Prediction Competition - March 2018

[Pete Whitehead]

Mar 23, 2018 17:40:17 GMT Davıd Boothroyd said:

I challenge all the 5 points. The Conservatives won in Thurrock, unequivocally, by one vote.

Have to agree. The Conservative candidate was elected - he won. Nobody who predicted that outcome should be getting any wrong winner faults

[middleenglander]

For week 4:

Authority	Aylesbury    Vale	Bassetlaw	Cheshire      East	Chiltern	Midlothian	Staffordshire Moorlands	Thurrock	Week 4
Ward	Central & Walton	Worksop   South East	Bunbury	Ridgeway	Penicuik	Leek West	Ockendon	faults
ajs	12.8	31.7	43.0	11.9	6.8	50.1+10	18.8+10	195.1
andrewp	11.5	14.6	4.3	6.4	22.5+10	40.1+10	11.6+5	136.0
David Boothroyd	27.3+10	16.6	5.7	26.5+10	10.8+10	46.1+10	9.6+5	187.7
dibs	44.8+10	34.6	45.0	37.9	19.7+10	44.1+10	22.4+5	283.5
greenrobinhood	17.6	49.7	39.0	47.6	8.5+10	56.1+10	32.4+5	275.9
hempie	10.4	27.7	17.0	19.9	12.5+10	44.1+10	16.8+10	178.4
k9	100	100	100	100	100	100	100	700
Lancastrian	11.9	34.6	25.0	10.4+10	18.8+10	38.1+10	20.8+10	199.6
Olympian95	14.4	40.6	2.3	43.6	11.2	34.1+10	2.4+5	163.6
priceofdawn	64.8+10	13.7	48.7	30.4+10	19.8+3	60.1+10	6.4+5	281.9
ricmk	19.5	14.6	25.3	23.5	12.8	44.1+10	16.4+5	171.2
Right Leaning	12.4	34.6	17.3	9.6	6.5	54.1+10	10.8+5	160.3
robbienicoll	14.4	5.7	11.7	25.9	8.5	46.1+10	6.4+5	133.7
Robert Waller	17.9	29.7	9.0	12.4+10	12.5	48.1+10	22.4+10	181.9
Yellow Peril	13.4	24.6	19.7	7.9	12.5+10	30.1+10	10.8+5	143.9
Total faults	393.0+30	472.7	413.1	414.0+40	283.5+3+70	735.7+140	307.8+90	3,392.8

and for the month to date:

	Weeks 1 -3	Week 4	Week 4	Weeks 1 - 4	Weeks 1 - 4
	faults	faults	position	faults	position
ajs	517.1	195.1	10th	712.2	8th
andrewp	499.2	136.0	2nd	635.2	2nd
David Boothroyd	549.9	187.7	9th	737.6	9th
dibs	551.7	283.5	14th	835.2	13th
greenrobinhood	520.8	275.9	12th	796.7	12th
hempie	569.7	178.4	7th	748.1	11th
k9	552.4	700		1,252.4	15th
Lancastrian	539.7	199.6	11th	739.2	10th
Olympian95	518.1	163.6	5th	681.7	5th
priceofdawn	687.9	281.9	13th	969.7	14th
ricmk	479.9	171.2	6th	651.1	3rd
Right Leaning	499.6	160.3	4th	659.9	4th
robbienicoll	550.8	133.7	1st	684.5	6th
Robert Waller	420.3	181.9	8th	602.2	1st
Yellow Peril	552.5	143.9	3rd	696.4	7th
Total	8,009.6	3,392.7		11,402.4

Objections please by noon Sunday when I will consider the 5 fault penalty given in Ockendon, based on the Official result showing 696 votes for each of the Conservative and Labour candidates, in the light of all representations.

There is just one election next Thursday - the final week of March.  Predictions on this thread by 9.00 on Thursday.

[andrewp]

Mar 23, 2018 18:00:28 GMT Pete Whitehead said:

Mar 23, 2018 17:40:17 GMT Davıd Boothroyd said:

I challenge all the 5 points. The Conservatives won in Thurrock, unequivocally, by one vote.

Have to agree. The Conservative candidate was elected - he won. Nobody who predicted that outcome should be getting any wrong winner faults

I declare an interest! But it does feel wrong to get penalty points for predicting a Conservative win. 

[yellowperil]

Congratulations to Robert for maintaining his lead for the month and given there is only one election to go next week he has almost certainly done enough to end the month still ahead. Congratulations too to Robbie for taking the week, by making the most of home advantage at the last stage- it was obviously Penicuik wot won it for him.

