Most Controversial Tobacco Blends

[hfearly]

Last night a friend and I were discussing why certain blends of pipe tobacco are so controversial in terms of peoples like and dislikes. You know, some blends are really polarizing and people either love or hate them.
My friend had the theory that only Aromatics cause this kind of love/hate polarization, and I had to disagree, remembering a lot of the discussion around VAs and English blends. Having access to the tobaccos and ratings database, we went ahead to check our theories
First, we defined whether somebody "likes" or "dislikes" a blend on whether the rating the person gave on a 10 - point scale (1-3 Dislike, 4-7 Impartial, 8-10 Like). Then we counted for each blend, how many likes and dislikes that blend received in total, and defined 'controversy' as a ratio of how evenly spread likes and dislikes are. For example, if 49 people disliked a blend and 55 people liked a blend, this indicates high controversy (55 / (55+49) - 50), as compared to another blend that received 81 likes and 2 dislikes (81/ (83+2) - 50). So, we receive a value on a scale of 0 (highly polarizing) to 50 (everybody is on the same page).
As it turns out, Aromatic blends are indeed the most polarizing type of tobacco, followed closely by Balkan blends. The least polarizing blends appear to be Va/Per and Oriental mixtures. What surprised me personally was the relative difference of about 12 points on the absolute scale between highest and lowest, which accounts to 24% difference on a relative scale, a very strong effect in the background of Psychology and Market Analysis.
In the following, I present to you our data, together with the top 5 most controversial blends for each tobacco type. All sorting is done from most controversial to least.
Results based on 1241 Blends of Type Aromatic (22.29)

• Captain Black - Royal
  
• Altadis - Classic Vanilla
  
• Cornell & Diehl - Organic Pipe Dreams
  
• Mac Baren - 7 Seas Royal Blend
  
• Lane Limited - Cherry Cavendish
  

--------------------------------------------
Results based on 269 Blends of Type Balkan(24.15)

• Iwan Ries - King's Oriental
  
• Brebbia - English Mixture No.8
  
• Robert McConnell - Latakia Flake
  
• Samuel Gawith - Perfection
  
• Two Friends - Valle Crucis

--------------------------------------------
Results based on 201 Blends of Type Va/Bur (25.97)

• House of Windsor - Briggs
  
• pipesandcigars.com - Old Companion
  
• Altadis - Saint Claude
  
• Altadis - Blue Note
  
• Benson & Hedges - Mellow Virginia

--------------------------------------------
Results based on 37 Blends of Type Va/Per/Bur (27.52)

• Just For Him - Old Toby
  
• Cornell & Diehl - Serenity Series: Comfort
  
• Cornell & Diehl - 107: Haunted Bookshop
  
• pipesandcigars.com - Sunza Bitches
  
• Butera - Kingfisher

--------------------------------------------
Results based on 447 Blends of Type Virginia(29.27)

• Mac Baren - Virginia Flake
  
• Ashton - Old Church
  
• A & C Peterson - Caledonian Virginia Flake
  
• McClelland - Bulk No.5105 Stoved Virginia
  
• Rattray - Black Virginia

--------------------------------------------
Results based on 465 Blends of Type English(29.55)

• Cornell & Diehl - Sam's Blend
  
• Cornell & Diehl - Canal Boat
  
• Benjamin Hartwell - Chesdin Mixture
  
• Mac Baren - Latakia Blend
  
• Cornell & Diehl - Red Stag

--------------------------------------------
Results based on 123 Blends of Type Va/Per (31.82)

• Tordenskjold - Virginia Slices
  
• Former - Straight Grain Flake
  
• Ashton - Black Parrot
  
• Cornell & Diehl - Bayou Morning Flake
  
• Bufflehead - King Eider

--------------------------------------------
Results based on 652 Blends of Type Oriental(33.43)

• McClelland - Oriental Mixture No.6
  
• McClelland - Oriental Mixture No.1
  
• McClelland - Rose of Latakia
  
• Rattray - 7 Reserve
  
• Pinkerton Tobacco Co. - Granger

 

[hfearly]

Roth: Yes, I've only ever heard good things about Kingfisher from hearsay, but the data tells that there is a like/dislike ratio of 43/26 for that blend (63% love it, 37% hate it).

 

[judcole]

Interesting to say the least.

