Saturday, September 01, 2007

Chocolate sack

I took this picture yesterday at my local bakery:


I hear that once you eat chocolate sack, you never go back.

Wednesday, January 17, 2007

Effective bumper stickers

I think that most bumper stickers have little or no effect on people. If you agree with one, then you smile. If you disagree with one, then you just think the person is a jerk.

I propose a new class of bumper stickers designed to elicit much more powerful responses from people.

If you're like me, you don't really like conflict. But sometimes, people just need to have their asses kicked. I hate to have to do it myself, so I propose using a new type of bumper sticker to get the job done by other means.

Here's now it works: when someone needs to have his or her ass kicked, figure out which car is theirs. If they are a coworker or neighbor, this isn't too hard. Then, secure yourself one of the bumper stickers as described below. Then, when nobody's watching, stick it on the back bumper of their car. These stickers will be small so that they won't necessarily notice them, but sooner or later, drivers behind them will see it.

The idea is to make the stickers say incredibly insensitive things to hopefully elicit anger from a driver behind them. Sooner or later, a driver will get out of his car, pull your target out of his or her car, and beat the shit out of this person.

Now, what would these bumper stickers say? Well how about the following:

(please keep in mind I am not promoting these statements; they are design to make others angry)

"Even if the Holocaust really did happen, I'm still happy it did!"

or

"I hate ni$&%*s" (obviously you'd use the real word on the sticker)

or

"Hey stupid Muslim: why don't you go blow up a cave in Afghanistan?"

etc, etc.

Sooner or later, anyone with such a bumper sticker will experience a world-class ass-kicking.

Saturday, January 13, 2007

Let's kill credit card applications by mail

If you're like me, you are annoyed by all the credit card applications you get bombarded with at home. I propose we all try something new to try to discourage the credit card companies from soliciting our business this way. Maybe we'll save a few trees along the way too.

Whenever I get a credit card application, I always send it back in. I don't fill out any of the information, and I certainly don't sign the application. Instead, I take a thick black marker and block out my name and address, as well as writing the word "NO" in big letters. Please see the image below.



Then, I take the application and, using their provided envelope (which does not require postage added by me) and I send it in.

There are two reasons why my doing this hurts the credit card company (just a little bit):

1. Based on the deal they have with the US Postal Service, they only pay postage on the return envelope when someone actually uses it. So they can print as many as they like, but when that barcode at the bottom of the envelope gets scanned, the USPS sends the credit card company a bill. So, the more of the envelopes that get returned, the more costly it is for the credit card company to solicit business this way.

2. They need to pay people to open these envelopes. The more envelopes they get, the more people they need to employ. Every envelope they get that does not lead to a newly-issued credit card is a loss for them. The higher the fraction of the "non-applications" they receive, the lower their profit margin becomes.

Now, me doing this by myself represents nothing more than the tiniest nuisance to them. But if we all get together and do it, that might make a difference.

For example, I return about 5 of these a month. If 50,000 of us do the same, that's 250,000 non-applications a month, or 3 million per year. That's about a million dollars in postage, plus whatever it costs to get them opened. Now, a million dollars is not a lot to these companies, but we know that every 50,000 of us who do this will cost them an additional $1 million.

Let's give it a try!

Wednesday, January 10, 2007

Tough interview question

In my career, I've gotten to do a fair amount of interviewing for both engineering and biological science (e.g. biochemistry) positions. Based on some things I picked up in grad school, I liked, late in the interview, to ask a pretty tough question to all candidates:
"Imagine that I had a drinking cup full of water. Now imagine that next to that cup, I had a second identical cup filled with a liquid that looked absolutely identical, but actually was completely inviscid. Design a test to discern which liquid was which."
The point of this question is to get at the candidate's problem-solving abilities, as well as their ability to design experiments. Given the unusual nature of this question (specifically the hypothetical existence of a theoretically impossible physical property), only the rarest of candidates would have an immediate answer. I liked to give the candidate time to think, and to see how he or she would approach the question.

