Unscientific, not guaranteed to be representative, age versus modelled era poll

[BMacdermott]

In terms of our individual modelling the era system is not really that relevant.  However as Brian says from a retailer perspective and also for the purposes of doing a survey such as this it has benefits.

 

All in all, and despite the obvious shortcomings, eras do provide a method of grouping our interests in time periods that we do understand, even if we don't agree with them.

 

Hello Andy

 

If the era system was 'valid', we could apply it to, say, the Somerset & Dorset Line. The problem is...we can't. 

 

If someone says they model the S&D in Era 5, it would give me some idea - but I haven't seen anyone model the line fully as running between 1957 and 1966. However, if someone says they model the S&D in 1961, I and many others would have a pretty accurate view of what they are doing (albeit given that there are 365 days in the year and the line was different in the winter compared to summer).

 

Brian

[Ravenser]

 

Hence I didn't accept there may be statisically significant age groups not represented by the rm poll. There is no physical evidence I see to suggest it's not representative . At warley it's clear the hobby is dominated by the same age group as evident by the poll.

 

Er no. At Warley and quite a few other shows I often feel I'm a youngster - the bulk of the crowd seem to be 15-30 years older than me. In this poll I'm right in the biggest cohort, and there are large cohorts immediately younger than me.

[Miss Prism]

Absolutely - but quite probably re-liveried to a later (but not necessarily the xx to yy) period.

 

The converse is also unfortunately typical, namely that a product is given an early livery inappropriate to its build state.

Edited January 20, 2018 by Miss Prism

[Colin_McLeod]

There are preserved coaches with such liveries.

[newbryford]

There are preserved coaches with such liveries.

So they could be modern image........?

 

Cheers

Mick

Edited January 20, 2018 by newbryford

[Ravenser]

A slightly alternative take on the "era" data

 

-The Grouping Era is 25 years.

- BR steam - that is early and late crest is 20-21 years (depending on how much steam you feel was hanging on in the early months of 1968

- post-steam BR is therefore 1968- mid 90s (I think the last BR train ran in 1997) - perhaps 27 years

- Overwhelmingly people model pre-Grouping railways from 1900 onward - Victorian layouts are pretty uncommon. One might say that the pre-Grouping era is also 25 years (1898-1922)

- The post-privatisation railway has been with us since about 1995 - say 23 years.

 

All these periods are roughly the same length - a quarter of a century . BR diesel and electric is a little longer than that, BR steam a little shorter. 

 

This means we can work with reasonably equivalent periods of time - which  makes comparing the levels of support much easier

 

* For BR steam, at time of writing, the combined total support is 262 votes

* For BR diesel & electric, at time of writing the combined support is 201 votes

* For post-privatisation, at time of writing the combined support is 65 votes

 

(there is a possibility that some people have voted for both crest eras, or Blue + Sector, but when comparing BR steam support and BR diesel /electric support , that error basically cancels out)

 

The two BR super-eras tower above everything else.

 

Strikingly , there are more votes for BR diesel and electric than for everything before 1948 combined (9+70+107 votes = 186 votes). This isn't really the impression you get from the RTR catalogues or the magazines, but it suggests that the gripes about lack of Blue 31s or Sector liveries might have some substance

 

Support for post-privatisation is about a third of that - which may suggest you shouldn't hold your breath for the early appearance of the new Bachmann 170

 

Support for BR steam now towers above that for the Grouping, with nearly 3 times as many votes. That was most emphatically not the case in the later 70s /early 80s , when BR steam was a dowdy niche

 

And pre-Grouping now runs Grouping surprisingly close. But -

 

- Post-privatisation votes (65) are slightly behind the pre-Grouping vote (70)

 

The two periods are very different modelling propositions, but the size of their constituencies are fairly similar

Edited January 20, 2018 by Ravenser

[Andy Hayter]

Hello Andy

 

If the era system was 'valid', we could apply it to, say, the Somerset & Dorset Line. The problem is...we can't. 

