Optimized Road Trips to Places You Would Actually Want to Go

Randy Olson made a fun little python program that uses google maps and some genetic algorithms to compute optimal road trips. The only problem is that his initial example road trip was constrained to visit a single destination in each of the lower 48 states, which means it skipped TONS OF GREAT CITIES.  Look at this thing, who would drive all over the country and skip LA, Seattle, and Chicago?

Optimal path for a single spot in each state. Click to interact.

So I did the only logical thing and forked Randy’s git repo and dropped in the top 50 metropolitan areas in the US. I figured I’d make optimal road trips for the top 3, 5, 10, 20, 30, 40, and 50 metro areas.  Click any map to embiggen and interactify.

3 5 10 20 30 40 50

I think the moral here is that you should never have to drive through Montana or the Dakotas.  Also, I’ve done two of the nasty blank stretches multiple times (Portland to the Bay Area and West TX).

This has also shown me I need to learn some Jekyll and move my blargh over to github.  Coming soon: yoachim.github.io

Anyway someone should go fork Randy’s git repo (or mine), and calculate an optimized bike path to visit every Starbucks in Seattle.

The Curse Returns

The Curse of the Simpsons has struck again with Leonard Nimoy (Seasons 4 and 8) dying at age 83.  Given that he was born in 1931 and appeared in season 4, I calculate there was a 57% chance of him dying before now.  So, like many of the other guest stars, he managed to outlive the standard life expectancy.


Dominant Tennis Players Over Time

As usual, I was starting my day with XKCD, when I read this gem:

This is great and all, but why isn’t there a panel for tennis players?  After all, they have a ranking system that makes it pretty easy to see who the dominant player is at any given time.  First, I was stoked to go plot it, then I was bummed because it wouldn’t look as cool as the XKCD plot.  But then I remembered that Jake, being a scholar and a gentleman, wrote code to make XKCD-style plots with matplotlib!  

Thus, I proudly present the men’s tennis dominant players over time, XKCD style:

Some notes about how I made the plot:

  • I only scraped data for the top 100 players (from here).
  • I’m only including players who held the #1 ranking for at least 52 weeks (need not be consecutive).  So no Patrick Rafter (who was number 1 for only one week).
  • Starting in 1990, I have listings for the actual point values in the rankings, so I gave the top ranked players a bonus based on their lead over the #2 ranked player.  The top players rank was computed as 2-points_1/points_2.  So if the top player has double the points of the #2 ranked player, I would plot them at zero.
  • I truncated the curves at the spot where players played their last match (they are still technically ranked after that, but the curves just decay away).
  • Player names start at the point where they first reached a #1 ranking

Some cool things you can see in the plot:

  • Agassi rose to #1 from being out of the top 10 THREE TIMES.
  • Lleyton Hewitt has had an amazingly long (but not particularly impressive) post-number 1 career.  But he did just win a tourney.
  • Nadal SHOT up to the #2 spot, and then was just parked behind Fed for a looong time.
  • Sampras and Edberg both managed to stop playing ranked #14.  The rest dropped out of the top 20.
  • Sampras was the only one to finish on a win (The US Open).
  • Federer wasn’t just #1 for a long time, he was #1 with a huge points lead for a long time.

Next up, I clearly need to plot the women.  I’d like to do both singles and doubles to show Martina dominating both simultaneously.  If I get motivated I’ll put the code on github in the hopes that someone else will help beautify it a bit more.

Rebooting Back to the Future

We’re coming up on the 30-year anniversary of the original Back to the Future.  This is of course extra meaningful since the movie involves time-traveling 30 years into the past, from 1985 to 1955.  A reboot of the movie practically writes itself:

 The Original  The ReBoot

The time machine is a Delorean, a ridiculous gull-wing sports car where the company that made it went out of business.


The time machine is a Hummer H3, a ridiculous sport utility car where the company that made it went out of business.


30 years later, the black kid becomes mayor of the city.


30 years later, the black kid becomes president of the country.


Joke about “Ronald Reagan the actor” becoming president.


Joke about “Arnold Schwarzenegger the actor” becoming governor.


Crazy Middle East terrorists


Yeah, that can stay the same.


Main character, Marty Mcfly


Main character, Martha Mcfly


Mom gets crush on own son (the whole incest plot line made it hard to get the original movie picked up)


Mom gets crush on own daughter (might actually make it an easier sell).


Climactic scene where rock classic Johnny B. Good blows all the kids away


Climactic scene where rock classic Smells Like Teen Spirit blows all the kids away


“It’s you’re cousin Marvin…”


“…Marvin Cobain!  You know that new sound you’re looking for, well listen to this!” And that’s why it’s so hard to understand Nirvana lyrics…


For some self-referential humor, there also has to be a scene walking past a movie theater in 1985, seeing the posters below:


“Wow, it’s like Hollywood hasn’t had an original idea for 30 years.”


See, it writes itself!  Roll credits and play this:

So who should direct, Joss Whedon or Seth MacFarlane?

Curse of the Simpsons Strikes Again!

Looks like Mickey Rooney died, marking yet another victim of the Curse of the Simpsons.  He was a guest star in one of my all-time favorite episodes, Radioactive Man the Movie (“The goggles, they do nothing!”).

I also see that over at the new fivethirtyeight site, Ben Morris did a very similar analysis to see if professional wrestlers die early.  Unlike being a guest on the Simpsons, which correlates with an additional 3 years of longevity, being a pro wrestler seems to be very bad for your health.

One good point on Ben’s analysis is that he used different actuarial tables for different years (to account for improving health technology), while I only used a recent one.  I don’t think this would change the Curse of the Simpsons much, and if it did it would be in the direction of making the anti-curse even stronger.



Fascinating plot on Nepotism over at Business Insider.

So it looks like wealthy people in Canada and Denmark are big on nepotism.  That’s not actually that surprising, what I’m curious about is the slight upturn on the left side of the plot.  Why do sons of poor fathers work for the same employer more often than average?

Born or Made? Football Kicker Edition

Following up on my post on birthday effects in pro sports that suggest that to some extent elite athletes get that way through training more than pure genetic gifts, there’s a cool article at the Atlantic looking at where the top football kickers come from.

The article claims that becoming an elite punter or placekicker is almost entirely an issue of training and getting high quality instruction, physical gifts are almost inconsequential.  Besides lots of training, one also needs to work the social aspect of the game and get noticed by attending high-profile camps.  Interesting stuff!