Category Archives: publishable

The best of the best.

Vilfredo on IPCC

One document, multiple point of views

Writing paper is a classic many to one process. With many authors convening together in a single final result. Final result that should contain several claims, often causally related to each other, reaching a final result, usually in the Conclusions. While this structure is useful it is not always easy (or indeed possible) to merge all the authors point of view into a shared world view. The usual solution to this is either to tone down the statements (finding very general ones), or that some authors suffer the indignity of having to signa paper they don’t fully agree with, or having the group split and produce different papers, with different author lists.

One classical example are the IPCC reports. Those are presented every 6-7 years by the Intergovernmant Panels on Climate Change, and should summarise all the recent findings on climate change and global warming. The report is “the largest and most detailed summary of the climate change situation ever undertaken, produced by thousands of authors, editors, and reviewers from dozens of countries, citing over 6,000 peer-reviewed scientific studies” [wikipedia]. Being written by so many scientists from all over the world leeds to serious problems over what is going to be written in such report. The solution was to accept claims only if a certain percentage of the scientists would agree with it. But this leds to a toned down report that under represents the threats we are facing. With reality being always as bad as, and often harsher than the predictions.

A similar problems happen in the authorship of wikipedia articles. Where many authors present their point of view, and often edit wars erupt. Wikipedia policy to “stay neutral”, and add references to each claim are useful but nowhere enough for the kind of disagreements about the world that are presented among the wikipedia authors.

In this paper we want to suggest that it should be possible to present inside the same article competing views, by presenting them side by side. The main part of the paper should remain co-authored. But when we reach the disagreeing paragraph, the alternative world views should be presented side by side, with the initials of the name of the authors that subscribe to each paragraph in tiny print below it.

Note that not all alternative views should be present, just the Pareto Front of them, as explained in the “Don’t vote, Evolve” paper. In brief this is done by listing all alternative versions of the same paragraph, and letting all the participants sign all the versions they subscribe to. Then we define a version A of a paragraph being dominated by a version B of the same paragraph if and only if, the people who subscribe to A are a subset of the people that subscribe to B. The Pareto Front of the versions will then be the set of all views that are non dominated (there is no other version that dominates them).

1) It is important that the authors don’t subscribe just to one version but to all the versions they feel they can accept.

2) Every person that has subscribed to at least one version will find at least one of the versions he subscribe to being present in the Pareto Front (the proof is trivial and left as an exercise). In this sense the Pareto Front represents the view of the whole community with no tyranny of the majority.

How VIlfredo defines groups depending on what people subscribe to

3) Each author that did NOT subscribe to a proposal, cannot have subscribed to any proposal dominated by it (or it would not dominate it). In this way the presented Pareto Front does not just represents competing views but also a minimal division of the group of authors. With people that subscribe only to version A, people that subscribe only to version B, people that subscribe only to version C…, Then people that subscribe to A and B, A and C, …, … . If we have n competing versions in the Pareto Front it will be possible to have 2n-1 groups. Each that represents an alternative point of view.

From a visual point of view, in a static medium the alternative point of view can be shown side by side. In a dynamic medium, one version can be shown with the alternative ones shown as tiny columns on the side. When the reader moves the mouse over them (hovers) the columns would open up into fully readable text (while the previous one closes by the side).

Vilfredo on IPCC

The suggested function v(x)=(M-1)x/(x+(M-1)+k)+1, for k=0 and M=100. Has an Asymptote at 100, and near the origin it is tangent to the function y=x+1. It represents how should the weight of the vote change, as more people mandates a person to represent them.

Delegating in eDemocracy, my Way!

The suggested function v(x)=(M-1)x/(x+(M-1)+k)+1, for k=0 and M=100. Has an Asymptote at 100, and near the origin it is tangent to the function y=x+1. It represents how should the weight of the vote change, as more people mandates a person to represent them.

The suggested function v(x)=(M-1)x/(x+(M-1)+k)+1, for k=0 and M=100. Has an Asymptote at 100, and near the origin it is tangent to the function y=x+1. It represents how should the weight of the vote change, as more people mandates a person to represent them.

