During the last 10 years, a few thousand software developers have participated in our coding contests. But it all started small. The curve starts with 50 participants in 2007 and ends with more than 1000 participants in 2016.
How many participants will we have at the 25th CCC on March 31st, 2017, when the contest takes place in more than 15 locations in parallel? In order to visualize it better, let’s have a look at the growth curve:
If you are wondering about the dip in 2010: Usually the coding contest was held on a Friday, but this time the Thursday before was a public holiday, so many pupils and students were already off for a 4-day weekend. That’s why we had fewer participants in 2010 than in 2009.
During those 10 years we experimented with various aspects around coding contests. Here are some of our learnings.
How long should a contest last?
We started with 4 hours, but we also tried shorter contests that took only 1, 2, or 3 hours. Hackathons can last anywhere from several hours to several days, and there are contests that last for 24 hours (e.g. Challenge 24).
The problems that we typically select for our coding contests are quite challenging. We break them down into a series of 5-7 levels. Understanding all those level descriptions and coding appropriate solutions takes time – even the best coders are busy for 3-4 hours. So the lower limit for the duration of our coding contests turned out to be 4 hours. We never tried to extend them. 4 hours seems to be enough.
Why should we have any physical locations instead of making it an online-only event?
We started offline. Since 2009, we also allowed people to participate via the internet. We thought we could reach many more people that way.
- onsite participation: you have to travel to a location; others see you; success and failure are public; you meet many other coders in a great location
- online participation: no travel, perfect privacy
We never did a lot of advertising for online participation. Actually, we incentivized onsite participation a little, e.g. we only had awards onsite.
If you take a look at the gallery you’ll see the energy and emotions of the onsite participants, as well as the great contest locations and understand why we and the majority of our participants like the onsite events better. Nonetheless, we still keep the online door open for now.
Interestingly, the fastest coders have always been onsite.
What about a participation fee?
In 2007 we had a participation fee of 15 Euro which we refunded for pupils and students after the contest (so only practitioners paid them). Basically, the participation fee was meant to make the registration more binding. Since collecting money and handling refunds was cumbersome, we discontinued the participation fee in 2009. Even without a participation fee we have less than 10% no-shows.
What about prize money?
In 2007 we handed out money to the three fastest coders (400 / 200 / 100 Euro). In 2012 we extended it to more than 12,000 Euro for the best 30 coders in Austria and 3,000 Euro for the best 15 coders in Romania (the contest took place in 2 locations at the same time).
- get more participants: realized
- get more media attention: not realized
- dozens of software developers traveled from Romania to Austria because the prize money was significantly higher in Austria than in Romania.
In the following years, we reduced the money again (7k in 2013; 5k in 2014) and discontinued it completely in 2015. Still, the number of participants kept going up.
Is it fair if pupils compete directly with experienced software developers?
For the first coding contest we had two separate rankings: one for pupils, one for grown-ups (students and practitioners). Before the contest started, one pupil asked whether he could also participate as a grown-up. In the end he was the fastest pupil and the second in the overall ranking. From that time on we only had one ranking for participants irrespective of their age.
But we got the feedback that the examples were too much of a challenge for inexperienced pupils with just 1-2 years of coding experience at school.
Hence we came up with the “School CCC”:
- Easier problem to solve
- Just 2 hours
- Participate either onsite or online from the classroom
- Larger teams allowed
- Teachers can get access to the problem definition ahead of time
Originally, the School CCC was run in parallel to the CCC. By now, we have the School CCC in the morning and the CCC in the afternoon. That way, the best pupils can participate in both.
How many locations can we serve?
In 2007, we started at the Johannes Kepler University in Linz (Austria). Four years later, we also organized the contest in Cluj (Romania) at the Babes-Bolyai University. Since then, the contest was typically held in two locations in parallel (one in Austria, one in Romania).
We do hope that scaling the event nationally and internationally will work fine. We are looking forward to the learnings from that scaling experiment. In the best case, we might have hundreds of locations in the future 🙂
If you have other questions regarding coding contests, just ask! => firstname.lastname@example.org