31 days, 103,342 tests, 15,700 new words learned… $1,000 won!
ShuylersNewJob entered the Memathon with a sense of purpose (hence the username). Today we are delighted to announce that he has fulfilled that purpose, and entered Memrise folklore as the first ever winner of the Memrise Memathon!
Schuyler has in fact been a long term Memriser under the name “Khagda” and has created a range of courses in a bewildering array of languages, from Kurmaji Kurdish to Onge, Jarawa to Sentinelese, Andaman to Yucatec to… French.
But since graduating from high school in the summer and heading off to travel in Europe, his old account had become “a mess of half-completed courses and unwatered plants”. So when when he heard about the Memathon he decided to start a new account dedicated to the purpose*
When we chatted over Skype he told me wistfully of his pride when he reached a million points on his old account – unaware of how much more he could do if he really went for it!
He decided to focus exclusively on Hungarian and Kurdish for the Memathon – languages he is currently learning. He started fast. While others nursed new year’s day headaches, Schuyler answered nearly 11,000 tests on Memrise on January the 1st. An impressive start.
But after the first 10 days, putting in around 8 hours a day learning, he found himself still well short of the leaders. You can read his anguish unfolding in this reddit thread where he goes by his old Memrise username, Khagda.
So on the 10th of January he was faced with a tough choice: continue to wrestle with the difficult orthography of the languages he was actually studying … or win the Memathon. He chose to win. Hungarian and Kurdish could wait until February – Schuyler switched to studying Spanish (then French, then finally Portuguese).
By the 15th of January, the leaders had started to draw clear of the pack. lucas.coffee and Kikisdeliveryservice had each learned nearly 12,000 new words and were blazing a remarkable trail. ShuylersNewJob was sitting in fifth place at the back of the leading group on a little over 9,000 words learned. The consensus at Memrise HQ was that the winner would have to come from among those 5 names – though some doubting voices suggested that Schuyler was already too far behind to chase the leaders down.
But as we now know, the plain facts didn’t quite tell the full story: once Schuyler had started learning Spanish, he was picking up new words at a remarkable rate. And not only was he learning new words fast, but he was also reviewing as he went along, meaning that the memories he was forming would stay healthy for longer, and he wouldn’t be faced with a review backlog at the end of the month.
He had started picking up pace in the middle of the month which brought him into contention. But it wasn’t until the last week that he really showed what he was capable of.
As lucas.coffee told me a little ruefully, “halfway through the month I didn’t think he (SchuylersNewJob) was a real threat… but then at the end he just came so fast! That last week was insane!”
It was a bit insane: 11,000 tests must have felt like a lot on January the 1st. But during the last week of the month SchuylersNewJob averaged 14,763 tests per day.
Before the start of the Memathon I tried to work out what the limits were of the number of tests that a person could take in a day. I spent several hours seeing how many I could do in an hour if I really pushed myself. By learning a “no typing” Spanish course and stretching my concentration to the limits I found I could do around 1850 tests in an hour. I wondered if anyone would be capable of keeping up that level of focus for 14 hours in a day to hit 25,900 tests. I reckoned that that would probably be around the limit of human endurance.
SchuylersNewJob has answered that question rather precisely: on the 31st of January he managed an utterly mind-boggling 25,879 tests. His powers of concentration are bewildering. He was also fighting a physical setback – after answering over 75,000 tests already in the week, his right hand was suffering from the strain of repeatedly hitting enter. So Shuyler spent that last epic day with his knackered right hand soaking in a bowl of warm water while he pounded the keyboard left handed.
That astonishing final dive for the line took the number of new words that he learned in January that were healthy at the end of the month to 15,700**. Nearly 800 clear of lucas.coffee in second place. Victory was his!
Massive congratulations to SchuylersNewJob, to lucas.coffee and kikisdeliveryservice – and to everyone else who pushed themselves a bit harder than usual in January. It was an incredible month. Here’s the summary of what was achieved:
* We’ve crossed-checked with his old account and we’ve removed double-learned words from the total count, though there were only a very small number of them
** This inevitably leads to the question of what exactly he has learned, and whether it was worthwhile to ‘learn’ 15k words through multiple choice tests over the course of a month. I think the answer has to be, “we don’t yet know”. As Schuyler himself is quick to point out, “I didn’t expect to be able to actually remember all those words… its kind of an odd experience. I’d call it passive intelligibility.” It will be interesting to see exactly how much is retained – as far as we know there has been no similar study on learning such large number of words over such a long period of time. We’ll be hooking Schuyler and some of the other leaders up with the Psychology team at UCL to see if we what we can learn from their experience.