December 10th is Eyewire’s 5th birthday. Five years..I can hardly believe it! It seems like just yesterday that we were planning a launch and hoping people would play. I’m so glad that you did! Three cheers to YOU, the Eyewire community. Without you this endeavor of citizen neuroscience wouldn’t be possible. On behalf of the whole team at Eyewire HQ, thank you!
To celebrate, and in addition to helping to drink the wine Nseraf sent, I took a trip through the archives while Kelv queried a bunch of stats to put together a summary of the impressive effort that this amazing community has put into Eyewire.
10 million cubes, WOW! Back in the early years it took weeks to finish a single cell. Now players complete multiple neurons per day.
By the end of the day of launch, Dec 10, 2012, there were 10,455 Eyewirers. Now there are 265,000 accounts and Eyewire grows by a little over 2,000 players per month. Amazingly, 138 of those original players have submitted cubes in the past 6 months. Thank to @hcom3 for suggesting this query in chat. Let us know if you’d like more stats than the ones below!
None of these cells would be completed without the careful insight of expert tracers. Before Eyewire, this was limited to the lab and for the first year of the project, oversight was conducted by admins at MIT. Then, in December of 2013, we launched the first player roles: Scouts and Scythes. From that point on, the game was never the same. Half a million reaps; 1.5 million completed cubes.. hats off to you!
You might be surprised that only 3,350 players beat Level 2 to unlock the excruciatingly difficult yet enticingly double-pointed cells. This just shows how tough the L2 challenge is. Of the 52,000 players who passed the tutorial, only 6% go on to pass Level 2.
But those who do have a special knack for Eyewire: 9% of L2 players go on to become Scouts. Only 40% of Scouts make it to Scythe, the top rank of player up until the recent launch of Mystic. A swan dive into the database reveals the very first reap to be… inconclusive. It was either @susi or @a5hm0r. Share the glory 🙂
To date we have completed over 3,000 neurons in Eyewire. For a sense of how our rate of progress has improved consider: it took us from Jan 2012 (public beta) until March 2014 to reconstruct 100 neurons. In November, 2016, we reconstructed 111 neurons. Even though many cells are smaller nowadays, this is generally representative of the extraordinary evolution of Eyewire.
On average, 3,805 cubes are submitted in order to complete one neuron. This is a huge improvement from the earlier days — originally 10-20 players did each cube. Now that number is capped at 4 enfranchised players. The average number of unique cubes per neuron is around 770. It is significantly decreased due to the recent smaller cells and bipolar fields. Compare these average numbers of cubes to the biggest, baddest neuron we’ve ever mapped:
Lightning 118 is one doozie of an cell! It’s type 82wo and you can check it out in the museum here. Or jump to Cube IDs #667258 and #667583 to check it out in Eyewire.
Eyewire is more than just stats, of course. Many players as well as folks at HQ have been here since the start. I went through all the history, and I do mean all of it – the blogs, github, social media, and even had some good old conversations – to compile this timeline of the birth and growth of Eyewire from a finicky, crashing baby of a website to a socially complex, stable(ish) kindergartener. Hope you enjoy this trip down memory lane! I encourage those of you who haven’t been on the forum lately to check out this post that players compiled showing the evolution of the Eyewire UI. For the devout players, I’m happy to share a link to this comprehensive history of all things Eyewire.
Thank you for being a part of the Eyewire community. We’re honored that you choose to spend your time playing and look forward to the next five more years. For science!
Illustrations by Daniela Gamba