It's only fair to share...Share on FacebookShare on Google+Tweet about this on TwitterShare on LinkedInEmail this to someonePrint this page

Despite filing over 20,000 pages of proof to substantiate their claims of a rigged election, the most probable outcome is that the opposition call would be rejected by the supreme court. Not that their evidence would be the main contention but forces playing against their bid may outweigh justice here.

For most rational individuals, the supreme court decision would sorely lie upon the evidence submitted in court and the findings that would result from the evidence, but the world has proved irrationality to be a greater influencer than rationality.

Judges do not make decisions in a vacuum and hence they would surely consider the impact of their decision on the country’s political and economic stability.

Furthermore, a ruling in favor of the opposition means that the courts would be lined up with 2017 poll losers who will claim the obvious – that a nullified presidential election would ratify the nullification of the whole election.

Anyway that is beside the point. We are here to substantiate our own claims through the only thing we know never lies; mathematical proof.

First we will call upon your basic understanding of mathematics to state some truth about real numbers functionality and the 2017 elections data stream.

Taking that each vote is an independent random variable, then, there are certain characteristics that would hold true for the real numbers generated from their random independent occurrences.

The characteristics can be summed up as below in the case of the elections:

  • Every vote is independent of another
  • Streaming of election data would be a random event since time for voting completion is independent for all polling stations
  • IEBC tallied results on a real-time basis from polling stations hence the data streaming was a random event
  • The results stream published would depict characteristics of a random event

Since we are all on the same mathematical page, it would be prudent at this point to note that slightly after the beginning of the streaming of election results from every polling station, IEBC results consistently displayed a 8-12% variance between the two leading candidates until conclusion of the results at which point Kenyatta won with a 9.23% gap as captured in the below images.

In our below analysis, we will prove through mathematical analysis using the same IEBC data set that the above gap resulted from a pre-calculated formula.

Relying on the published IEBC presidential data, we would generate three replica scenarios using random number generators that would randomly populate data from all the 47 counties and analyze how the variance between the two leading contestants fluctuated as votes streamed in.

We would then replicate a scenario that would depict what transpired during the election and then conclude on our observation.

Note: the scenarios would be populated using county data rather than polling station data but since the former is a subset of the latter, the equation holds true.

Final 2017 Presidential Election Results for the two leading candidates Source: IEBC

Random Data Set I
Using  the first random generator, the presidential results from IEBC populated as follows:

Bar Graph representing cumulative county tally from the first random set of the 2017 presidential elections results data. Source: IEBC and Kenya Brief Research

Line Graph depicting the variance fluctuation between Raila and Kenyatta from random data set I of the 2017 presidential election results

Line Graph depicting the variance fluctuation between Raila and Kenyatta from random data set I of the 2017 presidential election results. Source: Kenya Brief Research

From the first randomized set of data, we can note as the data randomly streamed in, Raila was in the lead by 3.00%, then 8.00% before Kenyatta started catching up and turning the tide to his favor and completing the win with a 9.23% gap.

Random Data Set II

Using  the second random generator, the presidential results from IEBC populated as follows:

Bar Graph representing cumulative county tally from the second random set of the 2017 presidential elections results data. Source: IEBC and Kenya Brief Research

Line Graph depicting the variance fluctuation between Raila and Kenyatta from random data set II of the 2017 presidential election results

Line Graph depicting the variance fluctuation between Raila and Kenyatta from random data set II of the 2017 presidential election results

From the second randomized set of data, Kenyatta started with a lead of 5.50% which shot up to 8.00% before Raila started catching up and recovered to be 7.75% ahead. Kenyatta later on took the lead and completed the win with a 9.23% differential.


Random Data Set III

Using  the third random generator, the presidential results from IEBC populated as follows:

Bar Graph representing cumulative county tally from the third random set of the 2017 presidential elections results data. Source: IEBC and Kenya Brief Research

Line Graph depicting the variance fluctuation between Raila and Kenyatta from random data set III of the 2017 presidential election results

Line Graph depicting the variance fluctuation between Raila and Kenyatta from random data set III of the 2017 presidential election results

For random data set three, at the beginning of data stream, the race was tight with leadership fluctuating between the two aspirants before Kenyatta started to gain an unassailable lead and closed with a 9.23% victory.

We can clearly note that irrespective of the randomness of the data, the outcome remains similar and the variance fluctuates randomly between the two leading principals for all the three data sets.

Modified Data Set
Now, using the same IEBC data, we will modify a data set to replicate the published variance of 8-12%.

Bar Graph representing cumulative county tally from the modified set of the 2017 presidential elections results data. Source: IEBC and Kenya Brief Research

Line Graph depicting the variance fluctuation between Raila and Kenyatta from the modified data set of the 2017 presidential election results

Line Graph depicting the variance fluctuation between Raila and Kenyatta from the modified data set of the 2017 presidential election results

For the modified data set of which the analysis was done using a predetermined calculation of ascending order, Kenyatta assumed the lead and was never to be seen again by Raila. The scenario was only possible after the analyst manipulated the data to paint the desired picture. It is virtually impossible to obtain such a graph from a randomized set of data.

Furthermore, attaining a 8-12% consistent gap which is similar to the one displayed during IEBC data would require a pre-calculated formula which would be independent of the data.

Conclusion
From the above analysis, we can conclude that for the variance to be maintained as static as the one from IEBC data stream, the calculation has to be predetermined.

Whichever kind of data set you may choose to test the above theorem, it will hold. Hence this is sufficient proof that IEBC controlled the data stream on the IEBC portal.

Control of the data stream indicates that IEBC had a predetermined formula that released results according to the variance stipulated by them which was about 8-12%.

Mathematical proof has substantiated that a random collection of data between two independent variables cannot raise a linear variance unless the data is manipulated to do so.

This analysis is sufficient proof that IEBC controlled the data streaming but further proof is needed to confirm that the actual data was altered although the need for IEBC to control the data stream is sufficient motive for them to alter results.

The burden of proof now lies with the opposition but without actual data from the servers, it would be a hard case to win since the same servers claim to contain scanned froms 34A, 34B and 34C of which no one can validate their accuracy, only IEBC who are part of the accused.

 

 

It's only fair to share...Share on FacebookShare on Google+Tweet about this on TwitterShare on LinkedInEmail this to someonePrint this page

Posted by Timothy II Aperit

True believer in numbers. Statistics never lie. Bsc Financial Engineering MBA Finance ACCA

  • Knowitall

    You cannot use a whole county as a random set of data, the counties are too varied in voting patterns. Additionally, 47 events cannot statistically represent over 40,000 events accurately