CASP9 Refinement Puzzle 2

Now, here’s a clever science game.  A game that actually generates valuable scientific outputs.

Foldit is a game from Seth Cooper and his colleagues at the University of Washington where players score points by squeezing as many  proteins as possible into a chemically stable configuration.  Understanding how proteins can fold together is essential to establishing bio-chemical processes and hence the creation of new drugs.  This is significant work.  Performed by gamers.

The team have just published a paper in Nature and their accomplishments have been picked by the Economist and others.

A few things jumped out at me as I read the research:

1. Players have performed better than the best available software algorithms.
2. Few of the highest scores come from experts
3. Many of the best results are reliant on collaboration.

Of course, none of these things are especially surprising but they are exciting nonetheless.  Although players haven’t proved as good as the algorithms at folding proteins from scratch, they have excelled at risk-taking – remodelling using structures that are temporarily unstable – a strategy dismissed by the software.

And enthusiastic amateurs have produced some of the highest scores.  This might just a consequence of the number of participants – there might be relatively few biochemists in the player pool or it might indicate that the ‘fools’ offer real advantages.  It’s not an insult.  The liberating aspect of being an enthusiastic amateur  is not knowing ‘the rules’ or being confined to play by them.  This has resulted in multiple novel strategies as the ‘unschooled’ have experiemented wildly.  It’s a winning approach.

And finally, many of the best scores have been the result of teams of strangers working together.  The effectiveness of collaboration demonstrating that again that (for adults particularly) dialogue is the core of learning.  It’s also an example of Clay Shirky’s Cognitive Surplus at work.  Together we’re capable of more.

Foldit epitomises many of the advantages of using games to serious purposes – it’s team-based fun, it’s unconventional and it’s generating real transferable scientific value.  Good game.