Irish Rail's latest reliability report

maggy_thatcher

I noticed that Irish Rail have put up their latest punctuality report up at the stations -- and one part confused me:

Reliability: 99.32%

This was for the last week in January + the first 3 weeks in February.

Correct me if I'm wrong, but for most of that period, they didn't run any trains to Greystones...does that mean that only 0.68% of trains ever actually goes there, or are they fudging the figures to exclude them? The passenger charter says that they'll exclude "conditions outside their control" from their punctuality report, but it doesn't say anything about their reliability report.

TradeSecrets

Sure when you look at the small print they say that a train that is 15 minutes late is still regarded as on time. You can take what you want from that !! Smudgey Smudge!!

KC61

maggy_thatcher wrote: »

I noticed that Irish Rail have put up their latest punctuality report up at the stations -- and one part confused me:

Reliability: 99.32%

This was for the last week in January + the first 3 weeks in February.

Correct me if I'm wrong, but for most of that period, they didn't run any trains to Greystones...does that mean that only 0.68% of trains ever actually goes there, or are they fudging the figures to exclude them? The passenger charter says that they'll exclude "conditions outside their control" from their punctuality report, but it doesn't say anything about their reliability report.

The trains did operate but operated only to Bray rather than Greystones.

Cookie_Monster

KC61 wrote: »

The trains did operate but operated only to Bray rather than Greystones.

That's hardly ok though:rolleyes:. what if the Galway train only ran as far as Athlone for a few week, would you be as dismissive about that.

There was no part of that closure outside their control either IMO. The landslides had no impact on the track itself and this was closed purely as a precaution. They made an obvious decision despite minimal risk to suggest it was nessicary.

maggy_thatcher

KC61 wrote: »

The trains did operate but operated only to Bray rather than Greystones.

I would interpret "operate as planned" as being "operated according to what the timetable says". That being the case, they didn't operate as planned, because they didn't quite make it to the finish line. Given they also cheat when giving punctuality reports (only show terminus-terminus, rather than showing how punctual they are from station-station), they effectively weren't punctual either.
At the very least, each individual train was only 96% reliable...

KC61

Cookie_Monster wrote: »

That's hardly ok though:rolleyes:. what if the Galway train only ran as far as Athlone for a few week, would you be as dismissive about that.

There was no part of that closure outside their control either IMO. The landslides had no impact on the track itself and this was closed purely as a precaution. They made an obvious decision despite minimal risk to suggest it was nessicary.

I am not being dismissive about anything here, merely trying to explain how the definition of "reliability" is (in my opinion) arrived at, even though it may seem unfair and daft.

I would interpret "reliability" meaning as whether the service operated or not, and I would certainly guess that if a bus is provided to replace part of the journey that it is included as having operated. Reliability is (I suspect) based on whether the service operated. Where a service is completely cancelled then it would be a negative.

In the case of Bray Head, this was absolutely out of the control of the operations department, as the civil engineers made a decision that the risk of a rockfall onto the track or the trains was too great. Now they are the professionals who made that decision in full possession of all the relevant information, not people on an internet messageboard, and as such I would consider that they made an informed decision. If I were an operations manager I certainly would not risk operating trains in that environment.

Therefore, I would say that it would have been the services cancelled such as the 1640 to Enniscorthy and 1923 return for which no bus was provided that would have been have been a negative feature of reliability statistics.

I know it appears daft, but those DART services were curtailed to Bray, due to the line closure at Bray Head and as they did operate as per the revised schedule, they would (I suspect) be included as "reliable". Similarly the Rosslare services did all operate, albeit with bus replacements for part of the journey.

As with all statistics there are different interpretations, but I suspect that is the one that Irish Rail are employing here.

Riskymove

on one level I know what you mean but then again a landslide is a landslide and does not really reflect the "reliability" of the service

maggy_thatcher

That's why I believe they should use inter-station results, not cheat and use terminus-terminus results (and then change the terminus when it suits them). The Greystones<->Bray DART was cancelled. The Bray<->Shankill DART ran, and so on. Similarly on whatever day it was they had to shutdown the inner-city stations -- the Howth<->Connolly service ran, and the Bray<->Grand Canal service ran, but the Grand Canal<->Pearse, or Pearse<->Tara St. services didn't.

But I suppose there's lies, damn lies, and statistics

maggy_thatcher

Riskymove wrote: »

on one level I know what you mean but then again a landslide is a landslide and does not really reflect the "reliability" of the service

On the contrary...external factors should always be taken into account. It would prompt somebody to research -- why did the reliability of the service drop so far that month, see that it's due to a landslide -- and perhaps do some preemptive work to reduce the likelihood of it happening again...

