Classifying Blends

I think it would be best to use the 2 main categories and, if possible, have the sub-categories as you have stated. That would be the easiest approach from the user standpoint.
Also, would it be possible to have a search by reviewer type set up that would when putting in the reviewer ID and blend would bring up that particular review, not just bring up the blend reviews as a whole and having the user scroll down until they find that particular review?

English and aromatics won’t even cover half the blends out there. Virginias, VaPers and Burleys don’t fit there at all (with Burleys I mean Aged Burley Flake, Stonehaven, etc., which are neither aromatics or English).

@Xeneize:

By definition, English is anything that is not an Aromatic, so yes, it would cover everything.

There are plenty of non-aromatics that aren’t “English.” For example, it would be difficult to justify classifying straight tobacco of a single type (for example Blending Perique) as “English.” Perhaps if a single dichotomy is needed it should be “flavored” and “unflavored.” Just my humble opinion, as an admitted (and often confused) newcomer to pipes.
Dave Webber

You will drive yourself to cigarettes or worse trying to classify blends. Besides, what if a new piper wants to access info on tobaccos, a new piper does not know what an English or a Burly flake is. This site was very helpful when I did not know a Vaper from a Chevy.

Just my 2 cents.

Jon, even when you’re right, most pipe smokers use the word “English” for blends with Latakia. If they had to enter a new blend when, I guess they would be quite puzzled when they’d get to the “English/Aromatics” option.

Even more, by your definition “English” means “not cased”. I would be tempted to put many English mixtures into the aromatics category (PS English Luxury, English Oriental Supreme, Brebbia English Mixtures No. 7 and 8, Davidoff Royalty, Presbyterian, etc.).

I’m not too sure how useful is that…

Check out the results of the poll conducted here to determine what the definition of an English blend is widely considered to be.

Essential, but tough job! MO got this one right. “You will drive yourself to cigarettes or worse trying to classify blends.” Or maybe like trying to figure out how many angels can dance on the head of a pin. Just to get the battle joined. Here’s how I do it, with my favorites as an example:
Aromatic Stokkebye Cherry
American Carter Hall
Full Virginia Stonehaven
Mild Virginia Ries Virginia Flake
VaPer Escudo
Full English Penzance
Mild English Ries Gourmet English
Stop Throwing Stuff, it’s only a start!

I would be concerned about doing any kind of definition that would break it simply into “flavored/non-flavored” categories. For one thing, how many of us actually know for certain which blends are flavored or not? Haddo’s Delight, 965, Stonehaven and Mephisto are just some of the blends that spring to mind as not exactly being “aromatics” but nontheless — to my tastes, at least — are clearly “flavored.” Also, I’m not sure that many tobacco blends which people consider “English” are actually unflavored. I think the current system actually works just fine, as it allow reviewers to speculate as to what degree a blend is flavored, squabble about catergories, etc.

What at first seems to be a relatively simple task
will at best wind up with some standard that will garner agreement among many, but certainly not all.

To get some idea as to what is involved here, take a look at http://en.wikipedia.org/wiki/Definition. And that is only the beginning!

A simple, but hardly a comprehensive standard that would avoid all argument as to what is English or aromatic (there are English aromatic mixtures out there)is to acknowledge that whenever we smoke a tobacco it will fall into one and only one of the
following:

1. A straight tobacco, e.g. perique, latakia, Virginia.

2. A blend consisting of at least two types of the same tobacco. Virginia blends are probably the
most common.

3. A mixture consisting of at least two different tobaccos.

If I’m not mistaken, Alfred Dunhill suggested such
an approach many years ago in one of his books.

Tony Gaito, aka tonyg

I have to agree with Mo. There’s simply far too many disagreements over what a particular brand is. I don’t agree that an ‘English’ tobacco can have latakia in it for example. That seems to be a definition originally arrived at in the USA which has spread around the world, but is not actually accurate. If you really have to do it; then Tony Gaito’s ideas seem the best route to follow (along with the ghost of Mr Dunhill).

I agree with Mo also. The only way to establish any sort of consistent classification scheme is to post YOUR OWN set of criteria and insist that everyone abide by it - not a very appealing idea, and one that would certainly cause more trouble than it would be worth.

I agree with jj. It’s easy enough to think of clear-cut examples of “English” and “Aromatic” blends that nobody would dispute. But there are plenty that don’t fall easily into either category.

What I think would be much more helpful, and accomplish a similar purpose, would be to have an algorithm that recommends blends you haven’t tried based on the ratings you submit on other blends. Netflix.com has such a system for recommending movies, and I think it works very well for them. The basic idea is that the program finds other users who have similar tastes to yours (both likes and dislikes), and tells you what else those users have rated highly. For example, if I give Escudo 4 stars and Dunhill Light Flake 1 star, the program would find other reviewers who have rated those two the same way I have, look for other blends that they have rated highly (e.g. 2015), and recommend them to me.

In a way, that allows each user to, in some sense, set their own criteria, at least in terms of what they like and don’t like. People who strongly favor a certain type, e.g. VaPer, will probably get mostly recommendations of other VaPers. These “categories” need not be explicitly designed. They will spontaneously emerge from the pattern of reviews.

I know this type of programming might be complicated, but I figured I’d make the suggestion just in case it’s possible.

Jon,

I’d love to see such a feature and would use it quite a bit, but as others have echoed above, I doubt that it can be done to anybody’s satisfaction. Years ago, William Serad talked about blend categories in Pipes & Tobaccos magazine and there is a very good section on Tobacco Touchstones on The Pipe Tobacco Aging, Storage and Cellaring FAQ. http://agingfaq.nocturne.org/indexphp

I don’t completely agree with either one of them and I doubt that anybody here would completely agree with my scheme.

