Back in February of 2017 I posted an analysis of all the geek critique scores and provided some commentary on what I'd observed. I have revisited that post with another year and a half's of geek critiques and some new charts and analysis for your enrichment. I excluded some cigars with only one or two reviews from the individual metrics, but those cigars are included in the aggregated data/analysis.
Some facts and figures:
* The total number of reviews grown from 932 to 1031 total reviews across 221 different cigars.
* The lowest average score with more than 3 reviews goes to the Montecristo Pilotico Pepe Mendez Toro, reviewed in March, 2018, with only 68.75 points average. This low average was due entirely to a terrible 13 point score earned due to a major defect the reviewer experienced.
The worst of the worst, the bottom 10 (includes cigars with 1 or 2 reviews).
1. Cooperstown Stinger 60.00
2. Cooperstown Grand Slam 67.00
3. Montecristo Pilotico Pepe Mendez Toro 68.75
4. Cuba Aliados (Core Line) Toro Extra 69.80
5. Di Fazio Picoso Doble Robusto 70.20
6. Dunhill Aged (formerly Dunhill Dominican) Condados 72.25
7. JFC Tobacco La Flor De Cuba 73.80
8. Evil Genius Cigars White Chapel Belicoso 74.60
9. Alec Bradley Black Market Filthy Hooligan (2013) 75.80
10. Leccia Tobacco Luchador El Gringo Frog Splash 75.80
* The highest average score has not changed and goes to the Sindicato Cigars Sindicato Toro Maduro reviewed way back in Feb, 2015, with an average of 94.2.
The Best of the Geek Critiques (includes cigars with only 1 or 2 reviews)
1. Sindicato Cigars Sindicato Toro Maduro 94.20
2. Paul Stulac White Blinding Light Torpedo 93.20
3. Partagas (Cuban) Serie D No. 4 93.00
4. Don Pepin Garcia (My Father Cigars) My Father (Rosado) No. 4 92.60
5. Drew Estate Herrera Esteli Corona 92.40
6. Vegas Robaina Famosos 92.20
7. Arturo Fuente Anejo Reserva No. 50 92.00
8. AKA (American Kick Ass) Nth Degree Apex 91.80
9. Valeroso Primavera Regalo 91.60
10. Crowned Heads La Imperiosa Double Robusto 91.40
* The cigar with the most consistent scores was the Illusione MJ12 (The Majestic 12) Maduro with an average of 88.66 (88, 89, 89).
The top 10 most consistently scored cigars (lowest standard deviation of scores) with corresponding score average:
1. Illusione MJ12 (The Majestic 12) Maduro 88.66666667
2. Drew Estate Undercrown Shade Corona 88.8
3. Leaf & Bean Leaf and Bean by Noel 90.4
4. Camacho Liberty 2015 90.2
5. Crowned Heads La Imperiosa Double Robusto 91.4
6. Sindicato Cigars Sindicato Toro 89.6
7. 1502 Black Gold Toro (Box Pressed) 87.5
8. Valeroso Primavera El Estado 89.6
9. Ave Maria (Core line) Crusader 91
10. La Flor Dominicana N.A.S. 86
* The cigar with the least consistent scores was also the Montecristo Pilotico Pepe Mendez Toro, likely due to the 13 point scoring throwing off the standard deviation.
The least consistently scored cigars:
1. Montecristo Pilotico Pepe Mendez Toro 68.75
2. JFC Tobacco La Flor De Cuba 73.80
3. Olman Premium Cigars Olman Toro 78.50
4. Di Fazio Maduro Doble Robusto 78.40
5. CyB (formerly Cuenca y Blanco) Lonsdale Club 77.80
6. Dunhill Aged (formerly Dunhill Dominican) Condados 72.25
7. Cuba Aliados (Core Line) Toro Extra 69.80
8. Pinar del Rio Liga Especial Reserva Superior Salomon 81.00
9. Alec Bradley Black Market Filthy Hooligan (2013) 75.80
10. CAO Flathead V660 Carb 83.80
What is interesting is the most consistently rated cigars averaged 89 points, while the LEAST consistent only averaged 76 points. If you see a lot of variability in a cigar's Geek Critique results, odds are its not very good or not going to be popular with most people.