Recently I wrote a very long post in which I tried, as exhaustively as possible to discuss if it was the case to let people delegate their vote in eDemocracy. The conclusion was that it would be better not to introduce it. Which is a bitter conclusion, because it halts the conversation before it starts. I also suggested that IF we wanted to allow delegated voting, it should be done in a “non linear” way. In other words, it should be possible to delegate someone, but it’s not a good deal. The vote of a person who is present should weight more than the vote of a person absent, which is asking someone to vote for him. Also no one, independently to the number of votes, should have a vote higher than a certain value M. How much is M? I was suggesting that we should start with 1/630-ith of the number of participating people. So no one person should have more power than a parliamentary in the [Italian] House of Parliament today. Without explicitly listing them I was suggesting that the characteristics that we should use to decide the weight of a delegated vote were:

  • v(0)=1  (the weight of who does not receive any mandate should still be 1)
  • v(1)< 2 (the weight of who receive a single mandate should be lower than the weight of two people that vote directly).
  • \lim_{x\rightarrow\infty}(v(x))=M (If the number of mandates that a person receives grow, the weight must tend to M)
  • \dot{v}(x)\geq0 per x\geq0 (as the number of mandates grows, so does the weight of the vote of the person receiving them)
  • lim_{x\rightarrow\infty}(\dot{v}(x))=0 (Each extra mandate grows the weight a bit less each time)

I even went as far as to say that there were an infinity of functions that satisfied those requirements. And in fact…


GewurzFew days later my friend Daniele Gewurz wrote me… Daniele is a great mathematician, very precise in his work; His blog is called L’Accademia dei Pignuoli (con la u!) [Hard to translate, maybe: Academy of Nitpickers?]. So I contacted Daniele, and after I wrote the blog post he came back to me with the answer.

[pause… silence… suspense… ]

The function, in fact the functions that we are looking for exist. In particular, each function of the form:

v(x) = \frac{(M-1)x}{(x+(M-1)+k)}+1 per k\geq 0

will satisfy those requirements.

As k grows the weight of the delegated person will go down. For K=0 the single mandate will give an extra 0.99, that is v(1)=1.99 (if I receive the mandate from one single person, my vote will then weight  1.99, we lose 1% mandating your vote to someone that does not receive other mandates). If we assume that no person should have a weight higher than 100, and we take the fastest growing function (K=0) we obtain v(x) = \frac{99x}{(x+99)}+1. That you can see here on the side.

La funzione 1+99x/(x+99) passa per il punto (0,1) e (1,2), inoltre ha un asintoto pari a 100 per x che tenda all'infinito.

Notice how the function has an asymptote equal to a 100 (that is equal to M), and passes through the points (0,1) and (1,1.99). It is also monotonically growing (which means, it always grows), but, as it grows, it grows slower and slower. So it is better not to delegate to the same person that others ate also mandating, when possible. And instead distribute your mandate to other people. It is a function that fights against the creation of an elite!

Per x che cresce, il valore della funzione 1+99x/(x+99) tende a 100, senza mai raggiungerlo. Ha cioè un asintoto.

In the second image you can see how the function effectively approaches 100 (actually M, with  M=100, in the example) as x grows.

Per valori di x bassi la funzione y=1+99x/(x+99) approssima molto bene la funzione y=x+1. Chi riceve poche deleghe, le riceve quasi completamente.

And finally in the last image you can see how for low values the function approximates very well the function y=x+1. Which is the function that is (implicitly) used when people delegate their vote. In other words who mandates someone which no one else delegated,, will see his vote transmitted nearly fully. While who mandates a very popular person will only change slightly the weight of their vote.

We still need to decide if we should let people delegate to someone else the mandate they received. The famous proxy voting. Also in this case the same reasoning holds: it would be better not, as each passage between the person mandating, and the person actually casting it increases the imprecision (the person voting, ending up voting something different from what he should have voted to correctly represent the person asking him to vote for them), but if we really cannot avoid it, we can use another function. At the end the principle is the same, so the simplest thing is to just use the same function v. So if x people delegate another person (who’s vote will then have a weight of v(x), and this person delegates another person, this second person vote will now weight v(v(x)). It should be noted that if a person sums the mandates from two people (x_1 e x_2) the result will not be v(x_1)+v(x_2), but v(x_1+x_2).

Let’s see some examples:

5 people delegate Anthony. Anthony’s vote, at this point, will have a weight of v(5)=5.76. So about 5 and 3/4. Anthony votes with Carl, and together have a weight of v(5)+v(0)=6.76.

Later Anthony needs to leave, and mandates Carl. Carl’s vote will then have a weight of v(v(5))=6.44.

Notice that if all have delegated directly carl, his vote would have had a weight of v(6)=6.66.

At the end of the day, passing through Anthony costed to the group v(6)-v(v(5))=0.21 (Note, 6.66-6.44=0.22, but 0.21 is the result of better approximations). Not exactly a full vote, but a fifth of a vote. On the other side, society, all together, will pay the price of a lower precision on the result. Who knows if Anthony, delegating Carl, is really doing what those 5 people wants?

What would be a better strategy for this group? Obviously if they could all participate, they would have a combined weight of 7. But if, instead, they knew that only 2 people would participate, the best thing to do would have been to split the mandates between the participants, and obtain v(3)+v(2)=6.87; having lost only 0.13 points respect to the fact of having voted all together. (There is also the option to split your mandate among different people, but this we shall discuss in another article).