 

If someone says they model the S&D in Era 5, it would give me some idea - but I haven't seen anyone model the line fully as running between 1957 and 1966. However, if someone says they model the S&D in 1961, I and many others would have a pretty accurate view of what they are doing (albeit given that there are 365 days in the year and the line was different in the winter compared to summer).

 

Brian

 

Brian 

 

I agree, if you read my earlier post I point out how 1923 has no relevance to a number of companies.  There will always be exceptions where the general rule does not apply*.  You will find it not only regarding the S&D, but in a whole host of other areas where a specific situation was out of kilter with the "general" state of affairs.  That is why detailed research will always trump eras. 

 

Before the arrival of the satnav, I described eras as a way to get you to roughly where you wanted to be - like a road atlas would get you to the right town - but for the detail you needed to go further - so having arrived in your chose destination town (= rough time period) you would need an A to Z, or to ask a local if you needed to find where 19, Acacia Avenues  could be found.

 

*everyone knows the Galileo experiment that showed that heavy and light objects fall at the same rate - by having dropped things from the top of the tower of Pisa.  Now try it with a lead weight and  a feather.  With the exceptions come explanations and those explanations are what makes our hobby so interesting.

 

Edited to add:  I think we are in danger of being in violent agreement here Brian.

Edited January 20, 2018 by Andy Hayter

[Mike Storey]

On t'other hand, one could have a jolly good debate about what the difference is between what people state they actually model, against what their preferred "era" is?

 

For example, 71 people say they model pre-grouping, but only 48 state that is their preferred period. Similar results for BR Blue.

 

Both eras are harder to model at present (using RTR and RTP or kits) than many of the others, but people have still chosen to model them against their preference. I can see this might occur where one is involved in a club layout, and so the response is what the club have chosen, but otherwise, I don't get it.

 

How would manufacturers conclude demand from that?

[Ravenser]

On t'other hand, one could have a jolly good debate about what the difference is between what people state they actually model, against what their preferred "era" is?

 

For example, 71 people say they model pre-grouping, but only 48 state that is their preferred period. Similar results for BR Blue.

 

Both eras are harder to model at present (using RTR and RTP or kits) than many of the others, but people have still chosen to model them against their preference. I can see this might occur where one is involved in a club layout, and so the response is what the club have chosen, but otherwise, I don't get it.

 

How would manufacturers conclude demand from that?

 

You were allowed to choose multiple periods.

 

Blacklade runs BR blue (in fact Blue/Sectorisation) 1985-90, and Privatisation 2000-7. What has gone to the shows is BR Blue, reflecting my personal preferences. I also keep my accumulated kettles active by running a "funny trains" steam period c1958, for which I've built a number of coaches.

 

The Boxfile is post war BR - a Y3 and 05 in BR black with cycling lion, a Hunslet tram in green and a Knightwing shunter which is I know not what but it came out of a Humbrol Emerald Green spraycan. Stock carries a mix of pre 64 and post 64 lettering - mostly pre 64

 

Tramlink - which has been buried under magazines for years - would be circa 2000 when I started building it - ie the original red/white

 

I actually ticked BR Blue TOPS (arguably it should have been Sectorisation) and BR late crest. Preferred period is BR Blue TOPS - because it is, however that might be flexed in response to the demands of a particular project. 

 

I suspect the way I simplified may be fairly typical

 

(And if I'd built an LT tram layout I could have "claimed" grouping as well)

[Dunsignalling]

Brian Macdermott's point about the 'era system' being set up for retailers to get some form of relative context to help them sell things is fair enough, but I've always thought the main problem with the era system is that if I consider I am modelling the period '19xx-yy', then I probably should be focusing primarily on locos and stock built 10 or even 20 years previous to that time span.

Definitely. The reason I lengthened my core 1958-62 period a bit at either end was that several classes of loco and types of coaches I like disappeared from the scene before 1958.

