Conflicting train cancellation data

[mrmartin]

I was caught up in the storm yesterday and the abysmal state of cancellation data did not help matters.

RTT seems to be the worst for showing actual cancellations, I assume this is a different data source
GWR app was showing as trains running (then the 2248 "vanished", which is when I knew something was wrong). The 2148 was showing as running though, which it wasn't. Then it showed a 2131(ish) train which didn't exist, but no cancelled trains
NR app did seem to show the correct data

It's highly confusing in an often time constrained environment when you are rushing to get an earlier train than you expected to!

Any idea what the best way to look at cancellation data like this is?

 

[30907]

I assume you were travelling from Paddington to somewhere beyond Swindon, as the 2131 is a Cheltenham?

It is always useful to look at what is happening on the way in - for example, an up train running 90 late (the 1823 from Swansea yesterday) might well cause problems for the return trip (the 2148, which started from Reading having (presumably) connected out of the 2153 down.

 

[Tom]

When everything goes to pot, the accuracy of any data source becomes questionable. NR and GWR will use the same data source (Darwin), RTT uses a different one (TRUST)- both have data input by different sources, the TOC and NR respectively. In theory Darwin should be more accurate than TRUST due to the latters limitations but that's not always the case.

I got caught up in major disruption last week on SWR and found similarly that no system was accurate, yet TRUST had some trains as cancelled that Darwin said was running and v.v. The actual answer is normally somewhere between the two. Staff were relying on whatever Darwin were showing and giving out that information as gospel. Ultimately my general feeling is that you can't rely on anything when everything goes wrong because controls are not staffed for major disruption rather just minor perturbation and get rapidly overwhelmed.

Agreed with what @30907 has said above, have a look at the inbound workings if you can work them out and also whether they have stock allocated as well. For some reason a lot of TOCs pull stock off trains before they cancel them.