Considering all I think that permitting delegation using the function v_{M,k}(x) expressed with an M well chosen, and maybe also with a K=0 would permit to everyone to participate, without going toward those excess that we have seen in the various liquid democracy systems, where few participants will obtain enough votes to dominate the decision. And not let anyone else decide. Note that M is defined at a 1/630-ith of the total number of participants, and if the number of participants, is equal to 630, M will be equal to 1. And the result will be

v(x) = \frac{(M-1)x}{(x+(1-1)+k)}+1=\frac{0x}{(x+0+k)}+1=1

In other words, if M=1 it is not possible anymore to delegate anyone.

And if M<1. In this case the function decreases as the number of the people delegating a person. In other words, if you are part of such a small group of people, do not mandate your vote, but participate directly!

Should we allow delegated voting in e-Democracy?

This post is the english translation of an italian post.

My dear readers, we must speak. We need to speak of the concept of delegating votes in eDemocracy. It will be long, maybe boring, but necessary.

What do we want from an eDemocracy system? I know what I want. It took me a long time to find out, but now I know. I want, actually I would like, a system that permits to everybody to participate, and that magically extracts from people’s interactions the best proposal. But what does best mean? In other moments (link to an example from a talk I gave in Italian) I left this quite vague. Now I think I can be more precise. Let us supposed that all the people would vote, and let us suppose that all people voted having a deep understanding of the issue. Then the result would be the best proposal… by definition. What everybody wants, assuming everybody has a complete understanding of the problem. Or at least complete within the limites that we, as human beings, understand it.

This brings us to two problems. The first is that not everybody votes, and the second is that not all, in fact no one, has a full understanding of the problem. Even experts, in general, only see one side of it. But we know that the solution is out there, somewhere. And we know that any proposal we make, and any vote we take (with any system) will approximate this answer. Sometimes the approximation will be better, sometime will be worse.

Asking experts (a sort of technicians’ government) does not work. They do not represent the general population and have a too limited point of view. Even if they had a complete vision (if they spoke among them, for example), they still would not represent el pueblo.

Asking people to vote directly (also known as Direct Democracy) also does not work. First of all not everybody votes, and then people are not prepared on the topic. Even less than technicians, much less than technicians (which at least do understand the problem).

Let’s consider further this point, as it is important. Why asking people to vote does not lead to the best answer (as previously defined)? Who is to judge when has a persone voted well or not? But, of course, only himself can. Each person can judge his own vote, after he has done it, when he learns more on the topic. In other words a good vote does not lead to regrets, later. In other words it is an informed and well thought out vote. People that vote, statistically speaking, in the great majority of cases, do not vote in an informed way. For example, if you ask to a 100 people if they are in favour of against death penalty, mostly people will be in favour. Often death penalty is seen as a deterrent for criminality. If, on the other side, people are informed about the effects of death penalty, for example on the fact that it does not work as a deterrent (nation that introduced it did not see their crime level decrease), and the always present risk of killing an innocent, at this point the number of people that support death penalty diminishes. We can suppose that if those people voted a first time (without being informed), after they would regret their vote. That’s a bad vote! And it does not help, and in fact it hinders, in looking for the best answer.

So, simply asking people to vote does not work. By the way, another reason why it does not work is that often voters cannot understand the consequences of their vote. In other word we vote for the sicilian fishermen to keep their job, and to make the sea around Sicily a natural reserve. For all human beings to have equal rights, and that mentally deranged human beings that have committed violent crimes should not be freed after serving their sentence. And so on, and so forth. In this context it is important how do we organise the vote. What are the options, how it is worded, and so on …

At this point people usually get up, declare the problem unsolvable, and focus on something else. But not finding the solution does not mean the solution does not exist.

A great approximation to the answer we are looking for is given by James Fishkin‘s Deliberative Opinion Polling. How does Deliberative Opinion Polling work?

A random, representative sample is first polled on the targeted issues. After this baseline poll, members of the sample are invited to gather at a single place for a weekend in order to discuss the issues. Carefully balanced briefing materials are sent to the participants and are also made publicly available. The participants engage in dialogue with competing experts and political leaders based on questions they develop in small group discussions with trained moderators. Parts of the weekend events are broadcast on television, either live or in taped and edited form. After the deliberations, the sample is again asked the original questions. The resulting changes in opinion represent the conclusions the public would reach, if people had opportunity to become more informed and more engaged by the issues. [source]

You understand the result from this process is a much better approximation of all the ones presented until now (technicians government, direct democracy). At least the method tries to consider both elements (population, and information) integrating them in a statistically representative group.

