This project explores KCBS competition data from 2013-2016. I used Python (selenium, beautiful soup) to pull the data from the KCBS website. Data analysis will include choropleths (colored maps) based on competition meta-data, and state-by-state score data comparisons.
The visuals folder contains choropleths on three measures in the data. These maps were made with Tableau. They include;
In Progress...
Plotting the cumulative distribution function of category scores from the 5 most competing states (Missouri, Kansas, California, Tennessee, Georgia) seems to show stronger scoring in Georgia and Tennessee.
Comparative CDF plots of category scores
The similarity between plot lines correspond to geographic proximity and/or the number of contest held in each state (Georgia and Tennessee are 4th/5th).
The higher scores in Tennessee and Georgia may be a significant effect, but additional testing needs to look at the significance of these states' distributions while controlling for the sample size. Another culprit could be regional scoring biases.
Creative team names are a tradition of the competitive barbeque circuit. Using the team names from my data set, I created a wordcloud to visualize the themes of these names. An automatic name generator might be a fun challenge, but the names are often pun-y which could be hard to automate.
A full list of names can be found on the KCBS website.
Competition Data: http://www.kcbs.us/events.
A change in KCBS scoring standards went into effect on July 2013. Competition data is limited to contests after that date.
Scoring Information: http://www.kcbs.us/news.php?id=687
Feel free to reach out for csv/pickle files (pandas dataframes). The largest file (~50mb) holds every competition from July 2013 onwards, and includes score result sub-tables (chicken, ribs, pork, brisket, overall).