Chirchir Forgot to add an END to the 54% Algorithm

[Omollo]

• Tweet

Both the screens at Bomas and the Portal are still adding Uhuru votes:

[Kim Jong-Un's Pajama Pants]

Doling out part of the remaining 0.31% of the polling station numbers that weren't  sent by typed text.  That means they didn't have the scanned Form 34As for these stations.  Or they did, but the typed text had for some reason not been received. 

Just trying to explain what this means, assuming that it's happening in good faith.

[Omollo]

There is no provision for "Provisional results". The Court of Appeal put an end to that mischief and reminded the IEBC they have 7 days to do the job before issuing a certificate.

Note also that Chibukati announced to the whole world that they had received ALL the forms and verified the same.

Doling out part of the remaining 0.31% of the polling station numbers that weren't  sent by typed text.  That means they didn't have the scanned Form 34As for these stations.  Or they did, but the typed text had for some reason not been received. 

Just trying to explain what this means, assuming that it's happening in good faith.

[Kim Jong-Un's Pajama Pants]

There is no provision for "Provisional results". The Court of Appeal put an end to that mischief and reminded the IEBC they have 7 days to do the job before issuing a certificate.

Note also that Chibukati announced to the whole world that they had received ALL the forms and verified the same.

Doling out part of the remaining 0.31% of the polling station numbers that weren't  sent by typed text.  That means they didn't have the scanned Form 34As for these stations.  Or they did, but the typed text had for some reason not been received. 

Just trying to explain what this means, assuming that it's happening in good faith.

I am on your side on this one.  I am just saying the more I try to cut the system some slack with genuine explanations, the more puzzling this gets to me.  It shouldn't be that way.

[Omollo]

These guys saw clearly that they would be walloped. They made a decision to steal and embarked on the long charade of campaigns when in the end it was the computer hackers led by Chirchir who delivered.

Someone suggested that I know nothing about IT. Before I changed to what I do now, I changed from two professions and one of them was IT.

I am on your side on this one.  I am just saying the more I try to cut the system some slack with genuine explanations, the more puzzling this gets to me.  It shouldn't be that way.

[RV Pundit]

Machines once they get network should sync

[Omollo]

Machines once they get network should sync

Explain further.

They do not have a network? What machines to be specific

[Kim Jong-Un's Pajama Pants]

Machines once they get network should sync

Explain further.

They do not have a network? What machines to be specific

There are scenarios where some packets can be received late.  Like in a message queue.  It would make sense to use a message queue for a system like that to

1) ease the load on the main server and
2) facilitate loose coupling and reduce the chances of losing messages that have been sent.

The message queue server receives each message(presumably the text)...Even if the main server is down or unreachable from the message queue server for some reason, these messages will not be rejected or lost.  As soon as the main server wakes up, the message queue server will connect and deliver the messages.  That way things keep moving even if the main server is not ready to receive results.

But you have to believe that this information has been in limbo for four days waiting for some servers to connect.  I have not gotten any indication from IEBC that they have this kind of problem.  Also I don't recall KPMG report mentioning any message queue server architecture in place in KIEMS.

This is mostly pure speculation on my part, - playing devil's advocate - IEBC are best placed to explain these things.  And their explanations could lead to more questions and so on.  Yes, this remains weird in a normal setup.

[Omollo]

Then explain the constant 54 vs 44

[Omollo]

See who we have joined....

[Kim Jong-Un's Pajama Pants]

Then explain the constant 54 vs 44

Near constant.  There were some slight fluctuations.  Raila led for a brief first few minutes.  And then the final pattern established itself like a trend.  If you look at very small time intervals, you should see fluctuations.  But if the intervals are big, the fluctuations smooth out.

[Omollo]

Do you have a plausible explanation (other than the ones you have already provided) why:

1. There are near perfect constants
2. The figures are still adding up while maintaining the % difference between the two leading candidates
3. Why the same is happening to ALL the 54% vs 44% cases?

[Kim Jong-Un's Pajama Pants]

Do you have a plausible explanation (other than the ones you have already provided) why:

1. There are near perfect constants
2. The figures are still adding up while maintaining the % difference between the two leading candidates
3. Why the same is happening to ALL the 54% vs 44% cases?

Number 2 and 3 could be explained by the relative size of the unadded figures.  Those numbers coming in are so small, they would not make a dent on the percentages.

I don't have any other explanation for number 1.  You do realize I think the results are suspect I hope.  But if I can explain something away, I am willing to allow for that possibility in order to be fair to IEBC.

[Omollo]

I disagree.

The percentages cannot change regardless of the number of votes unless one gets a figure out of the programmed spectrum. The 54% vs 44% are the constants to be maintained no matter how many new votes come in.

Note that the issue is not the percentages. Those are constant to the decimal. However the totals for each candidate keep changing.

I think if one examined the algorithm the increment was conditioned on time intervals as well as the number of votes with the percentages locked.

Do you have a plausible explanation (other than the ones you have already provided) why:

1. There are near perfect constants
2. The figures are still adding up while maintaining the % difference between the two leading candidates
3. Why the same is happening to ALL the 54% vs 44% cases?

Number 2 and 3 could be explained by the relative size of the unadded figures.  Those numbers coming in are so small, they would not make a dent on the percentages.

I don't have any other explanation for number 1.  You do realize I think the results are suspect I hope.  But if I can explain something away, I am willing to allow for that possibility in order to be fair to IEBC.

[patel]

what kind of synching is this that adds votes to Uhuru and deduct votes from Raila ?  ???????

[Omollo]

More votes, same percentage difference....