Hello everyone! I just recently migrated from the old board, and I have to say I like the ambiance here much better. I'm definitely looking forward to another season where I can cheer on my Dukes! and I have a new proposition for my fellow JMU/other CAA friends.
Last winter, I created a website hosting the "FCS Playoff Challenge" that allowed FCS fans to fill out a "March Madness" bracket for the 2012 FCS playoffs (congrats again to JMU Harpoo, last year's champion!). Although the inspiration was last-second, I got a great response from all of you with over 60 people submitting a bracket (over Thanksgiving, no less). Given the popularity, I decided to throw together a new beta effort titled "CAA Pick 'Em Challenge". The website I created can be found here:
http://caachallenge.site40.net (I apologize for the shady-looking URL. Just like last year, I hosted my website on a crappy free domain server. Having to deal with their partially-generated URLs is the price I pay for it, I suppose). Everyone is welcome to participate, whether you’re a JMU fan, frequent poster, lurker, etc.
The concept is simple and is very similar to another popular thread that frequently airs on this board. Each week, everyone will have the opportunity to predict the winner of each CAA game (defined as a game that includes at least one CAA team). Each correct prediction is worth 1 point. Each incorrect prediction is worth 0 points. The person with the most correct predictions at the end of the season wins the challenge. This website is not meant to compete with the other CAA Prediction Thread. Please participate in both!
In addition to predicting each contest's winner, I invite you to use the "Confidence" sliders to indicate your confidence in your selections. That is, the confidence that you submit for each game reflects the probability that you think you've selected the correct winner. For this year, the confidence measurement is merely informative and does not impact your score. However, I may elect to use aggregate confidence scores to break any ties. For example, a participant who displays an aggregate confidence of 80% across all predicted matches and an 80% aggregate win percentage will be deemed to have outperformed another participant who displays a 70% aggregate confidence and an 80% aggregate win percentage because the first person displayed a better overall grasp on his/her predictive ability.
Now for the question I've been asked a dozen times in the past 48 hours: "But [Lyric], why do you allow users to enter a confidence value of less than 50%??" I allow this because I'm simply recognizing the fact that a participant may sometimes choose an underdog, perhaps for nostalgic reasons. A <50% confidence value allows the user to make the selection while acknowledging the improbability of the outcome. I mean, if you REALLY want to take W&M over WVU.....
The website is still under construction, but according to my initial testing, it should be ready to support your predictions. If I get enough interest, I will continue to build the site (rather than simply post the leaderboard on the CSN boards). Website links I hope to create in the near future are "Password Reset", "Insights", and the "Leaderboard". If all goes well, I plan to run plenty of analysis on the user data to obtain some interesting insights. For example: Are users generally overconfident, accurate, or under-confident? Do user predictions get better throughout the season? Do prediction dynamics differ for ranked and unranked teams? Do users display bias for their favorite teams? Are any particular teams habitually underrated or overrated (could be useful info for that next FCS Playoff Challenge...). I will publish any/all obtained insights on the website so that everyone can enjoy the output of this social experiment. If this website works out this season, I may attempt to expand it to support every FCS matchup next year.
Thanks again to all of you for your interest, both past and present. It makes the website building a lot of fun!