We can consider under this light also Representative Democracy. The one we hold. We elect representatives, they are divided into working groups (in Italian: commissioni parlamentari). Each working group, for each law, interviews the experts, and proposes the laws to the parliament. It should be noted that the representatives are not thought to be statistically representative of the population. They should be the cream of the population. This also because Representative Democracy is a 19th century system. At the time only a minority of people knew how to read and write. Thus we now have two problems: the first is that by not being statistically representative of the population they tend to vote for laws that will favour people like them (same education, income, social class). The second is that since the same people are in power for a long time, they can be corrupted. It should also be noted that while the laws are developed in the working groups, they are then voted in parliament, that did not hear the experts. And here the member of the parliament vote, usually, how they are told by the party (so much for article 67 of the Italian Constitution). And this last passage eliminates completely the idea that the result should, somehow, mirror what an informed population would vote for. While making it a simple power struggle.

All this is a long preamble to speak about Delegation. Because, since not everybody votes (for many reasons: time, interests), considering that very few are preapared enough to vote, considering all this … does it have any sense to ask those people to elect someone that votes on their behalf? And how much should the delegated vote weight?

In other words, if I vote, and you delegate your vote to me, how much should my vote weight now? One? Two? Between one and two? More than two? Let’s take an example: John, Anthony and Sam are having a discussion. They decide to vote, and John looses against Anthony and Sam. No delegations, no problems. But now Sam needs to go, and leaves to Anthony his vote. Now John finds himself fighting against Anthony and Sam delegation. We can assume Anthony will win.

But now we add Joseph, and Joseph agrees with John. If they meet in four, John and Joseph, vote on one side (2 votes), Anthony and Sam vote on the other side (2 votes), the result is even.  Now let us suppose Sam cannot participate. So now we have John and Joseph on the one side, and Anthony and Sam’s delegation on the other. If before we had a problem with information, because John, Joseph, Anthony and Sam might have been misinformed on the issue, now the problem is even bigger. Sam’s absence raises the doubt that if he was present he might change idea. Or he might discover that Anthony does not represent him correctly. Everybody else is, instead, present. They could change their mind, but they do not do it. In other words, the situation is no longer symmetric. If Anthony, plus Sam’s delegation weight more than Giovanni’s vote; it is also reasonable to assume that, if we are looking for the best approximation to the best vote, mentioned above, that Anthony plus Sam’s delegation will weight less than John’s and Joseph vote.  We are thus supposing the best answer will, with more probability be on the side of two people that are present, than on the side of one persone present and one person delegated. So if we were to introduce delegation, it should weight more than zero, but less than one. You delegate, but in doing so your vote looses some of its power, for the legitimate suspect that you might have delegated the wrong person. This has also the effect of making delegation inconvenient.

If I am a mob leader, and it is possible to delegate, it is convenient for me that you do not participate, but delegate me. In this way I am sure that we all vote as one. But the option is more risky. The delegated becomes the weak link in the chain. You just need to corrupt that person to gain all the votes (at least for one time). But if, by delegating, we lose some power, the result is that it is no longer convenient for us, as a group, to have a single representative. In fact, the more we are, the better it is. But this brings us to another risk, the risk that some of us vote differently.  But, let us remember, our aim is not to win. Our aim is to approximate the best answer. So, if a person votes differently, because he is convinced that that is the best answer, he should do so. So also for this it is better if delegate voting is worth less than non-delegated voting; it induces participation.

But how much could someone’s vote weight? At the maximum? because if the vote of a person, thanks to all the delegations received, weights more than all the other votes connected, he has, essentially received absolute power. And he can do whatever he wants. So it is important, even here, that there is a maximum, M. And the lower is M, more the system will avoid centralisation of power. Even if two people together, could have an absolute majority we would have a centralisation of power. At the moment, in Italy, we have 2 Houses (Camera and Senato). Each law must go through both Houses. The Camera has 630 deputies. So there is a need for 316 people to have a majority. This is like asking to have an M such that:  316 * M = (Number of people voting/2)+1. At the last elections 50’449’979 people could vote. This brings us to an M=((50’449’979 / 2)+1) / 316) = 79’826. In other words, keeping the power division as it is right now, no one’s vote should weight more than 79’826 (Of course having two houses increases security, because each law must be approved twice). Of course assuming that all who can vote, will vote.  In reality an M of 80’000 is absolutely too high. At least because, as we said, not all will participate in the vote. While considering that no one should have more than 1/630th of the total weight of the people participating, is plausible.

NOTE: This part in blue does not need to be read to understand the article. 

