The 2013 MIT Sloan Sports Analytics Conference had a lot of key takeaways: That Larry Sanders is a great defender, that optical tracking data is the bomb, that as important as being smart is getting your point across in an accessible way.

Underrated, however, was a fascinating finding from smart guys Justin Rao (of Microsoft) and Matthew Goldman (from U.C. San Diego). Their presentation was about when teams should shoot 3s, but nestled in the paper was new and powerful evidence that in crunch time, the best teams get even better.

Yes, this confirms what a thousand jocks and pundits have been saying for a century. That's part of the reason it's so surprising! But also because, by and large, it has been tough to come up with real evidence that anybody could truly consistently raise their games in this way. And they have interesting ideas about what's causing it all.