NCT test results dashboard for you to explore

Reesy

Hi,

My company's created an analytics dashboard (that's a report that you can adjust yourself to dig into the data) of last year's NCT test results. You can use it to see which makes & models of car (and of what age) failed the NCT most, and what category they failed on.

We think it'll help people choose a used car. Hope you find it interesting - we'd love to hear your feedback.

The dashboard: https://public.tableau.com/profile/idiro.analytics#!/vizhome/NCT2016Top20MakesResults/NCT2016-20MostPopularMakes

Instructions on how to use it: http://idiro.com/2017/08/idiro-analysis-2016-nct-results-helps-car-buyers-choose-wisely/

It's covered in today's Indo: http://www.independent.ie/life/motoring/car-reviews/put-your-car-to-the-test-36049372.html

ht9zni1gs28crp

How accurate is the information? Especially model types and the quantity of testing?

Two cars of mine tested in 2016 dont show up in that data?

grogi

Kolten Lemon Teacup wrote: »

How accurate is the information? Especially model types and the quantity of testing?

Two cars of mine tested in 2016 dont show up in that data?

This dashboard displays a filtered version of the original data set, (NCT tests conducted in 2016, available from http://www.rsa.ie/) showing the 20 most popular makes of vehicle in the dataset, and for each of these, the models with at least 1000 tests carried out in 2016. The dataset covers the results of first tests only, and does not take into account mileage, maintenance or service history, and therefore may not directly evaluate vehicle quality.

Ubbquittious

The Nissan Puke is helped along a good bit by only having been in production since 2010 so it's quite meaningless unless you're only comparing cars of the same year. Also if cars are failing for consumable items like tyres and bulbs that's more a reflection on the owners of those cars

It wouls be nice to have an equivalent of http://howmanyleft.co.uk but you'll need contacts in the secret car registration inner circle and a hefty wad of cash to glean that info.

grogi

Ubbquittious wrote: »

The Nissan Puke is helped along a good bit by only having been in production since 2010 so it's quite meaningless unless you're only comparing cars of the same year. Also if cars are failing for consumable items like tyres and bulbs that's more a reflection on the owners of those cars

But that's still a good indicator of reliability and passability in future. A neglected car will fail more often...

zilog_jones

Ubbquittious wrote: »

The Nissan Puke is helped along a good bit by only having been in production since 2010 so it's quite meaningless unless you're only comparing cars of the same year. Also if cars are failing for consumable items like tyres and bulbs that's more a reflection on the owners of those cars

Yes, unfortunately it seems the pass rates are more or less directly proportional to the age profile of the model. Look at the Honda Civic or Toyota Corolla/Avensis, for example: They have unusually low pass rates despite being current models, but the majority of the cars doing the test are 2008 or older. If you search only 2009-2016 cars you get significantly different results.

Reesy

Agreed. Older cars are indeed more likely to fail - that's why we put the age selector slider at the top!

However, we think that if you're looking to buy a used car, this will be a handy tool for checking out cars before you go see them - to see how reliable the model (of that age) is, and to understand what it's most likely to fail on.

As for the 270k tests that we excluded, we had to cut it down otherwise it'd have been quite unwieldy. Also, there's a lot of bad data in there, where the model has just been input wrong (or at least inconsistently). However if you're interested in a car that's not on the list, you can download the source data from the RSA here http://www.rsa.ie/Documents/NCT/Make%20Model%20Data%202016.csv and filter / search the makes and models that interest you.

RandomAccess

The independent have done an article on this but they made a mess out of it. Despite mentioning that some models have been around far longer than others they still draw the headline grabbing conclusions they wanted.

Bloody media today.

zilog_jones

Did you expect any more from the Indo?

Avatar MIA

zilog_jones wrote: »

Yes, unfortunately it seems the pass rates are more or less directly proportional to the age profile of the model.

The Fiat Panda seems to be doing very well. And that was last produced in 2012.

RandomAccess

Avatar MIA wrote: »

The Fiat Panda seems to be doing very well. And that was last produced in 2012.

The Pandas I've seen are usually fairly clean and looked after. I've noticed that with certain models, they tend to be looked after more than others.

Avatar MIA

RandomAccess wrote: »

The Pandas I've seen are usually fairly clean and looked after. I've noticed that with certain models, they tend to be looked after more than others.

Remember fondly my sister's old (model) Panda Surprised (not sure why) to see them being so reliable.

edburg

Only problem I find the list and regarding 2nd hand car market and whats reliable is the jelly belly who just doesn't take care of the car or botches jobs to save 10 cents or just knows best because all garages are out to con you.

chicorytip

Avatar MIA wrote:

The Fiat Panda seems to be doing very well. And that was last produced in 2012.

No. The Fiat Panda is still in production today.

Avatar MIA

chicorytip wrote: »

No. The Fiat Panda is still in production today.

Ah, okay

Reesy

grogi wrote: »

This dashboard displays a filtered version of the original data set, (NCT tests conducted in 2016, available from http://www.rsa.ie/) showing the 20 most popular makes of vehicle in the dataset, and for each of these, the models with at least 1000 tests carried out in 2016. The dataset covers the results of first tests only, and does not take into account mileage, maintenance or service history, and therefore may not directly evaluate vehicle quality.

Incidentally, the lad who created the dashboard gave me a look at the data that we've had to exclude from the report. Seems lots of cars tested last year were had 'other' as either their make or model, and some were completely misclassified. For example the MX5 was represented in about 5 different ways, and no logical split. So sorry about that, but some of the data's of really low quality.

You can look at the source data on RSA.ie if you fancy digging into it.

Reesy

Hi all,

We've updated the NCT dashboard with the 2017 test results - here: https://public.tableau.com/profile/idiro.analytics#!/vizhome/NCT2017Top20Makes/NCT2017-20MostPopularMakes

And also, we created a simple utility to allow you to look up the detailed NCT results for any make / model / year - you'll find it at https://idiro.com/NCTchecker.

Detailed instructions for both are here: http://idiro.com/2018/08/will-my-car-pass-the-nct/

Feedback welcome.

i57dwun4yb1pt8

why would a 2016 or 2017 car show up
surely youngest would be 2013 for the 2017 results year ( 4 years from new ? )

beauf

DaDumTish wrote: »

why would a 2016 or 2017 car show up
surely youngest would be 2013 for the 2017 results year ( 4 years from new ? )

Bad data input I'm guessing...

colm_mcm

DaDumTish wrote: »

why would a 2016 or 2017 car show up
surely youngest would be 2013 for the 2017 results year ( 4 years from new ? )

Taxi?

beauf

I would hope they don't mix commercial with domestic. That would make a farce of the data.

colm_mcm

beauf wrote: »

I would hope they don't mix commercial with domestic. That would make a farce of the data.

You mean like PSVs?

beauf

I mean if you have a data set, taxi's or hackneys etc, or work vans, shouldn't be mixed with private use only vehicles as much as possible. As it would distort the data and any analysis of it. You should be able to seperate them.

Same with years age of vehicle. The results should be averaged per age of the vehicle. Not mixed across different ages.

Reesy

DaDumTish wrote: »

why would a 2016 or 2017 car show up
surely youngest would be 2013 for the 2017 results year ( 4 years from new ? )

It's a good question. We spotted that and we guess that you can get a car of any age tested, but why you would is a different question. I'll see if we can find out from the RSA (it may take a while).

cml387

UK imports?

zilog_jones

UK imports don't need an NCT until they're 4 years old either. And the MOT is now transferable.

barneygumble99

If you buy a brand new taxi with 0km it’s must still be tested before it gets a licence.