In the past, I’ve estimated that “fast” debaters speak at between 280 and 340 words per minute (WPM). This estimate was based on my experiences over many years working with high school debaters to improve their speaking, often including formal tracking of their WPM rates over time. It was also based on anecdotal reviews of debaters’ speech documents. When calculating WPMs, nearly all elimination round-caliber debaters seemed to speak somewhere in that 280-340 range.
While I was confident that my estimate was relatively accurate, I decided to attempt a more authoritative investigation into debaters’ speaking speeds. My goal was to answer two questions: (1) How quickly do “fast” debaters speak?, and (2) Have “fast” debaters slowed down for online debate?
To answer these questions, I downloaded open source affirmative speech documents from the openCaselist wiki from teams that cleared at the National Debate Tournament in 2019 (the most recent “in-person” NDT) and 2021 (the first online NDT).1
When possible, I used open source documents from rounds at the NDT. If NDT documents were not available, I used documents from another second semester tournament. Using this system, I obtained 44 documents from 23 of the 26 teams that cleared at the 2021 NDT and 38 documents from 17 of the 30 teams that cleared at the 2019 NDT. Teams that were not included did not share open source documents on the wiki or — in a few instances — formatted their documents in ways that prevented me from analyzing them.
After deleting everything except the 1AC (and making sure it was properly formatted), I used Verbatim‘s Stats2 feature to calculate the number of tags and highlighted words in the document.
After compiling this data in a spreadsheet, I calculated the 90th, 75th, 50th, 25th, and 10th percentile WPMs for the 2019 NDT, the 2021 NDT, and the combined 2019-2021 NDT. The results were as follows:
These results seem to suggest two conclusions:
1. “Fast” debaters speak at between 260 and 320 words per minute. This is slightly slower than my estimated (280-340) range.
2. Debaters do not seem to have slowed down for online debates, at least not significantly enough to show up in this data.
There are some important limitations of these results: my dataset was incomplete, 1ACs might be delivered slightly more slowly than other speeches (especially 1NCs and 2ACs), and I only evaluated data from one (college national championship) tournament.3
Nonetheless, these results can at least provide greater certainty about the range of speaking speeds demonstrated by the most successful college policy debaters. If a debater’s delivery exceeds 300 words per minute, they are indeed (to use the language of the Verbatim scale) “Super Fast.”
1. I decided to analyze 1ACs because they are the only reliably “complete” speech documents available. While 1NCs and 2ACs might be delivered at a faster speed — an important limitation of this study — they are often edited/amended in ways that preclude standardized restrospective WPM calculations. While it is possible that some 1ACs included “extra” cards that weren’t actually read in the debate, I am confident that this will have a relatively negligible effect on this dataset — especially given the reasonably large sample size.
2. The Stats feature — developed by Aaron Hardy for Version 5 of Verbatim, first released in 2015 — allows users to estimate how long it would take them to read a speech document out loud. This was a groundbreaking feature at the time, and it remains one of my favorite capabilities of the (still “industry standard”) Verbatim template in 2021.
Verbatim’s Stats function includes a link to an online reading test and the following chart of WPM speeds:
Super Fast: 450 WPM
Fast: 400 WPM
Average: 350 WPM
Pretty Slow: 300 WPM
Really Slow: 250 WPM
It has been obvious from the beginning that this chart is incorrect; it conflates “reading speed” with “speaking (out loud) speed.” Figuring out the proper scale was part of my motivation for conducting this research.
3. In the future, I would like to repeat this study for high school debaters — both experienced, national championship-contending debaters and younger, less competitively successful debaters. My guess is that the fastest high school debaters are generally just as fast as the fastest college debaters, but that a higher percentage of high school debaters fall below 280 WPM. Anecdotal evidence seems to confirm this theory, but a more comprehensive dataset would need to be created in order to be certain.