On the contentious issue of the 5-point deductions at Thurrock, I am glad we have middleenglander to make the difficult decisions for us. I can see both arguments. The Prediction Competition is a game of skill, not of luck,or so we try to kid ourselves. Why should the award of points then depend on a toss of a coin? On the other hand, if that coin toss decides who sits in the council chamber, and its good enough for that, why should it not decide who wins this much lesser prize?

We could (a) retain the penalties as originally decided by ME. The argument is that we were in the business of forecasting a clear-cut win, even if that had been only one vote in it, and everybody failed to achieve that. But those who forecast an Independent win were more wrong than those who forecast either a Lab or Con win.
or (b) delete the 5-point penalty for those forecasting a Conservative win, because they were ultimately right.
or (c) delete both sets of 5 point penalties on the basis that neither Lab nor Con were defeated in the actual poll.

My personal preference is for (b) (as I predicted a Con win I would do you might say),but I absolutely accept this is ME's call and I would entirely respect whatever decision he comes to, of course. The referee is always right.

[Pete Whitehead]

Mar 23, 2018 19:57:43 GMT yellowperil said:

Congratulations to Robert for maintaining his lead for the month and given there is only one election to go next week he has almost certainly done enough to end the month still ahead. Congratulations too to Robbie for taking the week, by making the most of home advantage at the last stage- it was obviously Penicuik wot won it for him.

On the contentious issue of the 5-point deductions at Thurrock, I am glad we have middleenglander to make the difficult decisions for us. I can see both arguments. The Prediction Competition is a game of skill, not of luck,or so we try to kid ourselves. Why should the award of points then depend on a toss of a coin? On the other hand, if that coin toss decides who sits in the council chamber, and its good enough for that, why should it not decide who wins this much lesser prize?

We could (a) retain the penalties as originally decided by ME. The argument is that we were in the business of forecasting a clear-cut win, even if that had been only one vote in it, and everybody failed to achieve that. But those who forecast an Independent win were more wrong than those who forecast either a Lab or Con win.
or (b) delete the 5-point penalty for those forecasting a Conservative win, because they were ultimately right.
or (c) delete both sets of 5 point penalties on the basis that neither Lab nor Con were defeated in the actual poll.

My personal preference is for (b) (as I predicted a Con win I would do you might say),but I absolutely accept this is ME's call and I would entirely respect whatever decision he comes to, of course. The referee is always right.

There are not two sets of 5 point penalties as nobody predicted a Labour victory. If there were people in that category it would perhaps be more arguable but as it is the only set of people with 5 penalty points are those who correctly predicted the winner of the election.

[yellowperil]

Mar 23, 2018 20:19:17 GMT Pete Whitehead said:

Mar 23, 2018 19:57:43 GMT yellowperil said:

Congratulations to Robert for maintaining his lead for the month and given there is only one election to go next week he has almost certainly done enough to end the month still ahead. Congratulations too to Robbie for taking the week, by making the most of home advantage at the last stage- it was obviously Penicuik wot won it for him.

On the contentious issue of the 5-point deductions at Thurrock, I am glad we have middleenglander to make the difficult decisions for us. I can see both arguments. The Prediction Competition is a game of skill, not of luck,or so we try to kid ourselves. Why should the award of points then depend on a toss of a coin? On the other hand, if that coin toss decides who sits in the council chamber, and its good enough for that, why should it not decide who wins this much lesser prize?

We could (a) retain the penalties as originally decided by ME. The argument is that we were in the business of forecasting a clear-cut win, even if that had been only one vote in it, and everybody failed to achieve that. But those who forecast an Independent win were more wrong than those who forecast either a Lab or Con win.
or (b) delete the 5-point penalty for those forecasting a Conservative win, because they were ultimately right.
or (c) delete both sets of 5 point penalties on the basis that neither Lab nor Con were defeated in the actual poll.

My personal preference is for (b) (as I predicted a Con win I would do you might say),but I absolutely accept this is ME's call and I would entirely respect whatever decision he comes to, of course. The referee is always right.

There are not two sets of 5 point penalties as nobody predicted a Labour victory. If there were people in that category it would perhaps be more arguable but as it is the only set of people with 5 penalty points are those who correctly predicted the winner of the election.

Ah my mistake - I'm getting a bit distracted and not concentrating properly, though probably if we were trying to work out the  first principles of this we should consider how we would have reacted to that possibility. I should have been aware there were no Labour win predictors as I was very nearly the one in that position- I only changed my mind at the last.

