Buses, trains and time: what’s really happening

The latest report on public transport performance, compiled from Metlink’s own (scarce) data by stalwart transport nerd Mike Mellor. Data to 10 March.

Editor’s note

As Mellor finds in his report, sent to all the relevant bigwigs on 14th March, the week to 10th March wasn’t any better for public transport performance in Te Upoko o te Ika. There’s been some more data released – a little more on patronage – but overall the picture is as grim as it was last week.

A few key points:

Both [bus] operators continue to perform much worse in Wellington than they do in Auckland. Their performance there is much better than in Wellington, consistently exceeding their much more demanding target.

Train punctuality: continues poor. Train reliability: a mystery, despite much public concern.

We’re still noting – (along with well-regarded local commentators like Dave Armstrong, and our own commenters) – a mismatch between public statements of how things are OK or getting better, and the actual data. They want to know your thoughts, though! Send your feedback.

Now for the data.

Metlink performance and patronage to 10/3/2019

1. Reliability of top 10 bus routes

1.1 Average reliability for both operators worsened, well below target and at equal-worst level since data has been published. All improvement initiatives have been ineffectual.

1.2 Tranzurban continues to outperform NZ Bus.

2. Punctuality of top 10 bus routes

2.1 Average punctuality rose, but improvement in the period that data has been published has been marginal.

2.2 Both operators continue to perform much worse in Wellington than they do in Auckland. Their performance there is much better than in Wellington, consistently exceeding their much more demanding target.

2.3 As the graph shows, there is no indication of any overall improvement in Wellington – all initiatives so far have been ineffectual.

3. Train punctuality

3.1 Train punctuality continues to be very poor.

3.2 Despite much public concern over train reliability, no relevant data has been published.

4. Patronage

New data is available!  We now have some for… December 2018.

5. Bus data

5.1 Notes on bus data:

a) italics on grey background indicate where the published data is deficient:

  • for Wed 19/12/18 and Thu 3/1/19 there is no punctuality data for routes 1, 2, 3, 7, 14, 22 or 24
  • for Wed 26/12/18 there is no data of any sort for any of the NZ Bus routes
  • for Thu 27/12/18 there is no reliability data for any of the routes
  • These deficiencies make any derived figures such as averages of dubious value, and distort trends, so they are not charted

b) italics on coloured background indicate where the data has been affected by NZ Bus industrial issues. Because these are outside of and do not affect the general trend, they have not been charted.

c) the absence of these figures from datasets, not published until over 8 weeks after the event and with no comment about these deficiencies, raises serious questions about the integrity of the data and fragility of the data-collection process.

d) such data irregularities are of financial concern because of the potential for missed contractual penalties for which bus operators may be liable, and also because of the revenue implications of irregularities in Snapper data. These issues need to be addressed.

6. General notes

a) the above measures are of contractual performance, and from a passenger perspective they are necessary but very far from being sufficient.

In particular, the punctuality figures give no indication of how late buses are running en route (from observation, bus bunching is still common), nor the effect on scheduled connections.

A measure such as Excess Waiting Time, calculated from RTI/Snapper data, is essential for that, and we don’t have it.

b) the format and presentation of the data are not user friendly, making it quite hard to analyse the data and identify the essential information about trends and comparisons. GWRC adopting Open Data policies (as is common amongst transit agencies) would make such analysis much easier.

c) the performance of routes 1, 2, 7 and 22 is particularly important because they are scheduled to make connections, meaning that the effects of poor performance extend across the wider network. (Routes 21 and 30x are also important in this respect.)

Report prepared by Mike Mellor, 14/3/19


Further reading:

Mike’s previous report, published on Talk Wellington with extra commentary

Dave Armstrong’s latest commentary.

A final point from Talk Wellington

Some people have criticised us for describing people like Mike as geeks and nerds.  One person felt we were perpetuating cruel stereotypes such as anorak-wearing.

Please note: we use these terms with deep respect and appreciation for their work, and the crucial role that public nerdery and geekery plays in keeping us laypeople informed.

All hail, transport nerds and geeks!

I guess that makes us geeks, nerds and dorks!

Image credits:

2 comments on “Buses, trains and time: what’s really happening”

  • Tim Jones says:

    First of all, I want to commend Mike for the outstanding job he is doing with this analysis: patient hard work that is removing the figleaf of bluster that GWRC officials have been hiding behind (if you’ll excuse the metaphor).

    Second, I personally do not like the use of terms such as “nerds” and “geeks” to describe people with the ability, persistence and determination to analyse data and write it up in such a useful fashion.

    But third, the first thing I did when I saw the cartoon that concludes this post was to check that the the cartoon dog had correctly included the colon in the title of “Magic: The Gathering”.

  • Fred Albert says:

    This is great work you are doing. One thing that the data doesn’t tell us is the availability of buses by time of day. I am a 14 and 20 bus user and certainly, in the past week, wanting to take a bus into town between 7am and 9am isn’t so much about punctuality but about what buses are actually showing up during that time. I would guess that going from Kilbirnie into the CBD possibly half the buses are cancelled during peak time morning (some ‘planned’ but others on the day), particularly the 14.

Leave a comment

Your email address will not be published. Required fields are marked *