The most popular thing coming first out of any given candidate's mouth was, "What does 'inviscid' mean?" In my mind, an engineering candidate would lose a bit (and only a bit) of merit if this question was asked, since anyone who has had basic fluid mechanics should know that an inviscid fluid is one with zero viscosity. I wouldn't expect biochemistry majors to necessarily know this. In practice, only about 25% of candidates would know what 'inviscid' meant (although, to date, every single candidate who knew the definition double-checked with me that I was, in fact, referring to a zero-viscosity liquid.)

Following are a series of responses I'd get after that typical first one, along with my impressions of the candidate.
"I don't understand the point of the question, because zero viscosity is impossible and therefore the question has no practical relevance."
This particular response bugged me. On the one hand, it is a valid point. For some candidates, this response may represent a strong rooting in problem-solving by wanting to address only practical and relevant issues. However, I feel that a job interview is the wrong time to call into question the relevance of a technical issue. This question (or really a comment) would be better saved until the end of the interview, when the candidate could ask something like, "Why did you choose to ask a question that was not technically possible?" I have, in fact, been asked this at the end of interviews, when I would explain that the goal was to see how the candidate problem-solved out of the box with a topic I could be almost entirely sure they had never previously considered.
"Well, you could just drink both cups."
I got this response a lot. In fact, it was usually a "gut response", from about a third of the candidates. It was almost always said with a little laugh, a sure sign that the speaker is not confident in what he or she is saying. (Notice how often President Bush laughs when he speaks.) My response to this statement was always to ask what the hypothesis was behind this test. At this point, the candidate will do one of two things: GOOD = to say "well, actually I'm not sure how you could tell from that...let me think some more." BAD = "well, I'm sure you could tell somehow." If they made the GOOD response, I'd chalk off their initial response to answering too quickly, and I'd wait for a real response after they thought some more. If they made the BAD response, I'd begin badgering them, asking WHAT measurable thing would happen differently if they drank an inviscid liquid. Rarely would I get a good response. By this point, you already know that this candidate does not understand the basics of experimental design and they do not know how to develop and test hypotheses. You already know that if you hire this person, he or she will require explicit instructions on everything to be done, and otherwise will do everything by basic, inefficient brute force.

(Just as an aside: what would actually happen if you drank an inviscid liquid? I'm not entirely sure, except that you might have trouble not spilling it due to the absence of surface tension as it exits the cup and touches your lips. I think it's also possible that the liquid would slide down your throat without swallowing, again due to the absence of surface tension.)

When I asked this interview question, what I really wanted to see was the candidate come up with a basic test that was measurable in some way. If it took them some time to get there, I didn't mind.

So, one response went like this:
"I would rotate both cups."
So I ask how they would behave differently and what could be observed to determine which liquid was which. The candidate said something I really liked which was, "I haven't figured that out yet, although I am pretty sure this will work. Just give me a minute." So we waited, for about a full minute. Then she said, "OK--if you rotate the cup containing water, the water inside will begin to swirl in the same direction as the cup, but I think the inviscid liquid would not swirl...is that right?"

So the candidate wasn't totally sure, but was absolutely correct. I really liked this response for several reasons: 1) The candidate picked an area where she was pretty sure the viscosity was going to be relevant, meaning she had at least a basic skill for narrowing down a broad problem using basic principles 2) The candidate was not afraid to sit in silence and take the time to think about the problem, plus not to mention the fact that she thought well under pressure--it's not like that's the most comfortable situation and 3) She wasn't afraid to represent a partial theory she wasn't entirely sure of, and to properly represent her level of conviction. This is just what you want to see someone bring to a team: present ideas, state their case as best as they can, and be open about their level of confidence. This candidate demonstrated to me at least a good starter level in problem solving skills, as well as at least a good starter level in communication skills and honesty. (She was hired.)

There are a whole bunch of other answers that are similar to the above one, in that they revolve around very basic tests. Another example:
"Pour a few drops of liquid onto the table. The water will bead up, while the inviscid liquid will spread out completely since it has no surface tension."
What an awesome answer. This guy gave the experiment, the expected result, and the scientific reasoning all in one complete answer. He gave me confidence in his ability to design experiments, as well as his problem-solving and communication skills, all in one-fell swoop.
"Place a drop of food coloring dye into each cup. The drop spreads by diffusion, which is limited by the viscosity of the liquid. Therefore, the drop in the inviscid liquid should spread to fill the whole cup much faster than the drop in the water."
Ditto on this answer as compared to the one above. The woman who gave this answer ended up with a job offer.