In other words we could consider a function v(x) that represents the weight of the vote of a person that is delegated by x people (for x > 0). Even when no one delegates a person, his vote will still be worth 1, so v(0)=1. If the power was transfered completely we would have that v(x)=x+1. If it was not possible delegating v(x)=1 for any x. But we said that we wanted 1 < v(1) < 2 so that the vote of a person receiving a vote should weight more than the weight of the vote of a person that no one delegates, and less than the weight of the vote of two person voting, together (to be pedantic we would have to write v(0) < v(1) < v(0)+v(0), but v(0) = 1 ). Then we want the function to be monotonically increasing, so that the more a person is delegated, the higher his vote must weight (or at least it should not diminish!), and we want a maximum of M. So v(x) ≤ v(x+1) ≤ M. We could even require the function to keep on increasing, as x increases, without ever reaching M. So v(x) < v(x+1) < M. With lim (for x→ +∞) v(x)=M. The fact that the function should be strictly monotonically increasing would be expressed by requiring the first derivative to be higher than zero: v’(x) > 0. With lim (per x→ +∞) v’(x) = 0. In other words, summing up:

  • v(0) = 1
  • v(1) < 2
  • lim v(x) = M (for x→ +∞)
  • v’(x) > 0       for x > 0
  • lim v’(x) = 0 (for x→ +∞)

Is there a function like this? Sure, there are infinite many! In a follow up post I will discuss a family of them (in Italian is already available here).

From all this we can desume that if we want to let people delegate their votes, a function like the one described above would be optimal. Who receives n votes will have a weight of v(n), with n described above.

But at this point we need to discuss on the real sense of those mandates. There are several reasons why a person delegates another. Maybe he is too busy, has already decided what to vote and delegates someone to give that specific vote. But this is not the most common mandate. Especially on the Internet, where voting takes little time, but deciding and informing one self takes more time. More common are mandates because a person does not feel proficient on a topic. Let’s go back to our definition of best answer. If I do not understand a topic, and I find myself facing a proposal that I do not understand I have few options. I can vote randomly, maybe intuitively, I can get informed, I can delegates someone, or I can decline the vote. Of those options, getting informed, or asking for help are surely the best (but they require more time). Voting randomly is terrible. Not only will I probably not guess right, but I risk voting for a different answer than what I would have voted if I were informed. So, basically, damaging my real position.  From the studies of Fishkin (see above) we know that people that vote intuitively, don’t have a great intuition. So even this option is not ok. And mandating someone else?

Delegating someone gives a sense of security. I don’t understand the topic, but this friend of mine understands it more than me. We have many values in common, I let him decide for me. This is ok in a conspirationist view of the world, that divides the world into good and bad. Bad people are in power, the good (we, of course) resist. And surely you will even find some cases where this is true. But in the great majority of cases the real difficulty is not choosing between good and bad, but which principle should prevail. And in this scenario (much more common!), mandating becomes harder, because the choice is less clear.

In this post I am presenting a series of ideas on how delegation works or does not in e-Democratic systems. Who does not agree with me, is not evil. Simply he does not agree. And if we have to vote, if we should introduce it, how can you mandate someone to vote for you? Surely you will not have spoken with your friend about those ideas. By the way what possibilities do I have of convincing a person that does not even make the effort of reading the article? Delegating your vote, like asking friends what should I vote? transforms a process of mass discernment, where all together we try to find what is the best course of action, into a simple power struggle. If you need to ask your friend, don’t ask what should I vote? ask what do the proposals say? which makes a world of difference.

And if I am really new to the topic? I can’t, in few hours on the internet, have a degree in Economy, Medicine and Physics. But you do not need a degree to understand most concepts. The most prepared people, instead of receiving a mandate to do what they feel is right, they should help others understanding. I am not speaking about an Utopia, but of a crowdsourced system to rewrite proposals.

And, what if the system, beside expressing how much do they agree or disagree on each proposal could also vote how much did they understood the proposal? And could, maybe, ask for clarifications? What if the system would put in touch people that ask a question with people that understood it? Maybe asking people who did understand it, clarifications. Or maybe it could invite who understood the proposal to rewrite it, in a clearer way. All doable things. The result would be a system where proposals are not only understood but also explained to the less prepared people. It wouldn’t be a perfect Democracy (i.e. the one that always let the best proposal emerge), but it would be a good approximation. I am going to speak about this model in a more detailed way in a future post. (The post in Italian is here)

And we still need to speak about proxy voting. If I can delegate you, can you then delegate another person? And what weight should my mandate has? The idea that I can delegate anyone, and he can delegate anyone, passing on my vote unchanged, leads to a system called Liquid Democracy. Since everybody can participate, it is democratic. Since the mandate can be taken back, and re assigned to someone else, it is liquid. Reality is that this system tend to create an elite of superusers, which receive the votes from everybody, centralising power. At this point fans of liquid democracy (or of liquid feedback, the only -or at least the most famous- system that implements it) will answer you, that you can always change your vote. Truly, try it, they will always tell you. And you must answer: “who cares?!”. Because, if you are going to lose time to follow who votes for what, and check that that is in fact what you would have voted, you can as well vote directly. And, finally, for a person that does check, and corrects, his vote there are many others that will not do it. Making those superusers, into a true elite into which you are not invited. In which your ideas might be discussed as an act of kindness and generosity. But not as a due act, among peers.

