Elezioni Iran

Ricevo da un amico iraniano questa lettera che contiene una prima analisi dei dati delle elezioni in Iran. Risulta evidente dai dati in loro possesso che i risultati delle elezioni sono un falso. Il metodo e' molto semplice: analizzano la dispersione dei risultati in 10 collegi e li confrontano con l'elezione precedente.
Si scopre allora che mentre nel 2005 i voti a favore di Ahmadinejad fluttuavano di un 10% ora solo di 1%! Confrontando poi i voti dei diversi candidati si osservano degli analoghi comportamenti "stranamente" lineari con fluttuazioni bassissime. Insomma sembra chiaro che questi risultati sono stati costruiti a tavolino senza preoccuparsi di proporre dei risultati realistici.

Vi lascio alla lettera originaria:

The Iran presidential election was held on 12 June 2009. Four candidates had this opportunity to compete in the election, Mahmud Ahmadinejad (the current president), Mohsen Rezaei, Mehdi Karrubi and Mir-Hossein Mousavi. There are a lot of doubts about the results of the election in the camps of three candidates who challenged Ahmadinejad. They believe that the results announced by the responsible organization, The Ministry of Internal Affairs (MIA), which is under the control of the government is just a fake manipulation. In this short letter we analyze the data announced by MIA. The data analysis is very simple and enlightening. It is highly expected that more detailed analysis will appear in the near future by many authors. Let’s start with the 9th presidential election (2005) results. In the first round, Ahmadinejad had in average %19.5 of total votes with a standard deviation %11. The least of his votes was from province Baloochestan with %5.5, and the most was from province Qom with %53. In the second round, he got %61 of total votes, with standard deviation %7, with the least %44, and the most %73, in the same provinces. So, in the 9th election the standard deviation was of order %10, with the difference of least and most about %40.
In 10th presidential election (2009) MIA announced in several turns the votes of four candidates, which are
coming in the table below:

If the numbers above are correct, the standard deviation of the fraction of Ahmadinejad’s votes to the total is of order %1. This means that, in 10th election, with respect to 9th one, all of Iran has become homogenized with a factor of 10. We should mention that the announced votes in each turn are due to the counts from different provinces and constituencies.
The other interesting observation, related to the mentioned above, is as follows. If one makes a plot comparing the votes for each of two candidates respectively (6 plots) one can draw a straight-line through the points with a surprising accuracy; the order of deviation from straight-line is 0.001. Again note that the announced votes in each turn are coming from different regions, saying that very different constituencies are similar with a great accuracy. This most strongly suggests that, first the lines are drawn, and then the points are selected.

The plots are presented below. Due to a inconsistency in Rezaei’s vote in turns 8 and 9 (the latter is less), just the 6 first turns have been used in the plots. On the plots, R is the regression, which measure how much the points are representing a straight-line (“1-R” is deviation from straight-line).