Belgian Chocolate Rankings: Brugge Edition
The second edition of our Belgian chocolate rankings!

After the wild success of our previous Gent chocolate rankings, we have returned with new, super official Belgian chocolate ratings! This time, we will be reviewing the chocolates we purchased during our trip to Brugge.

The Chocolatiers

This time around, we purchased chocolate from four different sources:

  • Neuhaus: a Brussels-based chocolatier founded in 1857
  • Pierre Marcolini: Established in Brussels in 1995
  • The Chocolate Line: Opened in Brugge in 1992
  • Mystery box: We also had purchased some generic truffles in an unmarked white box from a shop. We can’t remember the exact shop, but they said the truffles were fresh in from someplace nice…

Ratings Overview

Method: Chocolates were scored from 0 (worst) to 10 (best) in increments of 0.5 by Katherine and Jacob separately and (usually) independent of knowing the other reviewer’s score. The tier list below is based on the average of our two rankings and the specific tier thresholds are shown in the next subsection.

Without any further adieu, here is our tier list for the Brugge edition of the chocolate rankings! This follows the same rigorous framework we established in the Gent study.

Chocolate Rankings

We had a three-way tie for the top ranking chocolate between the Hazelnut Gianduja Cone (Neuhaus), the Violet (Neuhaus), and the Generic Truffle (Mystery box). Here are our notes on each:

  • Hazelnut Gianduja Cone: “Creamy and delicious, basic chocolate kinda like the rectangle bars. Maybe hint of nuttiness / nutella? Kat really likes it, but thinks the cone shape is unnecessary but cute”
  • Violet: “kat: ‘I would have that one a lot’ — darker chocolate, floral”
  • Generic Truffle: “yum, melty”

On the other end, three chocolates stood out as outliers in the “Chocolate Hell”. The worst was the Troïka (Neuhaus), the second worst was the Citron (Pierre Marcolini), and the third worst was Apero (The Chocolate Line). Here are our notes on these unfortunate specimens:

  • Troïka: “Gross, kinda tastes like grandma’s 10-year-old sugar cube, Kat: ’not even chocolate, sugar play-dough’”
  • Citron: “Like our bathroom cleaner - not nuanced. Or sweet medicine”
  • Apero: “Flavor a little off - it tastes like medicine”

Tiers Histogram

We separated the tiers using the following bounds:

Tier Range
“The Belgian Brilliance” [8.25, 10]
“Majestic Morsels” [7.25, 8.25)
“Decent Delights” [6.25, 7.25)
“Mediocre Melts” [5.25, 6.25)
“Chocolate Hell” [0, 5.25)

As you can see in the chart below, most of our chocolate ratings fell into the “Decent Delights” and “Majestic Morsels” tiers, each making up about a third of all the chocolates we tried. This distribution isn’t what you’d call “normal”—there’s a noticeable tail at the lower end of the ranking scale.

Tier Histogram

Detailed Analysis

To augment our analysis, we have evaluated the chocolate rankings by chocolatier, reviewer (Kat vs. Jacob), and date.

By Chocolatier

For this analysis, we have excluded the mystery chocolatier, as there was only one sample of chocolate flavor collected. Below, we see that grouping the mean ranking (Jacob and Kat ranking average) by chocolatier reveals a significantly higher mean ranking of Neuhaus versus Pierre Marcolini. The Chocolate Line did not have significantly different rankings compared to either of the other two chocolatiers.

Chocolatier Boxplots

Note: ranking differences tested for statistical differences with Mann-Whitney U rank test. 
*   =   p-value<0.05
ns  =   not significant

By Reviewer

Of course, we were also curious to see how similarly Jacob and I were rating the chocolates! Below, a scatterplot is shown with Jacob’s score versus Kat’s score, where each point represents one chocolate.

Our ratings were moderately correlated with a Pearson R of 0.43 (where the possible range from perfectly anticorrelated to perfectly correlated is -1 to 1).

Reviewer Scatterplot

Most Controversial and Agreed Upon Chocolates

Among our ranking comparisons, two cases were of special interest: 1) chocolates we ranked the most differently, and 2) chocolates we assigned identical ranks.

  • On the left, the top five “most controversial” (largest magnitude of score difference) chocolates are shown.
  • On the right, the six chocolates we scored the same are shown in order of descending rank from top to bottom.

Controversial and Identical

Our most controversial chocolate was The Chocolate Line’s Chill Pill. Here are our notes for this chocolate:

“Grass flavored, reminds jacob of harry potter jelly beans ‘6 or 10 depending on appetite for weirdness’ -J”

The top ranking chocolate that we scored the same was the Appeltje, which also came from The Chocolate Line. We noted:

“Reminded us of those caramel apple suckers but much higher quality.”

The lowest ranking chocolate we ranked the same was Pierre Marcolini’s Du Jour truffle. Our detailed description:

“It’s alright.”

By Date

After a particularly disappointing tasting on May 14, I formed a hypothesis that our chocolate ratings were declining over time due to a deterioration of the chocolate’s freshness. However, there was no significant relationship found between the rankings and time tasted, so we probably just had bad luck with our selection on May 14. Here is the plot of chocolate rankings over time nonetheless.

Time Series

As one final analysis, a word cloud of our rankings was compiled and is the image at the top of this page!


Last modified on 2024-06-15