KC61

maggy_thatcher wrote: »

That's why I believe they should use inter-station results, not cheat and use terminus-terminus results (and then change the terminus when it suits them). The Greystones<->Bray DART was cancelled. The Bray<->Shankill DART ran, and so on. Similarly on whatever day it was they had to shutdown the inner-city stations -- the Howth<->Connolly service ran, and the Bray<->Grand Canal service ran, but the Grand Canal<->Pearse, or Pearse<->Tara St. services didn't.

But I suppose there's lies, damn lies, and statistics

Indeed there are.

Unfortunately with something like this you could go into reams of detail. What appears to be taken is the broad statistic of whether a service operated or not. The DART services did operate in all the above cases, but were curtailed to Bray due to circumstances outside the control of IE.

maggy_thatcher

KC61 wrote: »

Indeed there are.

Unfortunately with something like this you could go into reams of detail. What appears to be taken is the broad statistic of whether a service operated or not. The DART services did operate in all the above cases, but were curtailed to Bray due to circumstances outside the control of IE.

Tell that to the people in Greystones -- from their point of view, the service did not operate. And their charter said that they'd exclude circumstances beyond their control from punctuality, but not from reliability. In my view, each individual service was 96% reliable. That seems like a reasonable compromise.

Also, who's responsible for maintaining the area to prevent landslides in the first place?

KC61

maggy_thatcher wrote: »

Tell that to the people in Greystones -- from their point of view, the service did not operate. And their charter said that they'd exclude circumstances beyond their control from punctuality, but not from reliability. In my view, each individual service was 96% reliable. That seems like a reasonable compromise.

Also, who's responsible for maintaining the area to prevent landslides in the first place?

I know - I didn't write these things - I'm only trying to interpret how they work.

I think you could argue about the landslide till the cows come home - personally I really don't think that it could be argued that it was within Irish Rail's control.

MOH

KC61 wrote: »

Indeed there are.

Unfortunately with something like this you could go into reams of detail. What appears to be taken is the broad statistic of whether a service operated or not. The DART services did operate in all the above cases, but were curtailed to Bray due to circumstances outside the control of IE.

The explanation given on the poster is the the reliability reflects the services that "operated as planned". The services were planned to go to Greystones, but didn't. Therefore the statistics, as presented, are not consistent with the stated criteria.

Whether the cause of the failure was within Irish Rail's control or not is irrelevant. If they want to state that the statistics show the % of services that "operate as planned, or failed due to circumstances outside our control", then that's fine - but they haven't.

As is, the reliability figure presented is false.

Carawaystick

What about the closure between Arklow and Gorey? That was a part of the embankment the line ran on. Surely that was entirely within Irish rail's fault for not maintaining the way correctly?
If Irish rail have small print saying 15 mins late is on time but don't put small print saying some get out about the line closure then the ad is misleading.

Can someone post a phone pic of the ad so we can see if it is actually misleading?

dowlingm

IE should not be allowed use these kinds of statistics in promotional materials because they have no context. I don't think there's a lot of difference between the posters above, I think we'd all agree that there is at least two and probably more criteria to benchmark this stuff on - one being adherence to timetable (similar to slot usage at airports) and one being adherence to schedule (which would be the first one but with credit for force majeure, maintenance etc.)

maggy_thatcher

Carawaystick wrote: »

Can someone post a phone pic of the ad so we can see if it is actually misleading?

Here's the ad (from Clontarf Road Station)

burgess1

The timetables on the website (where you enter your origin and destination and they present a list) gave exact times of shuttle buses between Greystones and Bray when the Dart wasn't operating. Then, at the station, you were told that outside the weekday rush hour, you were to use the normal 84 bus and that relying on the website was a terrible idea.

Not ideal when someone from, say, Kilcoole needs to get to Malahide on Sunday evening.

Carawaystick

Didn't the line close during the snow when roads were almost impassable and chronic tailbacks occurred?

