2017 - Rough MOAS : Improving turn out in NASA stronghold won't help.

[RV Pundit]

Absolutely right. We now have social media in addition to mass media to also gauge the political wind. All you need is to find popular gusii board and you've got a real feel on the ground. I have been to a couple - and it would seem it still 3 to 7; 3 for Jubilee;7 for NASA.

Pundit, yes I remember nipate conversations. It was I who posted that tidbit on nipate.com about Gusii voting patterns re elderly vs youth and urban vs rural. I engaged with Energizer about Gusii votes, and the very moment Jubilee dumped cash super close to the elections and got its percentage up, I admitted so, on nipate. This was VERRY close to the election when the shift happened. Before then, CORD was well over 80%. And when the shift happened, it was tangible, you could actually talk to people in shags who were openly planning to vote for Jubilee after getting 50 bob. Lots of alcoholics, no kidding!  This time, I have no access to the pulse of the community and I'm relying on the assessment of others but apart from one Raila-loathing nephew of mine, I honestly have not spoken to a single Kisii who is either openly saying they will vote Jubilee or predicting votes for Jubilee. If I were home this past year, I'd have my own feel for things, both Gusii and Luhyas, but alas, I don't.

[RV Pundit]

13th June
Uhuru -53%
Raila - 46%
Others -1%

Note -I have basically assumed 2013 turnout excepted in Turkana & Coast - where I have increased turn out by 10% -coz I think MRC affected turn out in 2013.

[RV Pundit]

Will turn out help Raila; the answer is NO.
When I turn Raila's all stronghold and battleground to 95% vote turn out....
Uhuru still win by 51.3% against Raila 47.7%

Seem Raila win is tending towards impossible from improbable.

[RV Pundit]

Parliamentary seats - Projected number of seats (including women rep)
Jubilee (including frontier,mcc,pnu) -183(54%) -190 with nomination slots
NASA(ODM,Wiper,ANC, Ford-K, CCM) -126(37%)   -131 with nomination slots.
Independents & Small parties - 28(8%)   -

Note:This time small parties will not benefit from nomination fallout because of change of rule that disallow party hopping - losers (some rigged) ran as independents.

Total Elected MP seats-(290 +47 women rep)337 - with 12 nomination seats - to be split based on party performances - with Jubilee taking 7-8, ODM(4-5), Wiper/ANC (1).

Still rough - Note in 2013 - Jubilee had 56-58% of the parliament.

[Nefertiti]

The real Nipate election is whether MOAS are venerable or will go bust like the US opinion polls - the MOAS referendum.

[Kim Jong-Un's Pajama Pants]

The real Nipate election is whether MOAS are venerable or will go bust like the US opinion polls - the MOAS referendum.

US opinion polls were not a bust - most, if not all of them, poll the popular vote and extrapolate from that.  It is usually a reliable extrapolation, but it can be upset once in a while, because the electoral college does not always align with the popular vote. 

The best were https://fivethirtyeight.com/.  They had Clinton winning overall(both popular and electoral college).  They also predicted that there were decent chances of her winning the popular vote and still losing the electoral college.

[Kichwa]

At least the US opinion polls were shifting based on a daily, weekly or even monthly basis based on the campaigns, debates and issues.  MOAS is a done deal. Ouru could stand in the middle of Mombasa Road and shoot Sonko dead and still win.   

The real Nipate election is whether MOAS are venerable or will go bust like the US opinion polls - the MOAS referendum.

US opinion polls were not a bust - most, if not all of them, poll the popular vote and extrapolate from that.  It is usually a reliable extrapolation, but it can be upset once in a while, because the electoral college does not always align with the popular vote. 

The best were https://fivethirtyeight.com/.  They had Clinton winning overall(both popular and electoral college).  They also predicted that there were decent chances of her winning the popular vote and still losing the electoral college.

[RV Pundit]

How many referendum do you want to hold on MOAS when over the years it's proven itself.

The real Nipate election is whether MOAS are venerable or will go bust like the US opinion polls - the MOAS referendum.

[RV Pundit]

Kenya politics is far from issue-based - it mainly tribal based politics- the places where changes or shifting happens are so called battle-ground or swing grounds - but their weight is so small - any shift has minimal impact - the big 5 are more than 70% of the vote and that is easy to read.

Uhuru and Sonko have done worse....allegedly..and still won Senate & PORK. What more can top up Crimes Against Humanity - murder of one person?

At least the US opinion polls were shifting based on a daily, weekly or even monthly basis based on the campaigns, debates and issues.  MOAS is a done deal. Ouru could stand in the middle of Mombasa Road and shoot Sonko dead and still win.   

[Nefertiti]

GEMA and Kalenjin turnout might disappoint you... which answers Omollo's question why Uhuruto are up and about if their numbers are so solid: they are scared of poor base turnout.

