1-800 GOT CHART JUNK

I was trying to get an estimate of how much it would cost to get rid of some stuff in a storage unit. 1-800 GOT JUNK trucks and billboards are all around and at their website you can get an estimate and see how much it would cost depending on how much of a truck you would fill. That data begged to be scraped and graphed to see what the set up cost and incremental cost of junk removal would be. They also provide pictures of how much of the truck would be filled as well as a picture of how much junk would fill it. I combined them all into the following chart, click it for a larger version.

Just a trip for a minimal amount of junk costs $142, and each increment of 1/8 of truck costs about $65. I added the little graphics of the amount of truck and junk next to each point so that a user could see it build, but I also wanted the full truck in the larger size. The 1-800 GOT JUNK website does do a nice job of letting you click to see each of these, but I like the graph to get the full view.

The payback time is based on a storage unit at $456/month. For a full truck costing $598 it only costs about 40 days of rent and then you have no rent after that. Might be worth it.

(source: 1-800 GOT JUNK, and my own imagination)

Another graphical anaysis of something or other

It is really quite clear from the chart above that my conclusions are flawless and that I am very clever.

(maybe to many of this, this, these or this recently)

(via Saturday Morning Breakfast Cereal)

The perfect Oreo is all filling

The perfect Oreo would be all filling and no cookies, but how would you hold it. The next best option would be a lot of filling with just cookies to hold it. Nabisco is moving in the right direction with the Double Stuf version. In some experimentation today I have been attempting to build increase the ratio of filling to cookie by cookie removal and filling half reattachment. If we count a regular Oreo as two cookies and one filling, filling ratio 1/2, then a Double Stuf is two cookies and two filling , ratio one. We can build from that. The following chart plots the experiment so far.


The results are below. On the left is four Chocolate Double Stuff Oreo cookies stuck together, on the right is a regular Chocolate Double Stuff Oreo for comparison.



Eating this cookie was unwieldy, I imagine ratios above this will result in a cookie that smooshes out all the filling when one attempts to bite it. I wasn't hungry anymore so I stopped the experiment to avoid filling and cookie waste. The experimentation will continue at a later date.

"Dynamic" baseball Payroll vs. Performance chart

Ben Fry, who can program, has created a dynamic salary vs. performance chart for this baseball season. His chart connects the current standings of each team to its 2007 salary.



We don't learn much more than I realized when I plotted the salaries vs. performance for last year - there is no correlation. You can also delight in the fact that the Yankees are paying a lot of money for a very mediocre team this year.

On his chart you learn these facts with more flash and javascript. I agree his chart dynamically changes the team performance as the teams win and lose games through the season, but I think if he has gone to all this trouble it would be great to see the payroll change as teams add and drop players over the season. Paul would point out that this is still a useless chart since we have beaten to death the idea that performance does not correlate to payroll in baseball (and in many other things). He suggests you read Moneyball.

(via Castro's Favorite Color)

Just how many prime ribs or prime numbers are there.

Just how many numbers are prime? Something on popurls.com inspired me to got to the Prime pages to figure it out. Experimentalist that I am, I just went and grabbed a list of the first thousand primes and then some points along the way to the first 15 million to see what I could see. I divided the number of primes by the prime number I had reached to get the fraction of numbers below that number that are prime. Plotting this yields a suspiciously smooth decreasing curve if you use a log x-axis, which is necessary because we have numbers from 2 to 275 million.

Since we don't have a straight line yet the next plot is of the inverse of the fraction of prime numbers. Bingo! We have a straight line. A quick check of wikipedia (Distribution of Prime Numbers. Know your search terms!) reveals that I have just empirically rediscovered the Prime Number Theorem. Typical engineer.

That theorem defines pi(n) as the number of numbers that are prime below n. The fraction of numbers that are prime that I was trying to figure out is pi(n)/n.

pi(n) ~ n/ln (n)

pi(n)/n ~ 1/ ln(n)

My plot is the inverse of the fraction on a log which gives me my straight line ln(n). The chance of a randomly selected number being prime is 1/ ln(n).

Team payrolls, last year's results and who is gonna win the pennant this year

We had discussion at work about baseball payrolls. I wondered just how high the Yankees payroll is and who the next place team was. A plot of the 2006 payrolls and the 2007 payrolls vs their rank shows that the Yankee salary is certainly disproportionate vs. the other teams, but they are not twice the next team anymore. The Florida Marlins and Tampa Bay Devil Rays hold the last two spots.


A high salary does not necessarily get you a good winning percentage. There is little correlation between salary and winning. Take that Yankees. Here is an example from 2006, both with and without the Yankees who skew the chart.

They say that past performance does not guarantee future results. That is also true for at least the last three seasons of baseball. I plotted the year's winning percentage vs. the year before and get an even worse correlation than for salaries.


Perhaps it is the quality of players that determines how many baseball games you will win. Team payroll or past performance doesn't seem to correlate.