 

I do try not to run anachronistic combinations though; preferring to use Rule One to stretch reality, not shatter it. 

 

My 2-year extension at the other end is very little exploited and is only really there to excuse one or two items I am especially fond of, like a Hymek with yellow panels and D0280 Falcon in 2-tone green livery (which may actually, I gather, be a little later still).

 

John

[jjb1970]

Maybe it'd be informative to include questions on how many people own chunky sweaters, sandals and drink real ale......

[Andy Hayter]

Chunky sweaters keep out the cold.

Sandals are a bit naff in 12cm snow - see chunky sweaters above

real ale?  I wish.  Biere artinisale is the closest  but real ale it isn't.

[Ozexpatriate]

The poll has been running now for about 10 days and we are still seeing people vote every day. We're now at 351 respondents, a number with which I am very pleased.

 

At the same time a few of us are doing some analysis of the data - which is welcome by the way. Based on that I will present the analysis that I wanted to do in my next post.

 

I did elaborate in the opening post that I used the Bachmann 'eras', not because I like them, but because the two largest manufacturers have now started using them. I completely agree that polling experienced enthusiasts by the specific years (or range of years) that they model is a better approach, but the poll software only allows (if I am remembering it correctly) a total of 25 answers.  Different software would be required to harvest year data for everyone, but more on that later.

 

While they do nothing at all for experienced enthusiasts I do think the "eras" serve a purpose.

 

What they offer is a first order of approximation which better than nothing for people who are inexperienced with railway history. As I posted on the Hornby announcements thread, the era concept is helpful insofar it might help avoid this conversation:

 

Gran: "I thought you wanted a Hornby"

Grandchild: "Yes Gran, thank you, the GWR Collett coach you bought me is lovely but it doesn't really go with my GWR HST."

 

Eliminating them for retailers may well throw the baby out with the bathwater. Only our retailers who post here can answer the question, but asking a novice what year they plan to model is probably a bit much.

 

Perhaps it's a strained analogy, but we don't have first graders in primary school study Shakespeare or Chaucer and with all their flaws the eras aren't really awful in terms of what runs with what for beginners.

Edited January 21, 2018 by Ozexpatriate

[Ozexpatriate]

Without further ado, here is my first-level analysis.
 
The poll set out to provide new data for people who asked for it in the Hornby thread and the purpose was to examine whether people modelled what they remembered - the idea being that if people model what they remember, there should be a big demand for post-steam models.
 
Here is how I wanted to present the data for the first three questions. 
 
The age cohorts have been associated with the "era" at which the youngest people in the cohort were 10 years old. Ten is an arbitrary number but it fits best with the age ranges. I suspect very few people possess a sufficiently eidetic memory where they are substantially aware of details of railways at five years, such that they can model the specifics.
 

 
Some will suggest that since the eras have different lengths - particularly the BR blue, pre-TOPs era - that this will distort the data. (Actually it doesn't distort it, but it is the reason for fewer voters being about 10-14 during that period.)
 
Binning the data into wider periods results (as suggested above) in more even ranges of years. Here is that data:

 

Edited January 21, 2018 by Ozexpatriate

[Ozexpatriate]

To draw conclusions it is easiest to examine the second of those graphs:

 

 

If we examine the region labelled "A" we are looking exclusively at respondents who model a period that they do not remember (except for the oldest 1% of respondents). First-hand knowledge is not informing their choice of subject.

 

If we examine the region labelled "B" we can see that at least 50% of the respondents who can remember the post steam period, don't choose to model it.

 

Those two observations, along with the answers to the direct question in the survey conclusively indicate to me, that while it is a factor for many people, more people don't use their first-hand experiences to choose a modelling subject in terms of period, much like we have seen in similar analysis before.

 

I will leave the poll open for now. It is easy to redo the analysis, but the trends haven't changed much since the beginning of the poll.

