« November 2005 | Main | January 2006 »

December 11, 2005

'Survivor' Game Theory

I wanted to get this shot in before tonight's finale of 'Survivor'. I love this show and have been watching it religiously since season 2. The reason I like it isn't for the drama or the "reality TV" aspect, it's because I think Mark Burnett and co. have created the world's most complicated strategy game.

Think about it. Chess is 100% strategy, but is sufficently less complicated because of the finite movements. Hold 'em poker is more complex because of the human reading/betting/bluffing angle, but still has the pretty important luck factor (at least over the short run). "Survivor" has very little luck -- only producer controlled things like tribe placement and challenge selection (and the mysterious, never used 'purple rock'). It also has the very sophisticated factor of human deception and alliances. That factor throws the game complexity off the map (and really into the realm of game theory as I understand it).

It's too bad the producers always cast the show with a bunch of dimwits. The case could easily be made that the smartest/best player in "Survivor" history was Richard Hatch in the first season. Nobody since has shown a very good ability to manipulate people or use what I would consider to be a complex, well-thought out strategy. Maybe this is editing. Maybe it's exhaustion. But I think it's stupidity. It would be really great to get a bunch of (in shape) poker/chess/backgammon professionals on the show to see what happens.

All of which is to say -- Thursday night, "Survivor" contestant Cindy made what may be the worst decision in the show's history. Of course, nobody has been given the same choice as her before. For those who aren't fans, (but are still reading -- good for you!) Cindy was given a $25K Pontiac SUV for winning a challenge, with the choice of either taking the car or giving the other 4 contestants the same car each. She chose to keep the car. Of course, there are other reasons why she shouldn't have, as this article points out. But let me just take the tack of why it was a bad idea from the possible cash return alone.

Let's compare this to a poker tournament for a second. If you have 5 people left at the table and the prizes remaining are $100, $10, $8, $7, and $6; assuming everyone has equal stacks of chips and equal skill, there is $131 prize money left in the game -- approximately $18 a person. Advancing from 5th place to 4th place in this situation gets you from $18 a person 'game equity' to close to $31 a person. Getting from 5th place and still alive to 4th place and still alive isn't just the $1 difference in the guaranteed payout, it's the equity in having the shot at the bigger prizes up the ladder. The assumptions above are big, of course, even for poker. "Survivor" has a bigger complexity in terms of human elements, and there is no way of knowing (especially with questionable editing) how secure players felt or who is really a better player.

BUT, for the sake of this discussion, let's assume that all the remaining players are equal. And (this is a big one) -- giving the other players the car guarantees you'll stay around for one more day, is that a good call from a money perspective? I think so. Let me take some shots at the math, and vague assumptions about probability:

1) Giving up the car means they won't vote you out that night.

2) If you don't give up the car, your chance of elimination goes up slightly that night -- from 20% to 40% let's say.

3) There is no effect on subsequent votes (for now, we'll leave it at just this vote).

The numbers above for the hypothetical poker tournament aren't pulled out of thin air. If you add 3 0's to each of them, they are the "Survivor" prizes. So given that, getting from 5th place and still alive to 4th and still alive increases your 'game equity' from $187857.14 to $313750. A more than $125K difference. If you go home it's a differnce from $60K to $313750. A $250K difference. Divide this by 5 for the aforementioned 20% more chance you go home, and it's still more than the $25K value of the car. The only way it's not a bad decision from this standpoint is if either you think there is no negative effect on the other players voting, or you don't think you have a chance of winnning the $1M prize (if your upside is 2nd place, there is just a $40K difference in prize possiblity).

This brings me to the second reason this was a terrible idea. THE JURY. Assume you think you do have a chance to win. If you make it to the final 2, 3 of the people voting whether or not you should win are people you gave a car to. You need 5 votes to make the leap from 2nd place -- $100K to 1st -- $1M. Are those votes each worth $220K? Does giving them a car sway 3 of them to some degree? Does it sway the other 6 on the jury at all? Can we quantify that at all? Probably not, but it doesn't seem like a trivial consideration to me.

So what really happened? Cindy got voted out that night. It is easy to write all this in hindsight, that she could have saved herself and bought herself some of the shaky concept of 'game equity'. Maybe my knowlege of game theory is off. But either way, wasting your chance at 1 million dollars for a $25K Pontiac hardly seems worth it.

December 09, 2005

Why Won't He Just Call???