The overall distribution of cigar ratings is a fairly classic Poisson distribution, and is not normal.
* The overall average is 86.11.
* The most frequently given score is 90 with 101 reviewers assigning this score.
* The standard deviation for any one review is 7.414 points. So the 1sigma score range is 78.7 - 93.5 points. By implication, a cigar with a single review within this range can be expected to be "about as good as any other cigar in this review data set." Cigars reviewed above 93.5 points could be considered "significantly" better than average.
(https://i.imgur.com/hwBI1aI.png) (https://imgur.com/hwBI1aI)
Thanks for reading, if you'd like to peek at the raw data, check out my badly organized and formatted spreadsheet with all 1031 entries!
https://docs.google.com/spreadsheets/d/11nVO64DxHQ6VMQGJQp9f9EN10MVl_JjOLUauwni5xHk/edit?usp=sharing
WOWZERS Diddy, you bored today?
Good info tho
Interesting info. Good job.
Interesting data. Thanks for putting the time in.
I was surprised to see the Evil Genius Cigars White Chapel Belicoso on the "worst of the worst" list. RDF and I recently discovered these this past Spring while visiting Famous Cigar Shop, and they were on the front counter being touted as a "new cigar made by a local guy". They were so good that we split a bundle! Maybe the recipe from 2014 was changed, or maybe these are the original cigars with 5 years of age on them and they have dramatically improved. Either way, we both loved them! :confused:
Very interesting results.
Dan the Man with a Plan for us geeks to Understand!! Well done my brother! :biggrin: :bigthumbs:
That is an awesome analysis Dan. Thanks for taking the time to put that together.
One of the things I like the most was when cigars got really consistent ratings, and what that rating is. Take for example the LFD NAS, it was consistently disliked compared to some others. :lmao:
Great info - thanks for doing the work.
That is very interesting. Thank you.
I know when I do a geek critique that I individually score each area and continue to make notes and adjust the number as I enjoy the cigar. So maybe about an inch in I start to assign numbers and then tweak those as I go. But when I get to the very end I add up the numbers to see what the final score is. I've done that on cigars in the past and thought, "82? I enjoyed it more than that." And then I go back and see if I can adjust numbers upward. And the reverse is true too. I've had some high final numbers and thought that the cigar just wasn't that spectacular and then I go back to see if I should lower some scores.
Part of issue with the weighted Geek Critique scoring is that it is weighted heavily toward construction, burn etc. Something that has been debated for a while now. Not that there is anything wrong with the Geek Critique model but it seems easy to skew the numbers.
I would love to work through some numbers pulling some of the either scores out and see what it looks like regarding flavor only.
Well done Dan, thanks for taking the time to do this. :biggrin:
QuoteMautrak - 9/24/2018 10:01 PM
That is very interesting. Thank you.
I know when I do a geek critique that I individually score each area and continue to make notes and adjust the number as I enjoy the cigar. So maybe about an inch in I start to assign numbers and then tweak those as I go. But when I get to the very end I add up the numbers to see what the final score is. I've done that on cigars in the past and thought, "82? I enjoyed it more than that." And then I go back and see if I can adjust numbers upward. And the reverse is true too. I've had some high final numbers and thought that the cigar just wasn't that spectacular and then I go back to see if I should lower some scores.
Yeah, I do this too. I'm not a fan of the way we weight numbers on certain categories. For example, why is "pre-light" 1-20 points? It's just a gauge of your first impression that lasts only a few seconds, and many of the attributes you are scoring in that category are also the same attributes under "construction". I'd be a fan of capping 10 total points for "pre-light" and give the other 10 points to "flavor and aroma". Just my 2 cents
QuoteBrlesq - 9/25/2018 6:03 AM
QuoteMautrak - 9/24/2018 10:01 PM
That is very interesting. Thank you.
