For my planning I have a master spreadsheet which shows the consequences of connections taking up to 3 minutes (e.g. if your connection is due in 2 mins it will then show the next connection after that (which may be 10 mins). I've also got assumptions for run times (which I now think are too conservative for the longer ones). It also assumes that no negative time changes are ever made. It gives a hierachy of outcomes (I think Hakan used to do something similar and assign probablities) and so you can see fairly readily which are the key connections to make and which don't matter too much as you will probably end up on the same train later. The most optimistic time this gives at the moment is 16h38m. Yesterday, I was ahead of this most optimistic schedule, which given towards the end effectively there is a 10 minute frequency on most of the relevant routes - potentially means 16h28m, except that my minute's delay at the start would have made 16h27m if the final train ran to schedule.geofftech wrote: too many questions to ask really ... so i'll go with the most pertinent one (which you may or may not reveal) is how fast you had hoped you might do it it, or how fast it might it have been if things had run better for you after 7pm... 16:18, or 16:08 ??!?
do you also think overall you got lucky (with buses, etc...) and that the same route on a different day would be slower... or faster?
Overall with average luck (but no major service delays or key trains cancelled), I think I should get 16h48m - so overall I think I have to say the luck was against me yesterday.
I think I need to look again at my assumptions about non-tube transfers and build in something more realistic for the interplay between running and busses. I think if I can get the connections I had in the morning of April 21 combined with those I had in the afternoon of yesterday we could be considering getting an earlier MHE which should lead to an earlier Hainault Loop (or vice versa) - so potentially 20 minutes ahead of 16h38m which would be 16h18m (but I haven't looked at any of this in detail, so this speculation may not be borne out once I check the timetables.) I suspect though that consistently trying to achieve these most optimistic times in rush hour in the centre is unrealistic (unless you choose Maundy Thursday ).
The recent opening of the the third platform at Stanmore means that more trains are running through to there which has been a big help.
What struck me yesterday was how much time tube trains spend just sitting around either in or just outside stations. My perception is that if LUL had really efficient scheduling/signalling software and these could be eliminated there could potentially be big gains made without the network having any more trains or them running any faster.
I'm not sure if I've really answered your question Geoff, I think that I probbaly need to do some more planning and test it out again soon