I am not sure why anyone would consider Granger an Oriental blend. It's a codger burley.

 

[papipeguy]

Okay, I'll bite. A facinating list to say the least; and one that will create some controvery here. So, well done on that. What data base did you access and how were the categories defined? I'm a frequent visitor to tobaccoreviews.com; did you use that? It's astounding to me that you had this conversation last night and compiled the data in such a short time. That has to be at least a 12 beer night; so well done on that point as well.

I'm just trying to get my head around the info you're presenting as I find it extremely interesting.

 

[mso489]

Granger advertises as a Kentucky Burley with the "Wellman process" as its secret sauce, more or less,

a dressing that has been part of it for its long existence. It's a good, if not inspired, mild blend, smooth

and long burning. I can't see it as an oriental, and calling it an aromatic seems like a technicality, but

I'd be interested in any information as to why it is designated as either oriental or aromatic. A codger

burely is closer to the truth, but still a good smoke. As an old song says, "After you've been having

steak a long time/ beans taste fine." Granger is a good change of pace.

 

[juvat270]

I did find it interesting that a puffer posted something recently about Granger being, IIRC, an aromatic -- that, too, I thought a bit strange.

I've noticed quite a few people on the forums that seem to think that anything with casing or topping automatically makes it an aromatic. I used to just chalk this up to them being misinformed. But maybe I am in the wrong. I always viewed an aromatic as a blend that is cased or topped with an artificial flavoring in order to give the smoke, overwhelmingly, that flavor. For instance cherry aromatics; when you smoke it, the idea is that all you taste (and smell) is cherry flavored smoke.
But like I said, maybe its my understanding that is wrong. For example, I've seen some posters refer to Carter Hall as an aromatic.

 

[zonomo]

I think you'd have to add Captain Black White to this list. It's controversial just because of what it is, its background and where it's sold. I think one of the problems here is to think about what is "good" or "not good" because honestly, everyone's tastes are so different. I have learned one thing in the 4 months since I've been smoking: When I mention Captain Black, I need to brace for the hit.....

 

[teufelhund]

Cool presentation; I just think there probably too many other variables to really take this at face value. If you factor in cost alone it starts to give me a headache. I'm certainly more pissed when I shell out more cash for something I don't end up liking and pleasantly surprised when drugstore baccy hits the spot for a tenth the price. I would have just left it at tobacco preferences are controversial and lit a nice bowl of my current favorite and just knew somebody out there hated it with wild abandon. Have a good one...

 

[Perique]

Really interesting and cool post. Thank you.

 

[andrew]

Mclelland stove Virginia is a blending tobacco not meant to be smoked on its own, like most blending tobacco

 

[romeowood]

Fascinating approach to a juicy question, hfearly. Evident already from the comments, though, is the ever-elusive (illusive?) nature of the beast: one man's aromatic is another man's english. Trying to categorize some of these blends down is an effort in futility, and therein lies one of the wonderful conundrums of the hobby.

 

[hfearly]

What data base did you access and how were the categories defined?

My private database of a dozen thousand reviews of tobaccos across the internet. Categories are defined as follows: a blend that contains

- only Virginia is a Pure Virginia

- only Burley is a Pure Burley

- Virginia and Burley is a Va/Bur (and maybe others but not Latakia)

- Virginia and Perique is a Va/Per (and maybe others but not Latakia)

- Virginia and Perique and Burley (and maybe others but not Latakia)

- if any of the above also contain Latakia it's an English

- if there are any Oriental tobaccos, it's an Oriental

- if there is Cavendish and saucing it's an Aromatic. If there is a saucing but no Cavendish, the reviews decide like below.
In addition, all review text's are matched on phrases like "... this is a really wonderful English blend..." and the above heuristics are augmented with the majority vote of reviews. I.e., blends that contain Va, Oriental, Latakia are off the bat classified as Orientals, but reviews might mention a lot that it's a Balkan blend, so that answer might be more true.

I am not sure why anyone would consider Granger an Oriental blend. It's a codger burley.

Jud, you are absolutely right, it's a Burley blend, and was mislabeled during copy and paste, I double checked.

If you factor in cost alone it starts to give me a headache. I'm certainly more pissed when I shell out more cash for something I don't end up liking and pleasantly surprised when drugstore baccy hits the spot for a tenth the price.