One candidate was being interviewed for a position that involved a lot of engineering on automated pipettors (for picking up and dispensing liquids in automated fashion.) He answered:
"Well, once you hire me, I can tell which liquid is which because I won't be able to work with the inviscid liquid. The pipettor relies on surface tension to pick up liquid, so I won't be able to get the pipettor to work with on the inviscid liquid."
Ha! This guy was brazen, but he really impressed me. A great answer, just like those above. But he did two other things I liked: 1) He related the question directly to the job description he was interviewing for, which shows excellent problem-solving skills by applying the knowledge at-hand to the problem and 2) He asserted himself as a candidate by suggesting that he was already getting the job. That's cocky, but it also shows strong self-confidence and ambition. Those qualities are difficult not to like, and indeed this guy was hired.

Keep in mind that this was just one interview question among many, and I was only one interviewer among many. The way the candidate handled this question was just one data point towards the full evaluation. However, I'd estimate that about 90% of the time, this question revealed the candidate's true level of problem-solving abilities

Saturday, December 23, 2006

San Diego Padres Trade Tree

We know that Kevin Towers has been a very active general manager over the years, making quite a few more trades than average for a major league baseball team. Here's a little diagram I put together, a so-called "trade tree". It is not exhaustive, but rather tracks as far back as possible, via trades only, from three current members of the Padres (Chris Young, Terrmel Sledge, and Adrian Hernandez) who were acquired in early 2006.

It's quite amazing that starting from just that one trade, the players involved go all the way back to 1976 - that's thirty years of continuity through just trading!

Finally, a few interesting things of note:
  • All the trades involving this particular tree are present. Some of the players (such as Doug Brocail and Phil Plantier) were acquired again by the Padres, but not as part of this specific trade tree.
  • One player, Craig Lefferts, actually appears twice in this trade tree! After the Pads traded him away in 1987, they re-signed him as a free agent in 1989 (and proceeded to trade him away again in 1991!)
  • Both MLB players named Greg Harris appear in this trade tree. The ambidextrous one was acquired for Al Newman in 1984, while the other one was drafted in 1985 and then traded in 1993 as part of the deal for Andy Ashby.
  • There are a heck of a lot of good players on this list, including some MVPs (Ken Caminiti, & Kevin Mitchell, e.g.), some Cy Young winners (Rollie Fingers & Mark Davis, e.g.) and some Hall of Famers (Fingers, and Ozzie Smith).
  • There are also some guys who went on to manage, including Jerry Manuel and Bob Geren, who was traded shortly after being drafted in 1979 and will manage Oakland this coming year.
Without further delay, here is the San Diego Padres Trade Tree:

Monday, December 11, 2006

Where's Washington?

I heard a story recently that reminded me something about how the Pacific Northwest differs from the rest of the US.

A friend of mine was moving from suburban New Jersey out to rural Washington state, out in the unincorporated part of Bremerton on the Olympic Peninsula. He bought a new cell phone in NJ a few weeks before moving, but was sure to ask the salesperson if the phone would get the same quality signal in Washington. The salesperson assured him that he would, in fact, get the same quality signal.

After the move, low and behold, he gets no signal at all out in Washington state. He takes the phone to his local branch for that cell provider, and they get the same salesperson on the phone from the first store, who corroborates my friend's story about assuring him it would work "in Washington."

Very quickly, they discover the source of the miscommunication. When people say "Washington" in the eastern part of the country (and most other parts as well), almost everyone assumes they are referring to Washington D.C. (i.e. the District of Columbia) whereas out in the Pacific Northwest, "Washington" refers to the state of that name. As someone who lived in Washington state for years, I can tell you that when the folks out there refer to the country's capital, they call it "D.C."

Friday, December 01, 2006

Really Awesome Halo

Here's a camera-phone shot of a full rainbow halo taken off the right side of an airplane.

Righteous!

Cinder Block Art at Tufts


This qualifies as a really awesome thing.

Behind one of the buildings on the Tufts campus (and visible from Boston Ave in Medford MA), somebody arranged a pile of cinder blocks to spell out the school's name.