On the subway today, my logic was impeccable.
There are four trains that run from the southern side of Penn Station to Times Square: numbers 2 and 3 on the express line and numbers 1 and 9 on the local line. The lines use different platforms. During rush hour, I must guess which platform to choose so that I waste the least possible time catching the next train. One and 9 run a little more frequently but 2 and 3 require less walking on my part. Moreover, if a train has just left, I must allow for the decreased likelihood of another train pulling up to that platform. Throw in the general existential uncertainty of the New York subway service and the belligerent jostling crowds that impede movement in any given direction (always the direction I take, for some reason) and you are faced with a dynamic optimization problem of dizzying complexity.
Yet my logic was unbeatable today. To wit:
I get up to the express platform and spy a number 2 train pulling from the station. It's almost empty, which means there weren't many people waiting for it when it arrived, which means (it being rush hour) that most people have departed on another train that immediately preceded this one. Conclusion: two express trains have just left the station within minutes of each other and it will be some time before the next one comes. Express trains run less frequently and judging by the crowd on the local platform, it's overdue a train anyway. Thus, my wait time should be reduced if I abandon the express platform in favor of the local one.
I proceed as logic dictates and quickly catch a number 9 train—just as another number 2 pulls up on the express side. The 2 leaves almost empty. Squeezed in by the shoulders of strangers, I watch it go with melancholy. My train doesn't move; presently they announce that it's being held at the station for police investigation.
I stand and ponder the usefulness of logic. The hands on my wristwatch move forward slowly, inexorably.