Blog Tools
Edit your Blog
Build a Blog
RSS Feed
View Profile
« September 2021 »
1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30
You are not logged in. Log in
Entries by Topic
All topics  «
On This Site
Tuesday, 24 June 2008
Boonsoft is born

Boonsoft is the closest thing to a software company that I have yet made. Right now, Boonsoft has 2 documents available for download. The core philosophy seems to be one of saving time. 

Posted by jonathan-m-katcher at 11:13 PM EDT
Updated: Tuesday, 24 June 2008 11:26 PM EDT
Post Comment | Permalink
Wednesday, 14 May 2008
Marx vs Tocqueville on revolutions

Posted by jonathan-m-katcher at 4:47 PM EDT
Updated: Wednesday, 4 June 2008 7:37 AM EDT
Post Comment | Permalink
Wednesday, 14 March 2007

What happens when you combine:


1) business simulation game

2)my expanding capabilities in microsoft excel and VBA

3)google's blogger (and in the future, other google tools)


I am learning more about how to use excel tools, and also discovering some universal truths about trading strategies and business. the nocompetition blog is more open to experimentation than the current blog here. I am going to post some of my excel tools and trading strategy ideas, so check it out at

Posted by jonathan-m-katcher at 7:32 PM EDT
Post Comment | View Comments (2) | Permalink
Sunday, 18 February 2007
business simulation game

After having found an interesting business simulation game, i have been developing some software tools in excel to make more informative and fun game play. Here is a picture of one of the features I have created:

This particular picture shows the in-game leasehold price that would be offered by the in-game government to a player were a plot of land sold by the government to a player. Land is represented as cells in a grid. Prices shown are in $thousands (in-game currency units). The price equation used is: 5k+1k*neighbors in a 3 square move area(no diagonals). Squares that are priced at 5k are the lowest possible price, and are colored blue.

This excel document of mine performs computations to represent data in another way from that provided in the game, in order to make game-play more fun and efficient. Now if I want to get new data, all I have to do is copy and paste the latest version of the in-game map into my excel document, and the new result will be instantly calculated.

Contact me for a copy (i intend to sell that excel software in-game, so i have not provided it publicly).

view my profile (to the left of your screen) for contact information.

Posted by jonathan-m-katcher at 2:28 AM EST
Updated: Sunday, 18 February 2007 3:17 AM EST
Post Comment | Permalink
Thursday, 14 December 2006
the SAD model of agent characteristics


According to my S.A.D. model of agent characteristics, every agent has 3 characteristics with respect to an event:

Sensor characteristic- ability or cost to get information from their environment about the probability of an outcome of an event.

Actuator characteristic- ability or cost to influence the probability of an outcome.

Demander characteristic- utility or payoff associated with each outcome.

Here is a copy of a paper of mine I brought with me and shared with some individuals at The Institute For Humane Studies' Workshop in Experimental Economics at Bryn Mawr College in July, 2006.

The S.A.D. Model of Agent Characteristics

The particular choice of mathematical model can be for example, a penny mechanism, a vote mechanism, an urn mechanism, a markov mechanism. Other mechanisms are possible, these are only examples of how the S.A.D. model could be applied mathematically.


What I like about the SAD model is that firstly, it is a positive not a normative economic model. It does not take sides. It can be used to run experinomic laboratory games that incorporate prediction markets. The agents are given private information about their own SAD characteristics. This substitutes for example in place of the double oral auction reservation prices given to each player, as famously developed by Nobel Prize winner Vernon Smith. But unlike commodity markets for pounds of apples or quantity of porkbellies, event markets like political markets in the IEM cannot be tackled experimentally using reservation prices. That is why the SAD model is useful. I will again repeat what I said at Bryn Mawr: To the best of my knowledge at this point in time, this is the first and only available model/framework that allows an experimental approach to the study of prediction markets. The SAD model allows the study of a controlled, repeatable artificial event with a probability that is known to the researcher.

What the SAD model does not do is define strategies for the agents. In order to perform a pure simulation using the SAD model, the simulated agents will be defined in terms of their SAD characteristics, but also must be endowed with behavioral strategies.

© 2006 Jonathan Moses Katcher, All Rights Reserved.

Posted by jonathan-m-katcher at 5:58 PM EST
Updated: Thursday, 14 December 2006 6:22 PM EST
Post Comment | Permalink
SimulatedUrn2.0b an Experimental Economics Engine

The following is a work in progress on an Experimental Economics Engine based on the S.A.D. model as applied to an Urn mechanism.

You will need to enable macros. Open excel. Select Tools>Macro>Security>Medium. Then Close Excel.