GEMA, Kalenjin, Luo turnout
92, 97, 02 there was Moi
2007 Kibaki tribalism & 41 vs 1
2013 ICC
2017 - complacency?

For MP independent candidates will rain on the party. I assume Jubilee-leaning independents are counted as Jubilee and same for NASA.

How many referendum do you want to hold on MOAS when over the years it's proven itself.

The real Nipate election is whether MOAS are venerable or will go bust like the US opinion polls - the MOAS referendum.

[RV Pundit]

We can play with turn out and see if there is any change. I have played with it by giving NASA 95% turnout (impossible) and they still lose to Uhuru. Admittedly turn out is hard to model. Last time (2013 elections) I used the average turn out for all election that I have been held but we saw historic turn out of 86% - and that gave Raila lift of 2% - and  I got it wrong in Western where I though MaDVD & Wamalwa would beat Raila - another 1% boost - so Raila instead of ending up around 39-40% - was up to 42%. Uhuru dropped from 52-53% to 51%. But that was still within the margin of error.

I think now the issue will only be turn out. My strategy is to seed the average turn out (from all elections or referendum) with 2013 turn out - because history is the best predictor.

GEMA and Kalenjin turnout might disappoint you... which answers Omollo's question why Uhuruto are up and about if their numbers are so solid: they are scared of poor base turnout.

GEMA, Kalenjin, Luo turnout
92, 97, 02 there was Moi
2007 Kibaki tribalism & 41 vs 1
2013 ICC
2017 - complacency?

[RV Pundit]

I finally got some tribal composition census for counties....based on 2019 data from Kenya socio-economic altas
Link http://www.kenya-atlas.org/pdf/Socio-Economic_Atlas_of_Kenya_2nd_edition.pdf

I have updated MOAS to reflect this - I'll work on fine-tunning it later this weekend.

But apart from turn out - we are nearly getting the FINAL MOAS - I don't see any new issues that could dramatically change anything now.

[Kichwa]

Why don't you let Ouru know about MOAS so that he can stop wasting time and money campaigning.

I finally got some tribal composition census for counties....based on 2019 data from Kenya socio-economic altas
Link http://www.kenya-atlas.org/pdf/Socio-Economic_Atlas_of_Kenya_2nd_edition.pdf

I have updated MOAS to reflect this - I'll work on fine-tunning it later this weekend.

But apart from turn out - we are nearly getting the FINAL MOAS - I don't see any new issues that could dramatically change anything now.

[RV Pundit]

I pity anybody doubting MOAS.

[Nefertiti]

Pundit, what is the mystery behind the excellent Jubilee performance in Mandera viz the rest of NFD, where the ethnic composition is basically the same? I smelt a rat in 2013 but it seems your MOAS maintain 90% washout.

[Kadame5]

The real Nipate election is whether MOAS are venerable or will go bust like the US opinion polls - the MOAS referendum.

US opinion polls were not a bust - most, if not all of them, poll the popular vote and extrapolate from that.  It is usually a reliable extrapolation, but it can be upset once in a while, because the electoral college does not always align with the popular vote. 

The best were https://fivethirtyeight.com/.  They had Clinton winning overall(both popular and electoral college).  They also predicted that there were decent chances of her winning the popular vote and still losing the electoral college.

I remember all those months right up until Florida results began to look unrelentingly RED, the unquestioned truth was that Donald Trump had no path to victory and a 1% chance of winning. They also had state to state polls that showed Donald Trump lacked a path to victory as he couldn't win any combination of swing states to catch up to Hillary. And then he ended up with over 300 electoral votes. This was the universal wisdom for months on end on virtually every news outlet on TV. I am simply never trusting opinion polls ever again. Whatever voodoo math they use to predict outcomes, it is proving reliably unreliable over the past year: Brexit, Trump, and Theresa May. Too many variables to be caught.

[Ole]

Pundit is right about Narok County. Tunai will win not because he has anything good or he is liked by the people but because the purko have alienated the small clans besides fielding two candidates both in Nasa (ntutu and musuni). As far as the presidential election Nasa is still leading but jubilee is worker way harder to improve its numbers. In the neighboring bomet, jubilee is still ahead. Isaac  (who I like very much) will lose but not by a huge margin like the one pundit is predicting.

[Kadame5]

Pundit is right about Narok County. Tunai will win not because he has anything good or he is liked by the people but because the purko have alienated the small clans besides fielding two candidates both in Nasa (ntutu and musuni). As far as the presidential election Nasa is still leading but jubilee is worker way harder to improve its numbers. In the neighboring bomet, jubilee is still ahead. Isaac  (who I like very much) will lose but not by a huge margin like the one pundit is predicting.