The next point is: “how long should a mandate last for?”. I would say that we do not even need to lose time on this. The longer it lasts, the lower correlation there will be between what the delegated person votes, and what the delegating person would have voted. We are soon going to move into the realm of random voting, except the people that receive the mandate, will concentrate power, and vote for a well defined direction. As I described before, on how proxy voting tend to create elites of superusers, is especially true for long mandates. In fact, the longer the mandates are, the bigger will the difference be between the people who have the power, and people who do not have it. I understand that some people claim that this is natural, this is how things always have been. Effectively there is not much difference between a person that is delegated through an online series of mandates (from people that forgot about it), and someone who is elected because his name is in the top list of one of the major parties. A party people by now vote for force of habit. This is the system we set up to change, and substituting it with another gerarchical structure is not the solution. Not even if some of us end up sitting in some of the high position of that structure.

But then, would a system where people delegate, and then ri-delegate each time work better? It would have other defects. For example, if people need to delegate each time, they can, as well, vote directly. Also, because of what we said before, delegations are always imprecise to understand what people want.

To all this we need to add that many people do not even try to delegate who, according to them, can represent them well. Instead they delegate some well known figure. An actor, a comedian, someone from the show business. Let’s remember how Bud Spencer tried to be elected for PDL (Berlusconi’s party), and Giorgio Gaber came back to vote, for his wife who was also in Berlusconi’s party lists. If those two examples haven’t convinced you that you cannot trust anyone with your mandate, just because he was a good actor or singer, I really don’t know what could.

And then there are those that delegate for convenience, to exchange favours, for promises made, because it is my wife, my husband, my son, my father, the cousin of my electrician, I can’t refuse his my mandate!

And the worse is that those people, mandate, and then they go away. Leaving to future generations the responsibility to handle a person whose vote now weights more.

Summing up, according to me if we want to build a system that let what people want emerge if everybody has studied the question, we should not use permit delegated voting. If we must use it, use it sparingly, use it for short time, more focused than more general. Better not to permit proxy voting than permitting it.

And let instead try to build systems in which everybody can participate; who has the time does not dominate the discussion (we shall speak about this in another post); who does not understand, can ask for explanations (and in this gets educated, which is never bad), and such that it does not let elite, club, groups, circles form that decide for other people.

I will add that in the systems I am developing (VilfredoBubblingAround, recently I am offering some free consulting for Airesis) I tend to follow (or suggest) those rules:

  • no one is allowed to delegate anyone else;
  • proposals are presented in an anonymous format, for those that vote them. So people vote the proposal, and not the author;
  • … [keep following me, and you will know the others, as I am writing]

Ebooks, the next revolution. But this time is BIG!

Another revolution is about to happen.

A revolution that is many times in size and importance bigger than the music revolution. I call it the e-book revolution.

In this moment a number of technologies are coming together:

On the one side OCR technologies are reaching a level of sophistication, where it is nearly as easy to photocopy a book as to make an ebook out of it. Do you remember when you would go to a photocopy shop, and ask them to make a copy of the book. Now it is that easy to have the ebook version of it, if you know how to do it. This means that more and more books are available in ebook format.

But the difference between the ebooks now, and the mp3, back then, is that when the mp3s came out, a song (5 minutes of fun) was about 5 MB. And since the internet was slower back then, it would take quite some time to download those 5 minutes of fun. Now a book, is often between 1 and 10 Mega Byte. And it can permit you to read it, study it, but also just to consult it.
More about this later.

I thought there were few ebooks around. That mostly you could find some old classics, but nothing really interesting. I was SO wrong.

Here is a collection available for download from pirate bay with more than 1000 ebooks, all on computer science. Here another with practically all of the ebooks from the “* for dummies” collection.

Those are not just some old classics. Those are good new books.

But why are users going through the whole work of digitalizing a single book to post it online? I guess this text will explain us: Continue reading

My Sunday Treasure Hunt: Enzymes and Digestion

If you go around the net, looking for information about enzymes and digestion, you often find detailed explanations. Explenations that generally say that enzymes come in a limited quantity, that if you eat raw food you are taking in new enzymes, thus not depleting your resources but if you eat cooked food you deactivate the enzymes, and eventually (over the years) you will have troubles digesting food. If you, instead, go to wikipedia nothing of this is present. Nor if you ask to a ‘normal’ doctor will they tell you any of this. What follows Continue reading

Tag Clouds are hard to Spam

I think the time have come to write my third, and hopefully last contribution to the topic of tagclouds.

