2018 Scripps National Spelling Bee: Analysis

As a coach and a bee aficionado, I like to keep a close eye on the percentages and statistical trends that one finds through the careful analysis of words and their origins as they are given on the national stage. The 2018 Scripps National Spelling Bee was incredible, both from an entertainment perspective and an analytical perspective, to say the least. It was both fantastically interesting and perplexing all at once; everything I predicted would characterize the 2018 bee was incorrect.

Allow me to elaborate by stating some trends I expected to observe in this year’s national bee. I expected, first of all, for root-based words (Latin & Greek words) to constitute about 40% percent of the words given; I observed, over the past five years, that the percentage of Latin and Greek words given in the bee had continually dwindled from well over half of the bee and to approximately 40%. Second, I expected about 20% of the list to be of French origin, as French words had been steadily on the rise over the past two years, leveling out at around one in every five words. Third, German was expected to decrease, and Italian was expected to take its place. Lastly, I anticipated an increase in the percentage of eponymous words, as more and more eponyms have been asked in each successive year in the past.

As you can see, my 2013-2017 graph was blown absolutely out of the water by the statistics gleaned from the 2018 word list. To summarize, Latin and Greek’s percentage shot back up to 52.4%, a number it hadn’t reached since 2014. French’s overall presence was less than in previous years, its percentage falling from 20.4% to 16.5%, which is not a terribly significant drop, but it was unexpected. Shockingly, German increased by one percent and Italian fell by one, the opposite of what I had predicted. The percentage of eponyms fell from almost 12% to about 7%, a pretty significant decrease in the difficult category. How is one supposed to explain this change? The first instinct of many would be to blame the extra competitors brought to Scripps by the RSVBee program, but I only account for rounds 4 and beyond in my analysis, and by that time, the excess spellers had been eliminated and the bee was back to normal size. My only guess at why all my trends and predictions were broken is the fact that so little of the championship list was used in the 2018 Scripps bee. In the four years prior, all or almost all of the championship list was used in the finals, adding a variable to my statistics that I hadn’t ever considered or accounted for. My hypothesis is that, because we saw only three of the championship words in the 2018 list, the percentages of the etymologies given didn’t get a chance to level out and conform to the trends exhibited in 2017 and the years before.

On another interesting note, something I noticed in this year’s bee that I hadn’t in other years was the presence of repeated words, and by that, I mean words that had been asked in the Scripps National Spelling Bee before. For example, nastaliq had appeared in the 1997 Scripps finals, propylaeum and chrystocrene both were asked in 2008, arrhostia was given in 2009, epipodiale and bondieuserie were asked in 2011, even bewusstseinslage was given on the 2007 written preliminary test at Scripps. These are just a few I happened to notice, I’m certain there were others. This phenomenon is definitely something to watch out for in future national competitions; it wouldn’t hurt to be familiar with the words asked in past national bees, as they could be asked again.

So what do all these numbers mean for spellers? The obvious implication is the importance of studying Latin and Greek roots, as those two languages together make up over half the bee. It also demands that spellers have a strong understanding of French and a broad knowledge of eponymous words. It reveals that a familiarity with past national lists can be beneficial. Of course, this is not all a speller needs to know, and statistics are not always entirely trustworthy, but these percentages and trends help a speller to know what to expect to encounter at the national bee and provide a factual basis for their studying. To all the spellers reading this, I wish you the best of luck in your preparation for next year’s big bee and I hope these numbers provide some valuable insight!