight three star cocked hat might give you a that warm fuzzy feeling (despite statisticians proving time and again that this can be misguided), but don’t do down the benefits of two-star sandwich fixing when the conditions are appropriate.  DaveP