Edited January 21, 2018 by Ozexpatriate

[Ozexpatriate]

What interested me the most however was to compare the answers to questions two and three.

 

 

Generally speaking, the responses to question two were very close to 2x the answers to question 3 and by and large this was true in every single era. This would be consistent with everyone (on average) modelling two 'eras' or one person modelling three for everyone else who just modelled one, etc.

 

There were two outliers - pre-grouping and BR Blue - Pre TOPS.

 

I think the wording of question three had an effect in terms of "ONE preferred subject".

 

What it indicates is that while people do model the BR Blue pre-TOPS period, respondents don't prefer it much and quite dramatically so.

 

It also suggests something aspirational about pre-grouping. It is clearly (and unsurprisingly) less modelled than say the grouping period for which there are many more models, but clearly many respondents "prefer" it (relatively speaking). 

 

What also stuck me is how even the popularity of the grouping and BR early crest periods were and that both of them were more popular than any post-steam period, though this changes when you aggregate the post-steam periods into larger ranges of time.

Edited January 21, 2018 by Ozexpatriate

[Ozexpatriate]

... All these periods are roughly the same length - a quarter of a century . BR diesel and electric is a little longer than that, BR steam a little shorter. 

 

This means we can work with reasonably equivalent periods of time - which  makes comparing the levels of support much easier

 

* For BR steam, at time of writing, the combined total support is 262 votes

* For BR diesel & electric, at time of writing the combined support is 201 votes

* For post-privatisation, at time of writing the combined support is 65 votes

This was an analysis I always intended to run, and as you can see above have done so, though I did not include the "Sectorization" data with the BR Corporate Blue data the way you have - which changes the aggregates somewhat.

 

As you say, the British Railways / Nationalization period is clearly the most popular.

 

Comparing British Rail and Grouping is less clear. The results for Grouping to question three are slightly higher for Grouping than the Corporate Blue period.

 

Aggregating all the post-Corporate Blue periods together is approximately as popular (within the confidence interval of 3.7% for those values) as the Grouping period. Whether we should aggregate the data differently is  separate question. It can of course be done. To my mind (and though the prism of an foreign enthusiast who never witnessed that railway first hand) the Sectorization period looked very different, but here we get into the mire of how useful the 'eras' are.

 

EDIT:

 

If we aggregated both BR Blue + Sectorization we get this:

 

 

Personally I think it's gobbling up too many buckets but we may all differ.  I don't subscribe to the idea that even periods of time are necessary to meaningfully analyze the results.

 

I never set out to identify the most popular period, just establish whether first hand experience is the primary driver of choice of period. (Conclusively it's not.)

 

Post Script EDIT:

 

There's actually some error that creeps in with the aggregated data. Say someone votes for BR early and late crests in question 2. When aggregated to "Nationalization" this now represents two votes instead of the one vote someone who only modelled grouping might have placed. When looking at the aggregated results, I believe the "one preferred" region is a more accurate indicator since there's no double counting there.

Edited January 21, 2018 by Ozexpatriate

[Dunsignalling]

One thing that stands out for me (if only intuitively) is that, for those who pick eras outside their 10 to 14 years age "slot", their choices seem to be heavily tilted towards earlier rather than later. 

 

An example of young enthusiasts absorbing the preferences of older "mentors" in their formative years, perhaps?

 

John  

[Nearholmer]

Isn’t the ‘blue pre-TOPS’ timeslot very short (c4years?) compared with the rest (even early crest was longer)?

 

It’s a bit like defining to have a category for ‘the late afternoon of June 21st 1976’, which was definitely shorter than ‘all of history up to 1923’.

 

Kevin

[Dick Turpin]

Interestingly, when I was younger i disliked the transition period because I found it dirty, depressing and moribund, but these days I rather like it for exactly the same reasons. It seems the further back in time those days become, the more the period appeals to my sense of sadness and loss. 

