Considering daily commute: people change schedules, family emergency, accidents on the road, and emergency vehicles, all disrupt the traffics in the city. Despite all these and other random changes every day, there are actually patterns in the commute. We don't need software and computer to do the data analytics, everyone of us has a"supper computer" ! All one needs is to pay a little bit attention to what happens on the road, and maybe collect some data to back up observations.
My daily commute is about 15 miles to 15.5 miles deepening on route chosen. There are two types of routes, local streets, and local plus toll way. The following observations are for local drivings.
Traffic pattern changes during the day. If I leave home before 7am, local street + 289 is the fastest route most of the time; even if there are occasional traffic jams at one or two major intersections. If I leave home after 7am, using local streets without 289 is better. There are more nuanced patterns, such as relations between traffic lights, yellow light duration ....
Using trip meter B on my car, I also timed my driving accurately, from start of engine in my garage, to stop at parking lot at work. It turns out the best average speed I can get is ~ 30 miles/hour during day time.
Using these patterns, I have less stressful commute, no need to "hurry up and wait", no need to frustrate about 1 second late to beat a traffic light (since it will be the same I beat it or not due to the next traffic light); it will make my car's gas efficiency better as well. My SUV's real local gas efficiency beat manufacture's EPA number!
NFL season is half way now, and NBA season just started.
At the beginning of the season, Patriots lost 2nd and 3rd games of the season, some sports writers started to question if they were going downhill all the way. On the other hand Jaguar had a hot start, beat Patriots in the process, some people were asserting that they turned around. There are many such "jump to conclusions" comments from sports writers and fans. My observations over the years of following NFL records is that, it takes 8 games, i.e. 50% of total games played to have a good indication on how an NFL team will perform for the season. For the aforementioned teams, after 8 games, Patriot is on top of AFC east, and Jaguar is at the bottom of AFC south.
Similarly in NBA, my observation from following the win-loss records is that it takes 20 games (25% of total games played) to have a good indication on a team's season performance. I saw the "jump to conclusions" by sports writers on Thunders at the beginning of the season, (0 - 4 start) ,they are now 7- 5. Even the NBA elite player like LeBron James, was not immune to the "jump to conclusions", when Lakers was at 1-6 earlier in the season. Now Lakers is 6-6.
There are many more chaos in life that do have their patterns, and statistical expectations. People don't need Artificial Intelligence to find them, average high school student level intelligence should be sufficient as long as one wants to find the patterns.