P_1

So IE's definition of "late" is 15 minutes?
I'd love to be able to ue that in work.

maggy_thatcher

On a related note -- this mornings 7:59 service from Howth to Bray was 6 carriages instead of the usual 8...personally I think it should only count as being 75% reliable for that (only 3/4 of it showed up)...but I bet IE will say "it showed up, so it counts", despite the fact that I'm sure some people couldn't make it on board...

laoisfan

dowlingm wrote: »

IE should not be allowed use these kinds of statistics in promotional materials because they have no context. I don't think there's a lot of difference between the posters above, I think we'd all agree that there is at least two and probably more criteria to benchmark this stuff on - one being adherence to timetable (similar to slot usage at airports) and one being adherence to schedule (which would be the first one but with credit for force majeure, maintenance etc.)

agreed. also....i believe (but am open to correction) that irish rail produce the stats themselves and there is no independent verification of the data.

have been traveling for year on irish rail (from Laois to Dublin) and would take these stats with a pinch of salt.

MOH

laoisfan wrote: »

agreed. also....i believe (but am open to correction) that irish rail produce the stats themselves and there is no independent verification of the data.

have been traveling for year on irish rail (from Laois to Dublin) and would take these stats with a pinch of salt.

The stats are supposedly independently verified.

kingshankly

maggy_thatcher wrote: »

On a related note -- this mornings 7:59 service from Howth to Bray was 6 carriages instead of the usual 8...personally I think it should only count as being 75% reliable for that (only 3/4 of it showed up)...but I bet IE will say "it showed up, so it counts", despite the fact that I'm sure some people couldn't make it on board...

better to leave two behind and go with six than cancel the train:) i was amazed to see their reliability for the belfast service ,its forever causing problems.were not there yet but were getting there:rolleyes:

maggy_thatcher

kingshankly wrote: »

better to leave two behind and go with six than cancel the train:) i was amazed to see their reliability for the belfast service ,its forever causing problems.were not there yet but were getting there:rolleyes:

Oh, I agree it's better the service ran in a restricted capacity than not at all, but I guess a single numeric statistic on how well the service operates doesn't really show the whole story.

KC61

kingshankly wrote: »

better to leave two behind and go with six than cancel the train:) i was amazed to see their reliability for the belfast service ,its forever causing problems.were not there yet but were getting there:rolleyes:

Indeed but again it boils back to did the service operate or not? And what is the definition of "as planned"?

I suspect that even if the Enterprise is replaced by a railcar, then IE deem it to have operated, and as such is included in the reliability figure, because the service did operate.

KC61

maggy_thatcher wrote: »

Oh, I agree it's better the service ran in a restricted capacity than not at all, but I guess a single numeric statistic on how well the service operates doesn't really show the whole story.

Exactly!

You would need very narrow definitions in order to get meaningful information, certainly far narrower than the statistics being offered at present.

maggy_thatcher

Is there any way to get the raw data that these statistics are built from? The only way I can see to do it is to screen-scrape the irish rail "next train due" page (which would probably violate their copyright). That way, we could get a more accurate picture of what the reliability / punctuality statistics are. Getting the information on what size a particular train is, and how long it is supposed to be would be a lot harder though, but at least getting the raw time information would be a start.

MOH

maggy_thatcher wrote: »

Is there any way to get the raw data that these statistics are built from? The only way I can see to do it is to screen-scrape the irish rail "next train due" page (which would probably violate their copyright). That way, we could get a more accurate picture of what the reliability / punctuality statistics are. Getting the information on what size a particular train is, and how long it is supposed to be would be a lot harder though, but at least getting the raw time information would be a start.

The "real-time" page doesn't work very well, anyway:
- the "current" time on it can be wrong (I've seen it at least 3 minutes out);
- trains no longer appear on it after their due time has passed (so a train due at 17:00 running 5 minutes late won't be listed at all once 17:00 has passed);
- trains which haven't started yet are always shown as on time (so even if the entire line is shut down, a train starting from Dun Laoghaire due in Blackrock in 15 minutes will show as 'ETA 15 minutes' when there's no way it will be;
- it can't deal with major delays at all without manual intervention - the last time a train failure caused a 30 minute backlog, it was still reporting everything as being normal almost 30 minutes later.

MOH

Saw the latest DART performance figures yesterday (22 Feb - 21 March, or something like that). Punctuality was down from the usual claimed 98% to 90.7%.

There's a note underneath explaining that due to customer feedback, they're now measuring punctual as less than five minutes late (instead of 10), and thanking customers for their input.

While this is a step in the right direction, it's still a bit daft. Customers complain that DART punctuality figures allow a 10 minutes window, instead of just measuring punctuality. IE response: OK, we'll use a five minute window instead.
Why not just give us the actual punctuality figures?

I also noted this morning they're back to their old trick for improving punctuality - leave every station early. Helps cut down on the number of pesky passengers who delay the train by getting on and off, since more of them miss it.