11. April 2008 by emeris.

This blogging thing is harder than it appears. There are a lot of things in real life that do a good job of providing ample distraction from this activity. On the bright side my life has been going well

Posted in random | 1 Comment »

11. March 2008 by emeris.

The shuttle is up again

bbc

Posted in random | 1 Comment »

29. February 2008 by emeris.

People always say it is important to take the time to document your code, keep your work organized, and to write scripts that generate your figures all by them selves.

They are not joking.

Posted in physics, math, random | 1 Comment »

23. February 2008 by emeris.

Some ideas, such as “Speak like a noun” days, are in general silly, this one takes the cake. If 3.14 is designated as such, does this mean I’m not allowed to talk the rest of the year?

Posted in random | 3 Comments »

7. February 2008 by emeris.

In my last post I refered to information cascades with out really defining them.   They are a form of collective behavior where each member of a group asynchronously making the same decision.   Each member of the group had private information known only to them and public information from the observation of how other members of the group have decided.   Each individual makes their decision based on their private information and the available public information combined in some way.    Because of this dynamic the consensus decision of the group will not necessarily be the same as the decision that would be reached if all private information.

A toy example of this is consider a group of people making a binary decision, up or down, who all have private information that says to choose down except for some very small subset.  Let the aggregation method be majority rule of the private information and last two decision taken.  If we assume that two of the people with the minority information for first and both decide up, then the third person will see two ups (from the public information) and one down (from private) and also choose up.  The same for the fourth and later people as well, thus the group will all choose up, even though a majority of the private information says to choose down.   One such example is choosing forks at a fancy place setting.  It is clear that this is a toy model and there are much more sophisticated  systems, however this displays the important behavior.

Posted in math, random | 1 Comment »

5. February 2008 by emeris.

Thinking one is clever is one of the most dangerous things to think. While it is true that cleverness can be a great boon for such things at solving math problems, cleverness can also be a great curse because one does not always see the true consequences until it is much too late. Hence, I am exhausted.

One such example in recent history is the Gulliani campaign method. It was greatly clever, ignore the small states and save money for the big ones, unfortunately it failed utterly because this momentum business matters. A lot. But, if it had worked it would have been brilliant and changed the dynamic of how the primaries are run. This does bring up the question of how much the media really matters in elections and what their responsibilities are. Edwards wasn’t doing that badly in the early races, but received very little media attention. After losing his home state he dropped out, arguably a victim of information cascade as much as anything else.

This raises the question if it is good that information cascades play such a large role in our political system, if political systems can exist with out information cascades (functionally or theoretically), or if any attempts are being made to control such cascades.

On one hand the cascade is a rapid way to bring a lot of people to the same view point, however it doesn’t prevent them from all flipping a different direction en-mass (see the realignment with in the republican primaries and Obama’s uptick after Iowa). They also serve as a multiplier of efforts.  If you can talk some number of people to your side, then some other people who may not have received any additional information other than that more people support you, will also join your side based on the judgment of others. This is the implicit reason that endorsements are good for anything at all.

I don’t think a system can exist with out such cascades, given the highly connected nature of this country, nor would it be desirable.  It is valid to extrapolate information based on how other people behave, ie you see people flocking to or from a candidate then it is worth looking into it to see why.

The far larger concern is if there are groups/powers that be that are trying covertly to manipulate the cascades (it is clear the candidates are overtly trying to manipulate them) and if we could even tell if someone was trying/succeeding at it.

Posted in politics, random | 2 Comments »