I know when I do a geek critique that I individually score each area and continue to make notes and adjust the number as I enjoy the cigar. So maybe about an inch in I start to assign numbers and then tweak those as I go. But when I get to the very end I add up the numbers to see what the final score is. I've done that on cigars in the past and thought, "82? I enjoyed it more than that." And then I go back and see if I can adjust numbers upward. And the reverse is true too. I've had some high final numbers and thought that the cigar just wasn't that spectacular and then I go back to see if I should lower some scores.
Yeah, I do this too. I'm not a fan of the way we weight numbers on certain categories. For example, why is "pre-light" 1-20 points? It's just a gauge of your first impression that lasts only a few seconds, and many of the attributes you are scoring in that category are also the same attributes under "construction". I'd be a fan of capping 10 total points for "pre-light" and give the other 10 points to "flavor and aroma". Just my 2 cents
Agreed. Flavor 50 points, construction, burn, appearance 50 points.
Sorry to send your thread off on a tangent, Daniel. But I tend to agree with you and Bruce that maybe some changes are in order, and that seems to be supported by your analysis / number crunching. The tough thing then, and it should probably start in a new thread so this doesn't get totally jacked, would be comparing old to new. There would end up being a pre-scoring system change group of reviews, and a post-scoring system change.
Food for thought. And you started it Daniel!
QuoteMautrak - 9/25/2018 3:01 PM
Sorry to send your thread off on a tangent, Daniel. But I tend to agree with you and Bruce that maybe some changes are in order, and that seems to be supported by your analysis / number crunching. The tough thing then, and it should probably start in a new thread so this doesn't get totally jacked, would be comparing old to new. There would end up being a pre-scoring system change group of reviews, and a post-scoring system change.
Food for thought. And you started it Daniel!
Well yes and no. I think the scoring needs to be revised, but I don't think it will make a difference. People tend to think, 80/100 is only "average" and not necessarily great. 90/100 is the bottom of the really good. You'll see similar distributions in academic scoring. I'm willing to bet you'd see a similar distribution of scores for cigars reviewed by Cigar Aficionado. I'll try to see if I can't scrape some data and do a comparison.
QuoteBrlesq - 9/25/2018 8:03 AM
Yeah, I do this too. I'm not a fan of the way we weight numbers on certain categories. For example, why is "pre-light" 1-20 points? It's just a gauge of your first impression that lasts only a few seconds, and many of the attributes you are scoring in that category are also the same attributes under "construction". I'd be a fan of capping 10 total points for "pre-light" and give the other 10 points to "flavor and aroma". Just my 2 cents
Could not agree more :bigthumbs:
Algebra. Statistics.
QuoteKid Montana - 9/25/2018 7:04 PM QuoteMautrak - 9/25/2018 3:01 PM Sorry to send your thread off on a tangent, Daniel. But I tend to agree with you and Bruce that maybe some changes are in order, and that seems to be supported by your analysis / number crunching. The tough thing then, and it should probably start in a new thread so this doesn't get totally jacked, would be comparing old to new. There would end up being a pre-scoring system change group of reviews, and a post-scoring system change. Food for thought. And you started it Daniel!
Well yes and no. I think the scoring needs to be revised, but I don't think it will make a difference. People tend to think, 80/100 is only "average" and not necessarily great. 90/100 is the bottom of the really good. You'll see similar distributions in academic scoring. I'm willing to bet you'd see a similar distribution of scores for cigars reviewed by Cigar Aficionado. I'll try to see if I can't scrape some data and do a comparison.
Don't disagree Dan, while the scores may stay in balance with historical trends, the score would better reflect the reason we all smoke cigars and that is flavor/taste.
I am one who wouldn't mind seeing the ranking weights change. My suggestions would be
Pre-light 10 - I don't that many of us would offer a cigar to critique knowing it was going to look bad, or even send a bad one out
Lighting and Burn 20 - burn I find more frustrating than construction, so I added 5 pts here
Construction 20 - I took 10 away from here, mostly because of the same point I made regarding pre-light, but also I tend to have much fewer construction issues vs burn
Flavor and Aroma 50 - Need to be the heavyweight of the rankings