Devildog, that's a very interesting observation and very reasonable. I agree that pricing can have a reinforcement effect on opinion building, be it positive or negative.
Another variable was touched on in Brian's latest Radio Show - people might be biased towards a certain manufacturer and blends from other companies just don't fly with them. I might check that theory next.

one man's aromatic is another man's english

Maybe that's a worthwhile lesson for Blenders? Don't sauce your Latakia blend, and then call it an English - it might just provoke both, Aromatics-biased, as well as English-biased smokers, and fit neither cup?

 

[Perique]

Fascinating. If you explore the brand-loyalty-bias element, please share your conclusions.

 

[hfearly]

Well, I will start off with the Top 100 most liked and disliked Blending Companies of all time
A quick note regarding the format: the numbers in brackets record the ratio of Love to Hate, a value of '1' means equal love to hate, a value larger than '1' means biased towards love (more love than hate), and a value smaller than '1' means biased towards hate. I took only data for which there were at least 100 votes (either love or hate) for at least one of the tobaccos they produce.
Some interesting observations:

1. 4noggins made it to the Top 10![/*]
   
2. G.L. Pease ranks above Orlik![/*]
   
3. Place 72 to 89 are those that produce tobacco blends, which about an equal number of people love and hate. We find old friends like Altadis, Planta, Sail and Lane Ltd. there.[/*]
   
4. The 'break even' point is very far down the top 100 (Altadis, place 88), so we have a ton of really good companies that produce very lovable tobacco. We do truly live in a Golden Age of pipe smoking![/*]
   

From there, I'll probably go on and use this 'love-to-hate' ratio to assign penalties in the brand-bias experiment down the road (a person may be more likely to love stuff from a Brand that everyone loves, so it's no real bias).
1. Compton's of Galashiels (39.999 )

2. Uhle's (9.599 )

3. Mick McQuaid (9.142 )

4. John Patton (8.181 )

5. Schürch (7.818 )

6. Smoker's Haven (7.333 )

7. Sobranie of London (7.307 )

8. Affordable Pipes (6.999 )

9. Peretti (6.772 )

10. 4noggins (6.483 )

11. Capstan (5.399 )

12. Reiner (5.049 )

13. Watch City Cigar (4.916 )

14. Milan Tobacconists (4.619 )

15. Solani (4.177 )

16. Boswell Pipes & Tobacco (4.162 )

17. Sillem's (4.136 )

18. Pipeworks & Wilke (4.074 )

19. G. L. Pease (4.056 )

20. Orlik Tobacco Company A/S (3.874 )

21. Esoterica Tobacciana (3.821 )

22. McCranie's (3.803 )

23. Paul Olsen (3.791 )

24. E. Hoffman Company (3.769 )

25. Fribourg & Treyer (3.666 )

26. pipesandcigars.com (3.637 )