Click on the link, and save to your desktop. Close all other open applications. Open SimulatedUrn2.0 and select "Enable Macros".

SimulatedUrn2.0b        :an Experimental Economics Engine based on the S.A.D. model as applied to an Urn Mechanism

© 2006 Jonathan Moses Katcher, All Rights Reserved.

I will simply say here that many hours of work went into this before I even started coding in excel.


Posted by jonathan-m-katcher at 3:21 PM EST
Updated: Thursday, 14 December 2006 3:34 PM EST
Post Comment | Permalink
Saturday, 2 September 2006
3 more terms


When agents with actuator abilities enter a population of agents engaged in trading of prediction securities, these information markets can turn into Actuator Markets.


While prediction markets act as information aggregators they simultaneously act as Incentive Distributors.


Moral Hazard is a term used to describe this phenomenon, but from the point of view of an interested party: a frame of preference. Economics (when describing things as they are) is amoral, so cannot take sides on a moral debate. Thus the term moral hazard is fine for the insurance industry from which the term was borrowed, but not sufficient for an amoral discussion of the process. Instead we speak of Actuator Incentives.

Posted by jonathan-m-katcher at 8:54 AM EDT
Post Comment | Permalink
"actuator" is the correct word

Now there can be no doubt in my mind that actuator is the right word to use. On Friday, August 4, 2006, on one of my many internet searches for the word Actuator over the past three years, I came across this paper that used the word actuator in an economic context in 2002, thus predating my use of the word actuator in an economic context (2003-present);  This is good, because Gerkey and Mataric's paper reinforces the legitimacy of the term actuator in an economic context. They use the word in a slightly more well-defined/narrowly-defined manner, and specifically within the context of a network of robot 'nodes' they call a Sensor Actuator Network (SAN). They speak of simulation as well as real robots. I was suprised to learn of this acronym, because I had independently developed a similar acronym Speculator Actuator Demander model of agent characteristics (SAD) model of agent characteristics. But now that I have read their paper, It seems that perhaps the word "sensor" is better than "speculator" in getting across the idea of the ability to get information. The biggest differences between my version and their version of ideas are:

1)that they seem to limit themselves to using the word actuator in describing robots, and I want to use the term more broadly in any context with players or agents or people or whatever.

2)I want to think of demander characteristics as being a part of the players themselves, so they have a natural stake in the outcome, not just a market stake that can be traded away. This means their natural preferences can be neutralized by an equal and opposite hedging stake in the market.

They said SAN I said SAD 

They said Node I said Agent

They said task I said change in outcome probability

Here is the paper by Gerkey and Mataric that reinforces the use of the word "actuator" in an economic context:

A Market Based Formulation of Sensor-Actuator Network Coordination

Brian P. Gerkey and Maja J. Mataric


Posted by jonathan-m-katcher at 8:45 AM EDT
Post Comment | Permalink
Sunday, 14 May 2006
Just for the record, here is a sample of my work on the idea of the actuator, and related event-based-betting paradigm. I created these ideas independently, and to the best of my knowledge at this point in time, no-one else has.
Feel free to talk about THE ACTUATOR, but please mention my name. JMK Jonathan Moses Katcher.


Jonathan Katcher
5/13/06, 5/14/06
from ideas developed since 2003
Relevancy: Social Sciences, Economics, Prediction Markets
? 2006 Jonathan Moses Katcher, All Rights Reserved


People do things. They have control over what they do. And generally they know what they are going to do before they do it. This is common sense, and an assumption of this paper concerning the event-actuating-agent, or actuator for short.

The simplest example of an actuator is a coin flipper who uses the forced coin method of flipping, as opposed to the fair or biased coin method of flipping. The fair coin is fair (50/50), and the biased coin is biased in favor of one party (ie 30/70), who generally has superior knowledge of that bias. The fair and biased coins are very well researched. But forced coins have been relatively overlooked by science. Therefore though forced coins are not the principle topic of this paper but simply an illustrative example, I must nevertheless describe them in at least minimal detail.

Let us consider two people: a person who calls heads or tails as the coin is tossed into the air (caller), and a person who flips the coin (flipper). The flipper uses a forced-coin method. There are only two possible outcomes: ?heads? or ?tails?. It is obvious that when the caller calls out ?heads!?, the flipper will force the coin into the outcome ?tails?. And when the caller calls out ?tails!?, the flipper will produce the outcome ?heads?. So no matter which choice the caller yells out as the coin is tossed into the air, the flipper will always produce the opposite outcome. This is because the flipper is rational and informed of his own abilities. Therefore he will always choose to behave in such a way so as to win the little zero-sum game occurring between him and the caller. Where cash is involved and each player is betting the other will be wrong, the flipper will be the one winning the money. What distinguishes this type of coin method from fair or biased coins is that the probability of heads is not set in stone so to speak. Instead, the probability is dependent on the decisions of the caller and the flipper.