Ole, if you don't mind my asking, who are you supporting for president?

[Kim Jong-Un's Pajama Pants]

The real Nipate election is whether MOAS are venerable or will go bust like the US opinion polls - the MOAS referendum.

US opinion polls were not a bust - most, if not all of them, poll the popular vote and extrapolate from that.  It is usually a reliable extrapolation, but it can be upset once in a while, because the electoral college does not always align with the popular vote. 

The best were https://fivethirtyeight.com/.  They had Clinton winning overall(both popular and electoral college).  They also predicted that there were decent chances of her winning the popular vote and still losing the electoral college.

I remember all those months right up until Florida results began to look unrelentingly RED, the unquestioned truth was that Donald Trump had no path to victory and a 1% chance of winning. They also had state to state polls that showed Donald Trump lacked a path to victory as he couldn't win any combination of swing states to catch up to Hillary. And then he ended up with over 300 electoral votes. This was the universal wisdom for months on end on virtually every news outlet on TV. I am simply never trusting opinion polls ever again. Whatever voodoo math they use to predict outcomes, it is proving reliably unreliable over the past year: Brexit, Trump, and Theresa May. Too many variables to be caught.

Trump's chance of winning tended to be between 10% and 25% if I recall well.  It was low but not 1%.  The battleground states were all within the margin of error.  He was at risk of losing Arizona for instance.  I think Hillary even campaigned there.  A mistake in retrospect, because even though the polls had her ahead in the firewall states they were within the margin of error.  It means they could flip either way and they did.  In Philly, she was simply unable to excite the Obama coalition.

4 days before the election, fivethirtyeight pointed out that a Trump victory, while highly unlikely, was still within a normal polling error https://fivethirtyeight.com/features/trump-is-just-a-normal-polling-error-behind-clinton/.  If something better than polling comes up, I would definitely not be averse to giving it a look. 

Even though people have correctly predicted election and even tournament winners just by shaking and throwing chicken bones where opinion polls got it wrong, I still think they are the best tools for the purpose.  If they are not working, one can go back and analyze what went wrong, and design tweaks to account for any shortcomings.

[Kichwa]

I accept the fact that in Kenya tribe is a very strong indicator of a person's political ideology, In the US, you can guess the political ideology of a person based on their race, education, gender and where they reside. Having said that, I believe that a politicians job is to change or maintain that ideology to their favor.  Many people predicted that Obama could not win because of his color but Obama managed to change that proved many people wrong.  This is the reason why I do not accept the notion that in Kenya presidential election can be predicted months before the elections just by plugging the candidates tribe and adjusting the previous voting numbers accordingly.  One can close their eyes and cross a busy highway twice without being hit by a care but that does not mean they are a genius as to matters of crossing roads with eyes closed. 

The real Nipate election is whether MOAS are venerable or will go bust like the US opinion polls - the MOAS referendum.

US opinion polls were not a bust - most, if not all of them, poll the popular vote and extrapolate from that.  It is usually a reliable extrapolation, but it can be upset once in a while, because the electoral college does not always align with the popular vote. 

The best were https://fivethirtyeight.com/.  They had Clinton winning overall(both popular and electoral college).  They also predicted that there were decent chances of her winning the popular vote and still losing the electoral college.

I remember all those months right up until Florida results began to look unrelentingly RED, the unquestioned truth was that Donald Trump had no path to victory and a 1% chance of winning. They also had state to state polls that showed Donald Trump lacked a path to victory as he couldn't win any combination of swing states to catch up to Hillary. And then he ended up with over 300 electoral votes. This was the universal wisdom for months on end on virtually every news outlet on TV. I am simply never trusting opinion polls ever again. Whatever voodoo math they use to predict outcomes, it is proving reliably unreliable over the past year: Brexit, Trump, and Theresa May. Too many variables to be caught.

Trump's chance of winning tended to be between 10% and 25% if I recall well.  It was low but not 1%.  The battleground states were all within the margin of error.  He was at risk of losing Arizona for instance.  I think Hillary even campaigned there.  A mistake in retrospect, because even though the polls had her ahead in the firewall states they were within the margin of error.  It means they could flip either way and they did.  In Philly, she was simply unable to excite the Obama coalition.

4 days before the election, fivethirtyeight pointed out that a Trump victory, while highly unlikely, was still within a normal polling error https://fivethirtyeight.com/features/trump-is-just-a-normal-polling-error-behind-clinton/.  If something better than polling comes up, I would definitely not be averse to giving it a look. 

Even though people have correctly predicted election and even tournament winners just by shaking and throwing chicken bones where opinion polls got it wrong, I still think they are the best tools for the purpose.  If they are not working, one can go back and analyze what went wrong, and design tweaks to account for any shortcomings.