Looking for a job is hard. I was really looking forward to the process when I started it a month or so ago. But it turns out it sucks. The constant barrage of the same questions, lots of unspoken rules (you shouldn't tell one company recruiter you are even looking at other places. What?!?), etc. I'm finding it hard to get thru a full day of interviews without making one cynical wisecrack. And if that wisecrack is taken wrong, well -- best of luck to you. NEXT!

But what is the most difficult for me is waiting for feedback. 1,2,3 days can seem like an eternity if you are excited about an opportunity. It's a lot like the waiting period dance after a first date or a phone number exchange. The "guy" (recruiter/hr person) is sitting there with some idea of whether there is interest. The "girl" (job candidate) can do nothing but wait at home for the phone call. And it is SOOOOO FRUSTRATING waiting. The constant temptation to call them (maybe they have some info, it'll be ok, I'm just checking in, etc.) I wonder if the recruiter is sitting there on the other side, thumbtacking the phone number on the calendar, "Swingers" style.

Newbie Recruiter -- "I should just call him now. We know we'd like to make him an offer."

Experienced Recruiter -- "No. No. Don't ever do that. At least a 3 day waiting period is mandatory. Make him sweat some."

NR -- "What? How can you do that? How long do you wait?"

ER -- "6 days. Just put it out of your mind."

So, as the 'female' in this exchange, that's what I'm trying to do too. Put it out of my mind. Read a book? Works for about 10 pages until the words on the page turn into (Insert company name here) over and over. Go snowshoeing? That's good, but the cell phone is still burning a hole in my pocket. Kill time on the internet? Oh no. That is an invitation to reload Gmail 45 times an hour.

Speaking of, I've got to go reload Gmail. And if you are reading this, Mr. Recruiter -- please call me (sobs into gallon of ice cream).

December 06, 2005

Urinals Are Broken

Urinals are broken. Even the big troughs they have at old football stadiums are better than the modern urinal. It almost makes me want to pee sitting down sometimes. You would think someone would have fixed the problems by now, but I guess not. So, maybe I can start the revolution. My beefs:

1) When you pee against the porcelain, the backsplash is dangerous. You don't get drenched, but there is a real mist issue and sometimes shoes or pants get wet.

2) For this reason, I often (when I am alone in a restroom) back away from the urinal after I get started. No more backsplash. But, those automatic flushers always flush the toilet when I back up. Why are those so sensitive, but the sink ones take a freaking full scale hand-jive to turn on? ANYWAY, this leads to this scenario:

- Start peeing, back up 6 inches or so.

- Flush!

- Almost done peeing, have to step forward again (hopefully flush is done so that doesn't splash me).

- Finish, step away -- no flush! Apparently the thing is smart enough to know I STEPPED BACKWARDS 6 INCHES, but not smart enough to know when I leave completely.

So... that's a problem. To say nothing of the chance that someone will walk in whtile I'm stepped back from the urinal. Then I get the real 'freak' label (as if you don't do the same thing!).

3) Why is there water in the bottom of some urinals? I thought water was in toilets for 2 purposes -- diluting the stains and the smells. No need for this in a urinal. Just drain the fluid and flush 5 seconds later. Peeing in the water doesn't help with the splashback. It just makes the whole experience loud and splashes more likely.

My proposed solutions:

1) Design a porcelain bowl for optimum spraying. There shouldn't be a rule that the pee has to hit the back wall at a 90 degree angle. Give me a decending slope to pee down. A 'pee-ramp' if you will.

2) Fix the auto flush laser. Come on, laser. You knew that you were hitting the other wall 6 feet away 30 seconds ago. Don't give up on me until you at least hit something 3 feet away or so (this allows for the case where another dude is back there waiting his turn). Trust me. That guy will be giving me 3 feet of room at least. Guys do not come close to each other in bathrooms.

3) Get rid of the standing water in the bottom. It's unnecessary and loud. I do like the 'urinal cakes' scented things though. They do freshen things up.

December 01, 2005

Overkill

A couple of weeks ago, I got inspired to clean out the garage so we could at least park one of our cars in there. One side of our garage is covered with a pegboard, so I picked up a peg kit. Below is the results:

pegboard.jpg

It turns out that we don't have enough tools to fill out the set, so I did things like hanging flashlights on the pegboard. Can you spot the other ridiculous items hanging on the pegboard?