I _DO_ like prof’s suggestion above about coming up with an algorithm that could predict what other blends you might like. Assuming that it’s feasible (sounds like a lot of work!!), make it a feature that would only be available to registered users.

– Doug Pearson

Just to further muddy the waters, GL Pease (the somewhat famous blender) contends that basically ALL blends are cased, even if it is only sugar water. He says that almost nobody has ever had a truly “straight, non-cased tobacco.”

At any rate, I think some kind of categorization is helpful. There will obviously not be a perfect system which will please everyone, but I would find it helpful if I was exploring say, Burley blends, to simply click a link that says “Burley Blends.” That is much easier than wading through tons of blends with non-descript names like “Vegas Nights” or “Rooster Nest.”

I’m just a newbie here, but categories that make sense to me are:

Aromatic
English
American English
VaPer
Virginia
Burley
Balkan
Scottish

Since many blends (including those that are intentionally “cross-over”) are difficult to characterize, why couldn’t a blends be tagged for more than one category? For instance, Morley’s Best from C&D could be classified as both a “Burley Blend” and an “American English Blend.”

Anyway, I think the classification is another tool that could be helpful to many. But individual users can take it or leave it. Seems to me than even a flawed and subjective classification is better than no classificaiton at all.

I agree with Illinois Hick on this one. As a relative newbie, it’s hard to sort through all the labels. The classifications would be helpful. If you’d like to try to please the masses on this one, you could take a poll of the classifications people would like to see and go from there. Democratic process, more or less. You could break new ground here. Who knows? Maybe our grandchildren will look back to the day when tobaccos were once and for all authoritatively categorized by their faithful forefathers on Tobaccoreviews.com!

Jon,

I agree with Illinois Hick and JJ and the rest - classification is a dangerous business!

Also - for the record - the current “advanced search” is a truly excellent tool that ALREADY allows searching by tobacco type, brand, even strength and room note, to say nothing of the recommendation!

For my money, a more obvious physical placement of the “Advanced Search” would really be all you need, but it would be helpful to have headings for each criteria.

Dave Webber

Seems to me than even a flawed and subjective classification is better than no classificaiton at all.

Agreed. It’s just for the purposes of making searches and comparisons easier, not establishing an infallible canon of classifications. They way the advanced search is set up now has very limited useability.

I have always thought it would be handy to have a search criteria based on the average rating categories, such as Strength, Flavoring, Taste and Room Note. I believe your site already has a search catgegory (By Rating) that lists those tobaccos that have Four Stars. I prefer the tobaccos that would fall into the Strength category of at least “Strong” and would like to have the ability to only search those blends. Don’t most English blends fall into the stronger categories and aromatics fall into the milder categories? Just a suggestion.

It seems to me, reading through the replies, that most people consider this to be a bad idea - one fraught with too much possibility for acrimony, and not accomplishing much in the long run.

Perhaps the best way to proceed is to, as several people have suggested, beef up the search algorithms to include more criteria and to add some new ways of discovering blends you might like. I’ll post a separate article on directions that could go for discussion.

I personally categorize into 3 types for my own personall use :

1) “True Aromatic” - fully and noticeably cased like Blue Note (Dan)
2) “Lite-Aromatic” - lightly topped with something however faint like Frogmorton or Flake DeLuxe
3 “Non-Flavored” - not topping whatsoever, like Ashbury, Nightcap or Christmas Cheer.

I have found “English/Balkan/Scottish/Gibberish/Tom-foolery” to be plain confusing and totally unrelated to whether I liked the blend and my smoking pleasure. What’s in a name? “A Rose by any other name would smell like Perique!”

Thanks,

Ethan “Glorfindel” Dickey

There are so many possibilities for blending and so many tastes that classification, even by “pure” standards seems too rigid — much like the classification of pipe shapes which we all know has no strict bounds.

What you may call a lovat, I may call a billiard whilst another may cry out, “Brandy, of course!” The ever-raging Bulldog versus Rhodesian may never be decided except by those members of the AKC!

Should pipe tobacco (straight or blended) be measured on such discrete metrics? Though many hold in their minds firm definitions of English, Aromatic, Vapers, etc., where does one end and another pick up? Does 3% Vanilla Black Cavendish make a blend with 40% latakia an aromatic? Is FM an aromatic? What about FMOT? (I can hear the “room” of smokers dividing.) Mississippi Mud is not just a clever name, but the name of a Vaper, no wait…English…wait, that’s not it…

I’m game for an algorithm based search…perhaps one utilizing similar criteria as one used for a certain online tobacconist? Hmmm?

Joe P

Jon,

First of all, thanks for setting up a great site!

Just thinking and decided I’d enter my 2 cents.

While classifying blends is worthwhile, most people don’t taste the same things the same way. It’s a scientific fact (Biology, 10th grade, 1974).

And most of the descriptions used by the tobacco sellers are subjective, to put it mildly. One man’s favorite blend can be another man’s gag-reflex!

My idea was that if someone liked a particular blend, there could be a way to rate it and at the same time suggest other blends that they also like as much. Not scientific for sure, but could be helpful. For example, I really love a blend- Gatlin Burlier, “Chimney Smoke”. I also love their “Mountain Mall Mix” and “Morning Moonshine”. They are all different “types” of blends, but have similar attributes that I enjoy without being able to put those attributes into words.

Off my soapbox. Just wanted to also add that it would be great if more tobacco shops (especially online) offered sampler packs. Randy’s Tobacco Shop in Indianapolis sells 8-one ounce samples for less than $20. Kudos!

Hey, how about a swap-a-sample program. Always brainstorming (with what little I have left :)).

Cliff Shelby,
Castle Rock, Colorado