Secret Algorithm For Predicting The Winners

[RGJohns]

Whilst wondering who was most likely to win the elections this year, it came to me - an INDEPENDENT, (hopefully) accurate and (for now) secret algorithm for predicting the winners...

A lollipop for who can guess what the secret algorithm is

None of these represent my own views - blame the algorithm!

Sabb predictions...

President:

Andrew Keenan - 56%
M (Malcolm Collins) - 9%
Simon Francies - 2%
Robert Fett - 21%
The Dalek - 12% (I know!)

Director of Representation:

Tim Woods - 26%
Harry Giles - 11%
James Shield - 31%
Lucy Green - 31% (oooh!)

Director of Events and Services:

Stacy Lee - 74%
Alison Hunter - 26%

DoSDA:

Royce Hunt - 6%
Sarah Secombes - 23%
Philippa Dunn - 26%
Brad McIlwraith - 16%
Terry Fulton - 12%
Jonathan Cooper - 17%

These are only current (Tuesday) figures, the algorithm will update as the elections get nearer...

[Dickie]

I note your algorithm is lacking R.O.N..


[hr]

http://facebook.com/p.php?id=37106107&l=217e435e0a

[Dickie]

Double posting.

[RGJohns]

Quoting Dickie from 17:44, 11th Mar 2008
I note your algorithm is lacking R.O.N..


[hr]

http://facebook.com/p.php?id=37106107&l=217e435e0a

Indeed, it just measures candidates' current predicted position relative to each other.

[munchingfoo]

My guess is that the algorithm relates to the relative frequency of identical letters of each candidate in relation to those same letters in the post running for.

If that isn't the case, then its quite a big coincedence.

[hr]

“Argue with an idiot long enough and people will fail to see the difference”

[Dickie]

Quoting munchingfoo from 18:07, 11th Mar 2008
My guess is that the algorithm relates to the relative frequency of identical letters of each candidate in relation to those same letters in the post running for.

In that case R.O.N should get a VERY hight percentage of the votes for Directorships of Events+Services and Representaion.


[hr]

http://facebook.com/p.php?id=37106107&l=217e435e0a

[JohnQPublic]

I think anybody posting any kind of numbers publicly should be a matter rapidly taken up by elections officers.

This thread should be deleted.

This could be very easily misread by the a young voter not wishing to look further down the thread (or read the post) as an accurate reflection of student opinion. It should be removed.

[MJC]

I would be shocked if this was not based on facebook group membership.

[munchingfoo]

JohnQ

anybody who is stupid enough to do that shouldn't be eligible to vote!

[hr]

“Argue with an idiot long enough and people will fail to see the difference”

[RGJohns]

Quoting MJC from 23:13, 11th Mar 2008
I would be shocked if this was not based on facebook group membership.

Lollipop to MJC!

Now, would be interested to see if this correlates to actual results...

[Lid]

See, I thought that, but Jonathan Cooper seems to be ahead of myself, despite having less members, and ahead of Terry Fulton despite also having less members. Or is it as a percentage of facebook friends?

Royce Hunt - nominated for Director of Student Development and Activities.

[hr]

Mathematical Anti Telharsic Harfatum Septomin

[Jono]

Quoting Lid from 01:46, 12th Mar 2008
See, I thought that, but Jonathan Cooper seems to be ahead of myself, despite having less members, and ahead of Terry Fulton despite also having less members. Or is it as a percentage of facebook friends?

Royce Hunt - nominated for Director of Student Development and Activities.

[hr]

Mathematical Anti Telharsic Harfatum Septomin

I certainly hope facebook group numbers aren't an accurate reflection of the likely outcome. If that's the case; I'm screwed!

[hr]

Jonathan Davies for SSC Societies Officer

http://www.facebook.com/group.php?gid=2 ... 144&ref=mf

[groovy]

Quoting JohnQPublic from 22:34, 11th Mar 2008
I think anybody posting any kind of numbers publicly should be a matter rapidly taken up by elections officers.

This thread should be deleted.

