Presumably because of the success of the NetFlix Prize, ESPN decided to hold a hundred thousand dollar purse to see who could provide the best algorithm to predict the outcome of upcoming college football games.  Fortunately, ESPN went looking for experienced help to design such a game.  Unfortunately, they chose TopCoder.

It’s a shame that ESPN chose to do this through TopCoder, as TopCoder’s general practices are poison for a machine learning contest. TopCoder chose to impose a gig memory limit and a nine minute runtime on any approach to this problem, which murders most machine learning tactics right out the door. It’s a shame they didn’t do this themselves on the NetFlix model, where contestants just submit predictions.

This contest isn’t to get football predictions. It’s to get football predictions under arbitrary ram and cpu caps. ESPN’s staff wouldn’t face such restrictions when using the work – one gig for nine minutes?  C’mon; there is literally no reason for this limitation to exist.

This contest’s design precludes most modern approaches to machine learning to no appreciable benefit, and is therefore fundamentally flawed. ESPN is going to get seriously quality-limited results.

Very disappointing.  That money would go to much better effect if the contest had been designed with the kind of foresight of which the NetFlix Prize had had the benefit.