Back in January, we hosted the third version of SparkFun Free Day. This year, we gave away $200,000 worth of $100 SparkFun credits over the course of the day. Once again, it was mayhem.
This time around, we wanted to see how people experience Free Day from the "other side", so we sent teams to Virginia, Texas, and Connecticut visiting North Street Labs, ATX Hackerspace, and New Mountain Innovations. We hung out with them on Free Day, documented their experience, toured their labs/spaces, and generally disrupted their lives for a couple days. It was a lot of fun and, again, we thank all of you who let us crash your party.
So today, we have the final installment of our documentary (Vimeo version here). Watch as we join our customers/friends as they attempt to win a share of the prize and get glimpse behind the scenes at how Free Day goes down at SparkFun HQ. If you missed Part I, you can check it out here. Thanks again to all those who participated in Free Day and congratulations to those who won!
Reading the comments I can't believe how many people seem to hate free.
Next year lets do Double Price Day so we can give those people something to appreciate.
I'd be satisfied if the Free Day would just stop resurfacing months after the event.
Why does it keep receiving air time? The results were announced in January. Some people won; some people didn't. Great, but it's done. Get over it. We don't need inane "documentaries" on Free Day. It's almost masturbatory.
My time is not free. We were spending time to save money.
At some point, it made more sense to spend time doing my normal job a bit longer to get the extra dough. I decided I'd much rather code than solve captchcas.
It's an intellectual rationalization, and not everyone sees it that way.
However, it is a great experience and caused us to interact with spf more. Even this comment about the event draws more attention to spf.
It's like The Game: By thinking and discussing about Freeday, we 'lose' and end up helping spf. Not sure that's losing, but anyways.
Can I pay SparkFun to work at Sparkfun? Please?
If I had the money......
Yet another "news" article on Free Day?
What's with the massive overdose of out-of-focus shots?
All those capcha's... absolutely worth it! Thanks for your free-day!
I hated this years FreeDay. So boring.
I preferred the quiz. That one was fun.
There was a train? People could watch the train? I thought it was a geiger counter!
Can't believe how tiny the age window to work in a place like that is. That tiny period between 22-30 is it.
FreeDay was awesome, no matter how it is implemented. I liked that it was different, that we didn't know what to expect, you should keep it that way and it'll allow for new things each year.
6.5 Million Captchas, 2000 winners: odds of winning = 1 in 3250. Say 10 Captchas per minute; it would take 5.5 hours to win (on average). Free day lasted 6 hours; should have won, if not for the mental breakdown. Still feeling uncomfortable every time I see a captcha, and longing for next FreeDay!
Dude, your numbers are misleading, and I don't know if I have the patience tonight to clarify sufficiently. Most nights I would.
Assuming your numbers are right (sorry, I don't remember) your post hoc chance of winning on your FIRST captcha was p=1/3250 (IIRC, at least one guy on IRC reported doing so -- but most who did would be unlikely to stick around, except for giggles) At 10 captchas/min, you'd have a 50% chance of winning after 2252.38 tries (about 3 hrs 45 minutes) or a 2/3 chance of winning in 6 hrs, assuming my watch calculator carries a sufficient number of digits of internal precision (12-14 IIRC) and I typed the digits correctly. It's a Poisson distribution or 1-p, because the right way to calculate it is by iteratively multiplying the odds of losing (3249/3250) not the chances of winning. The # of trials for 50% is e^(ln(0.5)/ln(1-p)) [or 10^/Log, if you prefer that to e^/ln] The chances of NOT winning in 3600 tries (6 hrs @ 10/min) are e^(3600* ln(1-p)) If you don't get why, I hope someone will explain. It just won't be me. Not tonight. Sorry. Also I may have it wrong (hardly unheard of), but that number accords with a lifetime of experience, and my Free Day observations. I ran the math in Sparkfun's IRC chat when I was in a better frame of mind, and a lot of folks got it.
I realize that doesn't dull your pain, and I sympathize
Also: kudos to Sparkfun on your site's excellent automatic superscripting of exponents. None of the servers I run handle that as well without explicit superscript tags -- but dammit, Jim, I've been a physician, not a programmer, for decades.
Point is, after six hours, the odds are very much in favor of anyone winning, if they stuck with it for that long (ie much better than 50 percent -- as you pointed out).
Here's an idea: sale day.
Who knows how your popularity will grow, but by free day time next year, I figured(with avg sales/day, using order # and days between my orders) that there will be 2100 orders/day(about 630 right now, I figure. How close am I Sparkfun?). Boost that by about 400(since you'll be having a sale), and you've got 2500 orders. If you give away $250,000, you could take $100 off of every order over $150(or even over $100 if you're feeling REALLY generous) for the first 2500 orders.
How many WRONG CAPTCHA?
Hey Sparkfun! Would I be able to get an audio snippet of the monome in action during free day? It was really pleasant to listen to.
I'm going to write a blog post about how it works at some point in the next couple weeks. As of right now the only recording I have of it is from a test before Free Day. You can find that here: vimeo.com/34868946
As much as i hate CAPCHAs ill do it over if it means i get another $100 free. :-)
I wish i got the first 100
Thanks Emcee!