This could be very easily misread by the a young voter not wishing to look further down the thread (or read the post) as an accurate reflection of student opinion. It should be removed.

I agree, the same way that opinion polls are barred in certain European countries as research has shown that numbers alone CAN AND WILL bias voters.

[hr]

Remember now, there's a big difference between kneeling down and BENDING OVER!

[Al]

For those predictions to come anywhere near being accurate, there would have to be an awful lot of electoral upsets.

Besides, given the nature of St Andrews I am sure that a lot of candidates have Facebook friends in common. How then do you work out who is going to vote for whom?

[Steveo]

I'm not deleting the thread, just so you know.

You can stop asking or expecting.

Stephen - Deputy Senior Elections Officer

[hr]

Get off my internet.

[RGJohns]

Just to confirm, it was based on number of members of each candidate's Facebook group, from Monday & Tuesday.

These may well have changed since then.

The one and only group that I couldn't find (on Mon) was Jonathan's - so I guessed that (sorry!).

[munchingfoo]

Here's a breakdown of who would win if the positions were awarded to those with the most letters in common with the position applied for:


Predicted Results in the Union elections 2008

President:

Andrew Keenan 27%
Malcolm Collins 7%
Simon Francies 22%
Robert Fett 22%
The Dalek 15%
RON 5%


Director of Representation:

Tim Woods 19%
Harry Giles 23%
James Shield 20%
Lucy Green 22%
RON 13%


Director of Events and Services:

Stacy Lee 38%
Alison Hunter 46%
RON 14%


Director of Student Development and Activities:

Royce Hunt 12%
Sarah Secombes 16%
Philippa Dunn 13%
Brad McIlwraith 16%
Terry Fulton 16%
Jonathan Cooper 19%
RON 4%


And here's the same algorithm applied with the same letter weighting that we use in a game of scrabble:


Predicted Results in the Union elections 2008

President:

Andrew Keenan 28%
Malcolm Collins 7%
Simon Francies 21%
Robert Fett 21%
The Dalek 16%
RON 4%


Director of Representation:

Tim Woods 19%
Harry Giles 22%
James Shield 21%
Lucy Green 23%
RON 12%


Director of Events and Services:

Stacy Lee 43%
Alison Hunter 43%
RON 13%


Director of Student Development and Activities:

Royce Hunt 12%
Sarah Secombes 15%
Philippa Dunn 16%
Brad McIlwraith 17%
Terry Fulton 16%
Jonathan Cooper 18%
RON 3%



The names were copied directly from the first post, if there are any spelling mistakes then I'll be happy to recompute the figures for you. If anyone wants the code to do this to mess around with leave me a message here.

P.S. All percentages are rounded down to the nearest whole number, so the percentage sums in each position will not equal 100%. The DoES position is a much more exciting race that way

[hr]

“Argue with an idiot long enough and people will fail to see the difference”

[Katherine_the_Noggin]

RGJohns your algorithm was spot on for the Sabb predictions! Andrew Keenan, Stacy Lee, Philippa Dunn and James Shield.

It seems the amount of facebook friends really does matter in today's world! Maybe you should do it for the U.S. elections...

[MJC]

Quoting RGJohns from 01:13, 12th Mar 2008

Quoting MJC from 23:13, 11th Mar 2008
I would be shocked if this was not based on facebook group membership.

Lollipop to MJC!

Now, would be interested to see if this correlates to actual results...

How will I receive the lollipop? Should I send you a private message?

[Steveo]

Quoting Katherine_the_Noggin from 10:53, 15th Mar 2008
RGJohns your algorithm was spot on for the Sabb predictions! Andrew Keenan, Stacy Lee, Philippa Dunn and James Shield.

It seems the amount of facebook friends really does matter in today's world! Maybe you should do it for the U.S. elections...

It might have predicted the winners, but the relative percentages were all off for DoR and DoSDA.

With the exception of DoES, the three other sabbatical results were the subject of some difficult redistributions of votes, and this didn't take any of that into account, nor RON, nor Spoils.



[hr]

Get off my internet.