Tom Scocca's dissection of Bill Simmons' columns ("pregame analysis of the postgame analysis") is brilliant, and I really loved this piece by "The Machines" about the use of instant replay in two late-game decisions in Game 3 of the NBA finals last night:
The Machines perceived a tone of dissatisfaction, then, from the TELEVISION announcers. The correct call had been made, but the humans were not content with it.
Why would humans prefer ERROR? Was this a problem with MACHINE-MEDIATION PROCEDURE? Or was it a problem with the limited parameters under which MACHINE-MEDIATION PROCEDURE was allowed to correct for ERROR? The Machines strongly advise the further incorporation of MACHINE-MEDIATION PROCEDURE into the basketball process.
Eventually, for reduction of ERROR, humans may consider replacement of human basketball referees with MACHINE REFEREES. Consider also the possibility of MACHINE PLAYERS. Modern technology can produce an entertaining "natural"-appearing variation in mechanical-physical results, or something close to it. It is our understanding that the algorithms employed in prototype basketball-shooting unit RAY ALLEN may have in fact gone beyond plausible human variation. We intend to recalibrate RAY ALLEN unit shortly.