Monday, July 25, 2011

optimize

at first glance, there's nothing remarkable about the following picture.  it's just a bunch of people riding the red or purple line to ... well, somewhere.  if you ride everyday (or even close) though, then perhaps you'll see what i saw in this scene.




on the near side (closer to the camera) of the train, you've got about eight passengers (some, including me, not pictured).  on the far side, you will see precisely zero (if it's not evident in the picture, you'll have to take my word).  this is not the usual spatial distribution of passengers.

turns out the next stop is the end of the line.  it also turns out that the near door is the closest to the platform's exit on this car (i'll leave it to the reader/rider to figure out which line and which end i'm talking about).  all these passengers have these two pieces of information within their personal crystal balls.  when dealing with the statistical world of bus arrival times, these sorts of optimizations can save up to an hour, though that may all change with nextrip.  you'll often see very specific train doors more crowded than the others.  i often wondered why, only to find that the crowded door opens right to the escalator or stairs at a particular stop and most of the crowd empties.

sure enough, when the train stops, four of the people near the door scramble to the exit just a few feet away, bound up the steps two at a time and scurry on their way.  like i said, this can save up to an hour, but this time i see all of them on the platform at their (and my) connecting train.  at least they avoided the crowd piling out from the other cars.

No comments:

Post a Comment