I have been hearing a lot of talk on how users should not use too many tags in linking to url. I also am the maintainer of the mindmap maker, and I often look at some of the maps generated (available to everybody). There is a number of people who tend to use an average of between one and two tags per URL. Their maps are often very ordered. No clustering, no hierarchy. (Forgive me if I don’t put a link to such a map, but since I am going to bash this way of using delicious, I’d rather bash a method than a specific human being. Just go to the list of maps and open a couple, odds are one of them will be of the type I am describing). This way of using delicious uses tags as folders, just with the modification that every now and then you can put an URL in more than one folder at the same time. A bit like big bookstore might carry several copies of the same book, and store them in more than one place (and the Tao Te Ching, ends up in New Age -God knows why- and in Religion).

Of course tags tend not to fit exactly. My Tag Clouds and Cultural Change will be under Tags or Folksonomy or Sociology… Whatever you chose you probably will not put it under Ajax. And yet most of the analysis was done studying the spreading of the term Ajax.

Let’s make a few simple calculations. Continue reading

Tagclouds and cultural changes

In the previous post I discussed how we can measure the relative importance of tags in a post, by calculating their weight, as

  • weight of tag t= (number of people using t)/(total number of people)

I also said that:

Not only we could study a culture by studying the differences in the power law approximated by the tag clouds used by people of that culture. But we could even measure cultural eartquake by measuring the difference between the tag cloud being generated before a certain event, or after a certain event.

Independently Clay Shirky was coming at a similar conclusion, although he more focused on temporal changes that seem more signature of a particular subgroup of people all bookmarking a site at a certain time:

During a period of about 120 users’ additions of OIO, 20 or of them used the tag ‘ia’, putting it between #7 and #10 during that period. Now it is down to #17. This suggests that one or a few IA-oriented sites or mailing lists posted the link, and it got a flurry of attention from those taggers in a narrower window of time. This in turn suggests a conversationally tightly-knit IA community.

Through this tool we can see changes in the culture we are living in. We are used to feel those changes, but generally we never were able to measure them. Maybe now we might start to be able to do it.

But let’s go back to the tag weight. Terrell Russell took the ball, and in one evening of programming presented a tool to actually see how the weights change in time.

Nothing to say about the tool. It works perfectly well, and although it can be enhanced in many little ways, it already is very useful. Not bad for one evening.

More interesting, from my point of view, is how, through this tool we can see changes in the culture we are living in. We are used to feel those changes, but generally we never were able to measure them. Maybe now we might start to be able to do it.

No change

Tag Clouds rapidly converging

First of all I would like to show you the graph of a part of the culture where no changes are happening:
From the site: Nifty Corners. 1859 people having bookmarked it by now. The values soon converge to what we can expect to be their definite value (for the culture we are in).

Little Social-Quake

Continue reading

On Tag Clouds, Metric, Tag Sets and Power Laws

Note: This entry is connected also to a mindmap. Some people were having problems in opening the page because of that. As such the mindmap has been stored in a separate page, and can be viewed from here.


As correctly pointed out by Jeffrey Zeldman tag clouds are becoming more and more popular. Yet I keep seeing services which should be using tag clouds that keep on using tag sets. It is not just a problem of programming a tool which can only support tag sets, but also but also of programming tools which might in principles produce tag clouds, but such that the users are not invited to use a tag if one already exists, and as such don’t generate a tag cloud.

Example of the first type of tools are Flickr, 43things, consuMating, tagsurf * , example of the second is the tagged version of the BBC* . In all those cases a tag set is used, where instead a tag cloud would be more appropriate. Some of the differences between a tag cloud and a tag set where explained in Explaining and Showing Broad and Narrow Folksonomies. Let’s see them again, and see some consequences of those differences, which should clarify when is better to use one tool and when is better to use the other. Continue reading

A house divided

As the price of houses rises, more and more people find that the best solution is to divide a house among friends. Usually each person gets a room. The problem then is: who gets what room and how much should he pay. Usually the total rent is fixed, and usually the rooms are not exactly all the same. Some might be bigger, some smaller. Some might have a better view, more privacy, closeness to the toilet, more silence, and so on. And what’s also important is that different people might value the various elements in different ways.

I present here two ways of splitting the rent and dividing a house. I personally favour (and has designed) the second, but while I was presenting this method to some friends to get some
feedback, I was told the other, it seemed simpler, yet interesting enough to add it. They both assume that:
a) the rent is fixed,
b) there are no favoritism among the will-be-housemate on
who gets to choose first.