In the above example, the state-actuating-agent (actuator) was the flipper who could force the coin into either of the two possible states. But the idea of the actuator extends beyond the petty example of the forced coin. Let us consider two other examples: A profit-motivated terrorist, and a profit-motivated cancer researcher.

In a binary prediction market centered around the event of an attack of a particular building, one can bet either in the affirmative (bet attack), or in the negative (bet no attack). Assume these two outcomes are the only possible outcomes. Our profit-motivated terrorist has no inherent preference for either outcome; his only motivation is profit. If he chooses to actuate the ?attack? outcome, it will certainly occur with 100% probability. And furthermore let us assume his costs are zero (he stole the required materials, and will never be identified or caught for any of his crimes). Let us also assume that he has no ability to decrease the probability of the ?attack? outcome. Unless the prediction market prices already predict 100% probability of attack, it will be rational (given this agent?s characteristics) for him to bet on, and then actuate the ?attack? outcome. When the outcome occurs, his contracts(the ?attack? contracts) will be worth $1 apiece, and the ?no attack? contracts will be worth $0 apiece.

The idea of the actuator is equally applicable to more socially desirable behaviors, such as in the example of the profit-motivated cancer researcher. In a binary prediction market topic of a particular kind of cancer, there are two types of mutually exclusive and completely exhaustive outcomes: ?cure? or ?no cure?. A person can bet either way. Assume that our cancer researcher has no inherent desire for either outcome. He does not care either way. If he does choose to actuate ?cure?, then that outcome will occur with 100% certainty. Let us also assume that his costs for actuating ?cure? are zero. He has no ability to decrease the probability of ?cure?. Unless prediction market prices already predict 100% probability of a cure, it will be rational (given this agent?s characteristics) for him to bet on, and then actuate the ?cure? outcome. When the outcome occurs, the ?cure? contracts will be worth $1.00 apiece, and the ?no cure? contracts will be worth $0 apiece.

To simply state that the agent is an actuator is a meaningless statement. The correct way to state the role of the agent is: ?The agent is an actuator with respect to such and such event?, or ?the agent has actuator characteristics with respect to such and such event?. If not stated explicitly, the event over which the agent has sway must be stated tacitly by context. Furthermore, one cannot gauge from a statement ?the agent is an actuator with respect to such and such event? what exactly those characteristics are. In the case of the forced coin example, simply stating that an agent has actuator characteristics with respect to the outcome of the coin flip is not sufficient information to tell you in what direction the agent can increase the probability. It does not tell you if that change is irreversible. It does not tell with what certainty the actuator?s action will produce the intended result. And it provides no information regarding the possible existence of other actuators in the environment who also have the ability to influence the probability of that same event. Furthermore it should be emphasized that the characteristics conferring upon an agent the ability to manipulate the probability of an event (actuator characteristics) are independent of that agent?s other characteristics and features. An agent?s actuator abilities state nothing about that agent?s desires, or preferences for the different possible outcomes of the event. An agent?s actuator characteristics do not even necessarily presume rationality. Though theories of rationality may involve a discussion of the actuator characteristic, the actuator characteristic is completely independent of the concept of rationality. Regarding information, an actuator can have inside information about what his actions will be in the future. In fact, he is generally the producer of this information. However, even an agent with no memory and hence no possibility of storing insider information can still be an actuator as long as he has the ability to influence the probability of the event. So the idea of the actuator can be completely separated from the idea of the selfish pursuit of gain. So while actuator and selfish characteristics often are related, they need not be. When considering real-life examples, these questions must be answered by searching for a mechanism. When constructing an ideal model, these questions must be answered by assumptions within the model about the mechanism.

Despite simplifications in the examples above, the idea of the actuator is something that should be recognized, and that can and should be used for more complex analysis of these types of problems, problems that fall into the category of the event-based-betting paradigm; Problems such as those involved in prediction markets.

Posted by jonathan-m-katcher at 5:24 AM EDT
Updated: Sunday, 14 May 2006 5:35 AM EDT
Post Comment | Permalink
Saturday, 13 May 2006
VB in excel for random sampling with replacement

google search: excel random sample
excel Sampling with Replacement
requires VBA code shown and explained at this URL

Posted by jonathan-m-katcher at 9:35 PM EDT
Post Comment | Permalink

Newer | Latest | Older