Wine in EDA vs Bullshit.
In the context of my information science course reading, “On Bullshit” by Frankfurt distinguishes between “bullshit” and misinformation or propaganda. Frankfurt posits that the primary distinction lies in the fact that individuals engaging in bullshit do not concern themselves with the veracity of their statements; they may not even subscribe to the beliefs they articulate. The central tenet of bullshitting is not the intentional persuasion of others to embrace a particular viewpoint. Conversely, misinformation, whether deliberate or inadvertent, involves the dissemination of incorrect information with the aim of inducing belief in unverified claims.
In my perspective, the validity of statements categorized as “bullshit” can often be debunked through Exploratory Data Analysis (EDA), as numerical data tends to be impervious to deceit. Growing up in Northern California, just a 40-minute drive from Napa Valley, I was encouraged by my parents to cultivate a basic understanding of fine wines. This led me to enroll in a high school wine-tasting course, where I acquired knowledge about how factors such as location, vintage, and temperatures influence the nuances of wine produced by Dane cellars. This early exposure ignited my passion for wine, extending from the intricacies of production to the art of bottling.
Having frequently encountered conversations at dining tables revolving around statements like “I possess a bottle of wine from [XXX] country, vintage [XXXX], acclaimed with [XX] points,” I decided to embark on a data-driven exploration. I aggregated a comprehensive dataset of 200,000 entries from “Data World” to conduct a thorough analysis. My objective was to ascertain which countries consistently produced the highest-quality wines. Furthermore, I aimed to investigate whether the quality of wine correlated with its market value. Lastly, I sought to identify the price points and associated quality ratings that align with affordability for a broader audience.
I utilized multiple different softwares for my EDA, including python, R, and Tableau. The first interesting finding that I found was the correlation between prices and points. Imagine if the correlation between these two variables were high; it would imply that there is no inherent cost-performance ratio across different wines. A high correlation would suggest that higher expenditure unequivocally results in superior quality, thereby challenging the notion that cost directly dictates the caliber of wine — a belief I seek to disprove.
However, a closer examination of the data unveils a correlation coefficient of only 0.408 between price and points. This value falls below the conventional threshold of 0.5, indicating a relatively weak correlation. This observation serves as a compelling demonstration that possessing a bottle of exceptional wine does not necessarily entail a hefty price tag. In essence, the data refutes the oversimplified assumption that the quality of wine is intrinsically tied to its cost.
Based on my analysis, I recommend individuals who relish dining with wine to consider acquiring bottles with a rating of 83 points, priced at $12.6. For those with a penchant for high-end selections, I suggest exploring wines with a rating of 98, even though they come with a higher price tag of $198.2.
This recommendation is substantiated by the data presented in the chart, where the 83-point category emerges as the most cost-effective choice, offering superior performance in relation to its price when compared to wines rated 80, 81, and 82. Notably, this implies that wines achieving an 83-point rating are not only more affordable but also outperform their counterparts rated in the lower 80s.
An intriguing pattern revealed in the chart is the robust correlation between price and performance after reaching a point rating of 90. However, the exceptional case of 98 points challenges this trend. For instance, the price gap between wines rated 99 and 100 is $290, while the difference between 97 (priced at $192.37) and 98 (priced at $198.2) is a mere $6. In the context of a nearly $200 investment in a bottle, this nominal price difference becomes inconsequential, offering a considerably enhanced performance with the 98-point wine.
Another bullshit I hope to be answered by EDA is that countries such as wine from Italy and France are expensive and luxurious. In reality, if you look at the graph, the top three countries that are having the most expensive wine are Switzerland(avg price 99), England(avg price 62) and Germany(48); moreover, the 3 countries that have highest quality of wine are England(91.82), Australia(90.07), and Germany(89.83). Thus, England is in fact the country that sells high-end, and high quality wine not France or Italy. And my recommendation via the analysis would be do not buy wine from Switzerland, because it has the most expansive price but the quality of it cannot even make it to top 20. On the flip side of the coin, Australia has the second best of wine worldwide with a very affordable prices.
In conclusion, with my analysis and visualizations, my recommendations for wine purchasing is that 83, and 98 points of wine should be considered if you are looking for affordable prices for dining wine and high-end wine. Australian wine is suggested due to its quality and prices, and stay away from wine that is made from Switzerland. Last but not least, my analysis has refuted two bullshit theories of wine. According to the analysis, it is possible to buy good quality of wine within a affordable price; also countries like France and Italy’s wine are not that luxurious compared to other countries, in fact England is the country that should come out from your mind.
Data retrieved from: https://data.world/datasets/open-data
