Numbers & stats

[MetBlue]

With lots of carp on television the last couple of days (touch of lysdexia there), I thought I'd do something more interesting and take a look around the "how Many Left" data on line for our cars, particular any trends since SORN records started being available around 2007.
No idea how to post anything other than Jpeg files, so ended up photographing my monitor to get the pictures below, but they seem readable.







The obvious and not surprising first observation is that number recorded generally increases year on year. Explained I guess by anyone buying a barn find having to put it on SORN as part of transferring the ownership in todays "we're watching you " world. At least Elites and Eclats do - Excels are pretty steady. Could that be that Excels are just too new still ???
It surprised me just how many Eclats are still known to the DVLA. In percentage terms, twice as many as either Elites or Excels.!
It also surprised me that Excels are least represented of the 3 models, both by known numbers and % on the road. !! Look at that trend!
And that only the Elite has an upwards trend regarding number on the road.

Tony

Disclaimer : Like all good statistics, I'll accept now that the data is not 100% factual, but it's supposed to be based on Gov data, and it's the same basic source for all three models, so they should all be wrong in the same way.

[Mr Locust]

It surprised me just how many Eclats are still known to the DVLA. In percentage terms, twice as many as either Elites or Excels.!
It also surprised me that Excels are least represented of the 3 models, both by known numbers and % on the road. !! Look at that trend!

It could be explained by how many Excel are registered as Eclat with DVLA.

[Lotus-e-Clan]

On of the issues I have with Excel (spreadsheets!) charts and graphs is that the axis data range is set to 'dynamically auto-scale axis' by default. Which is OK-ish for a single standalone graph but not ideal when you want to compare rates of change in data between multiple charts/graphs.

If you manually set the y axis (car numbers) to the same data range (eg 0 to 700) for every graph then the rate of change ( line gradient) for car numbers 'sorned' or 'taxed' is easier to compare between car models. My kids would tell you I'm either a PITA or I'm clearly on the autistic spectrum for pointing that out...

[hodders]

On of the issues I have with Excel (spreadsheets!) charts and graphs is that the axis data range is set to 'dynamically auto-scale axis' by default. Which is OK-ish for a single standalone graph but not ideal when you want to compare rates of change in data between multiple charts/graphs.

If you manually set the y axis (car numbers) to the same data range (eg 0 to 700) for every graph then the rate of change ( line gradient) for car numbers 'sorned' or 'taxed' is easier to compare between car models. My kids would tell you I'm either a PITA or I'm clearly on the autistic spectrum for pointing that out...

Absolutely with you on that. My geography & maths teachers drummed that into me.

[richardw]

On of the issues I have with Excel (spreadsheets!) charts and graphs is that the axis data range is set to 'dynamically auto-scale axis' by default. Which is OK-ish for a single standalone graph but not ideal when you want to compare rates of change in data between multiple charts/graphs.

If you manually set the y axis (car numbers) to the same data range (eg 0 to 700) for every graph then the rate of change ( line gradient) for car numbers 'sorned' or 'taxed' is easier to compare between car models. My kids would tell you I'm either a PITA or I'm clearly on the autistic spectrum for pointing that out...

Absolutely with you on that. My geography & maths teachers drummed that into me.

Yes, I’m often horrified at the abuse of statistics we see from government and media these days involving exactly this kind of misrepresentation. Even some academics are just as guilty!

Richard

[Lotus-e-Clan]

On of the issues I have with Excel (spreadsheets!) charts and graphs is that the axis data range is set to 'dynamically auto-scale axis' by default. Which is OK-ish for a single standalone graph but not ideal when you want to compare rates of change in data between multiple charts/graphs.

If you manually set the y axis (car numbers) to the same data range (eg 0 to 700) for every graph then the rate of change ( line gradient) for car numbers 'sorned' or 'taxed' is easier to compare between car models. My kids would tell you I'm either a PITA or I'm clearly on the autistic spectrum for pointing that out...

Absolutely with you on that. My geography & maths teachers drummed that into me.

Yes, I’m often horrified at the abuse of statistics we see from government and media these days involving exactly this kind of misrepresentation. Even some academics are just as guilty!

Richard

You're spot-on.

I spent the last twenty years teaching biomed @ Durham University. Academics deliberately exaggerate scales to show small changes within data for impact. Sometimes this is appropriate if the experimental model is robustly designed. Climate scientists wanting to demonstrate significant trends within 1 deg C data ranges are a good example. But they sometimes mislead. Almost all undergrad dissertations contain this error. Post grads weren't much better either.

And then there's Chris Witty et.al.....

[MetBlue]

I've set off a right discussion here about graphical representations. Must say I likewise get very annoyed when trends get exaggerated by changing the scale to try and prove a point ( Climate scientists and 1 degree example above the best example). I've even seen examples where even the numbers get left off the vertical scale !!

Anyway, back to Excel numbers....

It could be explained by how many Excel are registered as Eclat with DVLA.

.
Good Point Ian. I looked again and there are 283 "eclats" registered after 1983 !!!- So I've redone the graphs with those moved to Excels.
Also looked at the Elite data in a bit more detail and there are 101 models included manufactured in early 1960's. - The original Elites have crept in !!, so again, redone the graphs without those in.

- And for the record, I set all the scales to the same range






It would seem Elites are gradually returning to the road, but off set by fewer Eclats and Excels on the road each summer.

Time to get those restoration projects re-invigorated. - It'll soon be the 50 year anniversary. - What a taget.

[Pete Boole]

Lies, damned lies, and statistics! That what they say! Interesting data though Tony.
Pete

[bash]

If the DvLA are so bad at adding up its a good job they dont do 'times bys' and 'guzinters'.
Bash
(BA hons. (Failed) university of life).

Ps, Im surprised that there are so many excels taxed seeing as there seem to be so few of us on here or on the facebook judean front page.

[richardw]

Time to get those restoration projects re-invigorated. - It'll soon be the 50 year anniversary. - What a taget.

Excel is 40 in 2022!

I have to say that your numbers reflect my impression that Eclats are the rarest of the breed. Excels are used more regularly than the earlier cars, and are seen more regularly, and Elites are hidden away somewhere. At Oliver’s funeral, all but one of the wedges were Excels.

Cheers Richard

[Zag]

You should really also adjust the number produced to include the transferred eclats to excels assuming they have always been registered "incorrectly".

This comes from someone who may have previously used a log scale on a graph to hide an upwards trend!

[Atking53]

Thanks for the interesting data. Cant beat a good XL graph. Some of the cars may have drifted overseas in that time. I know mine did in 2012. As export regulations get tighter though due to asbestos etc this might slow.

[Phantasm]

Yup, I know of 8 excels that have now skipped town to the u.s. that were mostly if not all u.k. cars.