The ‘find the objective value first’ method.

Before the rooms are assigned, get together and agree on what are the objective value of each room (i.e. 20% of the rent for this, 50% of the rent for this). The total value must of course be the whole rent. Then randomly select who gets what room (at the agreed price), and as a final action people are allowed to exchange rooms if they want to.
Positive element: it is simple and quite straightforward.
Negative element: it assumes that people can easily agree on the actual relative value of the rooms, and that such value does not change respectively to the persons.

The ‘each person gets the best room’ method.

As I said this is the method that I love most. First of all let each person inspect all the room. Then each person, writes, secretly, the relative value of each room in a piece of paper. The sum of the values must be equal to the requested rent. The idea is to divide the house so that each person gets a room, and pays for that room the value THEY wrote on the piece of paper, while the sum of the valued paid by each person totally covers the requested rent.

Obviously, very often, the collected money would then be higher than the rent. Let’s call the collected money minus the monthly rent, the ‘extra money’.

Often there is more than one solution, that permit to have a some extra money each month. When this happens, the solution that permits to maximize the extra money is chosen. The extra money is then used to pay for the light, any extra expenses, or whatever is needed for the house.

Sometimes there are more than one optimal solution, that is some solutions generate the same extra money, everybody is paying the requested cost for each room, and all other solutions are less optimal. In that case the adopted solution will be one of the optimal one, randomly chosen.

Examples, examples:
Let’s suppose we have a house with 3 rooms (a, b, and c) and 3 persons (A, B, and C). Let’s suppose the total rent being 100.

Person A might find the three rooms equivalent, so he might just write (a: 33.3, b: 33.3, c: 33.3). Person B might instead favour room B, because is more sunny, and she likes to paint, and then she thinks that room ‘a’ is slightly better than room ‘c’, infact she would prefer not to be in room c at all, so she would write: (a: 35, b: 40, c: 25). Person C instead does not care about the sun, but has noticed that room A has more privacy, plus is near the toilet, and since he likes to have his gf as a guest, thinks that having room A would be a better deal. So he votes (a: 40, b: 30, c: 30).

Then the papers are revealed.

Generally when a room has a person that values it more than all the others, and he values that room more than all other ooms, then that room gets taken by that person at the price he has choose.

In our example we have:
A: (a: 33.3, b: 33.3, c: 33.3)
B: (a: 35, b: 40, c: 25)
C: (a: 40, b: 30, c: 30)
which would give us that A would get room ‘c’ paying one third of the rent. B would get room ‘b’ paying 40% of the rent, and C would get room ‘a’ for 40% of the rent… and the collected money each month would be 33.3+40+40=113.3 . The extra money would be 113.3-100=13.3 and would be used to pay for the electricity, water, gas, or whatever.

It is also possible to rinormalise the prices, by lowering them so that the total sum becomes exactly the cost of the rent, while the relative ratio remains the same. In our example
A: (33.3/113.3)*100=29.4
B: (40/113.3)*100=35.3
C: (40/113.3)*100=35.3
and person A would pay 29.4 of the rent (since he took the room nobody wanted)
person B would pay 35.3 of the rent (and took the sunny room)
person C would pays 35.3 of the rent (and took the room with more privacy)

So, what if the situation is not that easy. There isn’t a person that prefers each room? For example you could be in a situation like:
A: (a: 45, b: 45, c: 10)
B: (a: 40, b: 40, c: 20)
C: (a: 40, b: 30, c: 30)
well in this case it is obvious that person A will get either room a or room b. But it is also obvious that room c will go to person C. So C get’s c at 30% of the rent. Both A and B value the room a and b equivalently. But once the room will be assigned person A will pay more than person B, so it seem fair to me that person A chooses a or b and pays 45, and person B gets the remaining room, but pays less (40).

But things can get even more complicated if some people
value some rooms exactly the same:
A: (a: 45, b: 45, c: 10)
B: (a: 45, b: 45, c: 10)
C: (a: 40, b: 40, c: 20)
in which case A and B have obviously to randomly choose who gets what.

Or if the situation is symmethric among the rooms:
A: (a: 40, b: 30, c: 40)
B: (a: 40, b: 40, c: 30)
C: (a: 30, b: 40, c: 40)
In which case you randomly choose if A gets a or c, and then the other follow obviously.

So here we have the first mehtod, where everybody chooses the value together, this is equivalent on the second method if everybody agrees on the relative value:
A: (a: 35, b: 40, c: 25)
B: (a: 35, b: 40, c: 25)
C: (a: 35, b: 40, c: 25)
After which, also in this method, you would randomly pick who gets which room.

Please, let me know if you have tried it and if it was succesful.