Edited January 21, 2018 by Dick Turpin

[Colin_McLeod]

Isn’t the ‘blue pre-TOPS’ timeslot very short (c4years?) compared with the rest (even early crest was longer)?

It’s a bit like defining to have a category for ‘the late afternoon of June 21st 1976’, which was definitely shorter than ‘all of history up to 1923’.

Kevin

https://www.onthisday.com/date/1976/june/21

[BMacdermott]

The poll has been running now for about 10 days and we are still seeing people vote every day. We're now at 351 respondents, a number with which I am very pleased.

 

While they do nothing at all for experienced enthusiasts I do think the "eras" serve a purpose

 

Eliminating them for retailers may well throw the baby out with the bathwater. Only our retailers who post here can answer the question, but asking a novice what year they plan to model is probably a bit much.

 

Perhaps it's a strained analogy, but we don't have first graders in primary school study Shakespeare or Chaucer and with all their flaws the eras aren't really awful in terms of what runs with what for beginners.

 

Hello Michael

 

Whilst I agree that the era system has value for retailers/novice customers, I disagree with the way some retailers/makers portray its veracity/accuracy. It cannot be fully relied upon to show what runs with what.

 

Bachmann's era system description in the catalogue says (they use bold italics): Please note this classification is not absolute and is intended to be a rough guide only. Correct me if I'm wrong, but I don't think others emphasise that fact, and well done Bachmann for doing so.

 

To concur with Andy Hayter earlier... If someone asks me if I can get from A to B, I can answer that yes, I can, as I have a car. It won't matter a jot what type of car I have. However, if ask for insurance on my car, the company will want to know age, colour, make, model, etc.

 

In other words, one is for the generalist; one is for the specialist. Both have a place at specific times.

 

Brian

[Mike Storey]

Interestingly, when I was younger i disliked the transition period because I found it dirty, depressing and moribund, but these days I rather like it for exactly the same reasons. It seems the further back in time those days become, the more the period appeals to sense of sadness and loss. 

 

Totally agree (although for me, that is why I model BR Blue!).

[tractionman]

One thing that stands out for me (if only intuitively) is that, for those who pick eras outside their 10 to 14 years age "slot", their choices seem to be heavily tilted towards earlier rather than later. 

 

An example of young enthusiasts absorbing the preferences of older "mentors" in their formative years, perhaps?

 

John  

 

 

This assessment matches my experience—my ‘formative’, 10-14 year old period falls right at the peak of the BR Blue era, ‘77-’82. Love it! About ten years ago, as I got back into railway modelling in a reasonably serious way, I focused pretty much exclusively on recreating aspects of my formative years in miniature form, and my plank layout “Spencer Park” was the result, plus lots of purchases of BR Blue stock and associated books, such as the Strathwood ‘70s Spotting’ series, visits to diesel galas etc. Excellent stuff.

 

Ten years later I find myself, though still drawn to the BR Blue of my youth, rather more drawn to the inter-war period, the era of my grandparents, whose stories I heard as a 10-14 year old caught my imagination and have stayed with me over the intervening decades. Plus the 1970s modelling world (of my formative years) were I think rather dominated by layouts representing the Big Four, which I read about and saw at the time in 'Railway Modeller' and exhibitions, so in a curious sort of way the Grouping era (though I never knew it first hand) is instilled in me and part of my 10-14 year old experience of four decades ago.

 

So my principal modelling interests / era now is 1930s LMS, beloved by my maternal grandfather, whose stories about the big red engines of his youth captured my young imagination.

 

Cheers,

 

Keith

[eastglosmog]

Isn’t the ‘blue pre-TOPS’ timeslot very short (c4years?) compared with the rest (even early crest was longer)?

 

It’s a bit like defining to have a category for ‘the late afternoon of June 21st 1976’, which was definitely shorter than ‘all of history up to 1923’.

 

Kevin

Not as short as the "Contemporary" slot, which whilst a great deal longer now than when Ozy set up the poll, is still only 21 days as it is defined as 2018!