27. Charles Fairmorn (3.624 )

28. Wessex (3.605 )

29. Hermit (3.458 )

30. Bufflehead (3.449 )

31. A & C Peterson (3.374 )

32. William P. Solomon (3.199 )

33. Samuel Gawith (3.160 )

34. McClelland (3.123 )

35. CAO (3.117 )

36. James Fox (3.081 )
^---- Like a LOT more than Dislike
37. Rattray (2.999 )

38. Robert Lewis (2.944 )

39. Dunhill (2.943 )

40. ThePipeRoom.com (2.894 )

41. Edgeworth (2.866 )

42. Just For Him (2.754 )

43. Gawith, Hoggarth & Co. (2.698 )

44. Condor (2.696 )

45. Robert McConnell (2.669 )

46. Cornell & Diehl (2.538 )

47. Peter Stokkebye (2.498 )

48. Astley's (2.454 )

49. Two Friends (2.439 )

50. Nat Sherman (2.384 )

51. Ashton (2.327 )

52. J. F. Germain & Son (2.247 )

53. Peterson (2.246 )

54. Balkan Sasieni (2.230 )

55. Bell's (2.114 )

56. Dan Tobacco (2.098 )

57. Benjamin Hartwell (2.038 )

58. Ogden (2.037 )

59. Brigham Enterprises Inc. (2.032 )

60. W.O. Larsen (1.909 )

61. Prince Albert (1.902 )

62. Butera (1.890 )

63. Murray Sons & Co, Ltd (1.766 )

64. Treasures of Ireland (1.749 )

65. Fader's (1.739 )

66. Mac Baren (1.737 )

67. Torben Danske (1.690 )

68. Stanwell (1.660 )

69. Sutliffe Tobacco Company (1.660 )

70. John Middleton, Inc (1.650 )

71. Tinderbox (1.631 )
^----- Like more than Dislike
72. Erinmore (1.509 )

73. Pinkerton Tobacco Co. (1.482 )

74. Davidoff (1.449 )

75. Edward's (1.403 )

76. Kentucky Club (1.366 )

77. Brown & Williamson (1.362 )

78. Sail (1.306 )

79. Brebbia (1.294 )

80. Holger Danske (1.249 )

81. Imperial Tobacco Group, PLC (1.238 )

82. Lane Limited (1.228 )

83. Troost (1.170 )

84. House of Windsor (1.121 )

85. Savinelli (1.099 )

86. Planta (1.054 )

87. Iwan Ries (1.045 )

88. Altadis (1.025 )

89. Skandinavik (0.999 )
^----- Like about as much as Dislike
90. Dream Castle Tobacco Company (0.835 )

91. Alsbo (0.799 )

92. Smoker's Pride (0.772 )

93. Captain Black (0.756 )

94. Half & Half (0.625 )

95. Consolidated (0.612 )

96. Clan Shaw (0.499 )
^----- Dislike more than Like
97. Theodorus Niemeyer (0.489 )

98. H. Sutliff (0.321 )

99. Paladin (0.269 )

100. Borkum Riff (0.267 )
^----- Dislike a LOT more than Like

 

[hfearly]

Fascinating. If you explore the brand-loyalty-bias element, please share your conclusions.

So results fresh off the mill:
Out of a sample of 2,093 reviewers who reviewed tobaccos from at least 5 different brands, only 185 show a bias (one or two brands loved, all others hated), while 1,908 do not show a bias.
From the data, I'd very much conclude that there is no bias towards a certain brand.
However, from looking at the data so far, there appears to be a very strong bias towards certain tobacco types. Brand love or hate appears to be an artifact of whether that brand mainly manufactures tobaccos that a particular pipe smoker is biased to or not.
Indeed, I ran the test, and of a sample of 2,050 reviewers who have rated at least 3 different types of tobacco (Aromatic/English/Oriental/VaPer/ ...), only 85 show no bias towards a particular type of tobacco, while the remaining 1,965 do so (1,860 very strongly so!).
Alas, we are not MacBaren-smokers or G.L.-Pease smokers, we appear to be Aromatic-guys and English-guys. There is a grey zone (Aromatics smokers who like English blends) but it appears to be very small. The vast majority of the data suggests that most smokers have one or two single types of tobaccos they favor above everything else with the occasional 'fling' outside of their comfort zone.
That's a very interesting result to me as it will much influence my "Tobacco Recommendation" work.

 

[spartan]

I always thought Altadis blue TV was an aromatic.

 

[hfearly]

Roth: these folks appear to have recreated the Compton's blends of old. No idea if they are any good ... http://www.cubancigars4u.com/blends.htm
Spartan: That may very well be the case, however, there is no indication of an added flavoring from the manufacturer and the reviewers don't really say so, alas the computer wouldn't know

 

[rjscha]

This is incredibly interesting. Thanks to the OP for all of the work putting this together.
I'm wondering if your data takes date of review into consideration. I'm interested if there is ever a trend where a particular blend or category of blends become popular and thus reviews tend to be higher. You might be able to predict when certain blends will become harder to find.

 

[bigboi]

Wow this is great work. Very interesting indeed. Hfearly you could definitely work for some tobacco companies research and development team by compiling data to see where trends may be taking a turn towards!

 

[misterlowercase]

I'm wondering if your data takes date of review into consideration. I'm interested if there is ever a trend where a particular blend or category of blends become popular and thus reviews tend to be higher.

Such a popularity timeline chart would be quite interesting.
I'd wager that Murray's took a "hate hit" after they started making Dunhill tobacco.
Mac Baren gained love points after they developed Old Dark Fired.
Other brands are fueled by "mythic status" if discontinued, and if the blends are still being produced but always hard-to-get it would seem to generate a backlash at some point, detracting "likes" into the "over-rated" category?
Interesting stuff!