Reliability stats on an all line rover

[whole diary]

[grahame]

An interesting spread sheet - Robin, thank you for permission to reprint, from an all line rover on train performance.

	Total delay (mins)	Total delay (mins)	Average delay	Average delay	No. trains	Total mins	Total mlles	Overall av.
TOC▸ (Train Operating Company)	on departure	on arrival	on departure	on arrival	used	in transit	in transit	speed
GWR▸ (Great Western Railway)	102	236	3.1875	7.38	32	1634	1708	62.72
XC▸ (Cross Country Trains (franchise))	59	72	4.21	5.14	14	1720	1663	58.01
Northern	15	32	0.63	1.33	24	947	514	32.57
London Midland	14	28	1.40	2.80	10	218	131	36.06
Virgin	31	25	5.17	4.17	6	692	918	79.60
Merseyrail	14	10	2.33	1.67	6	147	53	21.63
Arriva	4	19	0.80	3.80	5	294	227	46.33
South Eastern	0	4	0.00	1.00	4	156	108	41.54
LNER» (London North Eastern Railway - about)	7	17	1.75	4.25	4	560	794	85.07
East Midlands	0	0	0.00	0.00	2	161	158	58.88
South Western	5	7	2.50	3.50	2	216	178	49.44
Trans Pennine	2	-2	1.00	-1.00	2	67	52	46.57
Grand Central	0	0	0.00	0.00	1	53	21	23.77

I've had a look at the data and have reached some thoughts / drawn some views - I will leave others to comment before I do - see if we come up with the same things.

[eightonedee]

One thing rather sticks out, even allowing for the fact that the number of journeys on most operators probably means the results are not statistically significant, is GWR^▸ (Great Western Railway)'s poor performance. They lost most time on average - over 4 minutes per journey, worse than XC^▸ (Cross Country Trains (franchise)) who historically have a reputation for increasingly losing time as they cross from one other operator's territory to another and pick up the cumulative effects of their individual problems, lose slots etc.

It also highlights how Journey Check's optimism in usually showing late running trains making up lost time during the remaining parts of the journey is not well founded.

Also - not an bad performance from Northern considering the beasting they normally get, although I guess it reflects perhaps no peak time commuter services in the major cities?

[eightf48544]

GWR^▸ (Great Western Railway) doesn't look to good with 7.38 mins late on average and only 52,72 av speed.

Virging and LNER^» (London North Eastern Railway - about) come out pretty well.

Virgin average late with 5.17 av speed 79.6
LNER                               4.25 av speed 85.07  

East Midlands, and Grand Central were on time but av speed were quite low.

Trans Pennine, and Northern  despite their reputation didn't do too badly with TP early

I think what it illustrates is the randomness of late running you pays your money and hope the trains on time.

Have you included any cancellations? In my opinion They are worse than late running.

A reliable railway is a punctual railway.

[Robin Summerhill]

This was my analysis and came about during the validity of a 14-day all line rail rover ticket in July 2018. The original objective was to revisit a number of my "old haunts" whilst recording the end of steam traction on BR^▸ (British Rail(ways)) between 1965 and 1968, and to compile some "then and now" photos. They are now beginning to appear on my Flickr site (as an example:  https://www.flickr.com/photos/93122458@N08/45215021952/ ). The analysis was an "add on," not particularly scientific, but just a random sample of trains that I happened to use over the 14 days, and I suppose something of a post-retirement return to the data collection and analysis work that I did professionally for some 30 years!

What it does do, however, is to give a picture of what the average passenger may experience from the various TOCs^▸ (Train Operating Company). There is a tendency for "familiarity breeds contempt" to kick in if one was to carry out a survey of passengers ie. if you asked a random member of the public standing on the platform at Bath Spa who the worst TOC in the country was, the most likely answer you would get would be "GWR^▸ (Great Western Railway)." Do the same thing at Blackburn and the answer would probably be Northern; at Ludlow it would be Arriva Wales, and so on.

I thought that the comparison between GWR and XC^▸ (Cross Country Trains (franchise)) was particularly interesting, because I used their trains more than any other and they operate in similar environments (over relatively long distances with relatively frequent stops). It was suggested to me that XC might "look" better because of lengthy dwell times at some locations, although a cursory examination of the WTTs^▸ (Working Time-Table) will show that this is hardly unique, and many TOCs appear to shoehorn as many minutes as they can get away with into the timetable at the end of journeys to improve their timekeeping stats. Another matter I thought revealing was the relatively good performance by Northern Trains, who have taken much "stick" in the press since the timetable debacle in May. It appears that when they do manage to run trains they tend to run them on time!

Such a summary as this does of course not show individual train timekeeping performance, but if any one is interested in that level of detail, if they email me I will send them the spreadsheet.

[grahame]

What it does do, however, is to give a picture of what the average passenger may experience from the various TOCs^▸ (Train Operating Company). There is a tendency for "familiarity breeds contempt" to kick in if one was to carry out a survey of passengers ie. if you asked a random member of the public standing on the platform at Bath Spa who the worst TOC in the country was, the most likely answer you would get would be "GWR^▸ (Great Western Railway)." Do the same thing at Blackburn and the answer would probably be Northern; at Ludlow it would be Arriva Wales, and so on.

A big welcome to the forum, Robin ... such data is fascinating.

I'll give away my first general thought when I saw the data - looking at our discussion of a five minute connection from a GWR train to a bus at certain stations and wondering if that was going to worsen rather that lessen the current connection failures with train averaging more than 5 minutes late!

Would people know who there TOC was ... GWR and Northern are well enough know I suppose --- Arriva Wales is now historic thought it was current during your travels ...

[Robin Summerhill]

GWR^▸ (Great Western Railway) doesn't look to good with 7.38 mins late on average and only 52,72 av speed.

Virging and LNER^» (London North Eastern Railway - about) come out pretty well.

Virgin average late with 5.17 av speed 79.6
LNER                               4.25 av speed 85.07  

East Midlands, and Grand Central were on time but av speed were quite low.

Trans Pennine, and Northern  despite their reputation didn't do too badly with TP early

I think what it illustrates is the randomness of late running you pays your money and hope the trains on time.

Have you included any cancellations? In my opinion They are worse than late running.

A reliable railway is a punctual railway.

The average speed figure for GWR was 62.72 not 52.72. However, this has now been amended to 63.14 because I checked the data again before I offered to share the spreadsheet with others and I found a mistake on the first day (showing timings from Bristol to Chippenham against a train from I used from Swindon to Chippenham).

LNER and Virgin both came out with high average speeds because of the nature of the journeys I took with them (Leeds to KGX, KGX to Darlington, Preston/ Wigan/ Lancaster to Euston)

As regards cancellations, there were two and a half! A train I had planned to use between Bradford and Ilkley was cancelled (so I didn't go), and XC^▸ (Cross Country Trains (franchise)) summarily cancelled a Newcastle-bound train at Birmingham on my last day (this was the day the weather broke, the OHLE was down north of York and I suspect that the crew booked to take it north from Brum were stranded somewhere, and there was also no guarantee that the train would get anywhere near its destination anyway.

The "half a cancellation" was due to a complete signal failure at Wigan Wallgate. I was at Manchester Victoria wanting to go to Rose Grove on a Southport to Blackburn via Manchester service that was stuck the wrong side of Wigan. However. local staff at Manchester conjured up a 2-car class 144 from somewhere and ran the Manchester to Blackburn section with that. This did rather confuse the automated train indicator information for passengers, as it was alternating between "on time" and cancelled" every minute or so!