[Andrew_S]

Is this the first time there's been a tie in an election that's been part of the competition?

[Davıd Boothroyd]

Rule 49 of the Election rules: "Where, after the counting of the votes (including any recount) is completed, an equality of votes is found to exist between any candidates and the addition of a vote would entitle any of those candidates to be declared elected, the returning officer shall forthwith decide between those candidates by lot, and proceed as if the candidate on whom the lot falls had received an additional vote."

[ricmk]

Well done robbienicoll and Andrewp in particular. And thanks middleenglander as always.
I tend to agree that there shouldn't be 5 points penalties in Thurrock; the aim is to call the winner and several of us did so. Maybe I'm biased as I was one.

Thinking the unthinkable, why do we have these penalties anyway? if you make a crazy prediction you'll be penalised with a high score anyway. And in a genuinely close contest, it's a tossup anyway. If you've correctly guessed it'll be close you've done pretty well in my view.

[robbienicoll]

Thank you both YP and ricmk, was a very closely fought contest but very glad to have picked up my first week win!

[middleenglander]

Thank you for your thoughts on how to determine the additional faults for a wrong winner in Ockendon. Those responding all disagreed with my interpretation. But then many more did not object.

The aim of the competition is to predict the %age share of the votes and consequently who wins - we have not had a tie before. In this case the Official Result published by the Returning Officer has the Conservative and Labour candidates each polling 696 votes with the Thurrock Independent 150 plus behind. These are the votes used to calculate the shares and therefore the faults. If the official result had given the Conservatives one extra vote, that would have been included in the share calculation - but it was not how the result was published. After the counting of votes we did not know who had won, but it was either Conservative or Labour. That is therefore the rationale for giving those who predicted a Conservative win 5 faults - if anyone had predicted a Labour win they would also have got 5 faults.

The winner was determined by the Returning Officer picking one of two ballot papers from a box. That is when a new Councillor was elected.

However, I am aware of the objections and am minded to withdrew the 5 faults from the 10 entrants who predicted a Conservative win unless there is a significant adverse reaction. I will not change the table above - I do not think anyone's position has been affected - but will credit later this week when there is just one by-election.

The 10 faults for a wrong winner was a number of modifications to the rules made in the light of experience shortly after the competition started. There were in fact a number of instances where entrants were being highly placed even though they were getting several winners wrong. I consider this rule continues to be appropriate as long as the completion continues in its present form.

[yellowperil]

Thank you, middleenglander , for your considered and judicious reply. I think most objectors were concerned for the principle and the precedent rather than caring overmuch about the 5 points at stake which as you say may not change any positions on the list. I think you are right to retain the 10 point wrong winner faults. I am assuming that if this were to happen again(heaven forbid), the 5 points against the eventual winner would not be imposed, but the 5 points against the eventual loser by lot would be charged?

I notice that in fact it is still possible to score quite well in spite of getting several winners wrong, and there would be an argument if you want to give priority to picking a winner for the wrong-winner penalty to be significantly higher than 10 points. However 10 points is a reasonable compromise.

[ricmk]

Indeed a considered response from middleenglander - thank you. It's clear that in the rare event of a tie, returning officers have a couple of different approaches, either by declaring the scores level and separately noting who won the tiebreak, or adding the extra vote into the scores with or without a note on the declaration paper. Whatever is adopted here should be consistent for both approaches, as scores shouldn't be subject to the whim of the returning officer (it's bad enough being subject to the voters!)

However anyone who thinks that the referee is always right has clearly never been to a League One football match!

[Davıd Boothroyd]

I agree with a 'wrong winner' additional fault.

I've also long been of the opinion that in Scottish byelections, predictions should also state who wins after redistribution of preferences, and the wrong winner faults distributed on that basis not on plurality.

[Deleted]

Mar 26, 2018 9:35:26 GMT ricmk said:

Indeed a considered response from middleenglander - thank you. It's clear that in the rare event of a tie, returning officers have a couple of different approaches, either by declaring the scores level and separately noting who won the tiebreak, or adding the extra vote into the scores with or without a note on the declaration paper. Whatever is adopted here should be consistent for both approaches, as scores shouldn't be subject to the whim of the returning officer (it's bad enough being subject to the voters!)

However anyone who thinks that the referee is always right has clearly never been to a League One football match!

i used to go to conference south games; the referee was always alright I thought, except the year AFC Wimbledon got promoted

[yellowperil]

Mar 26, 2018 9:50:53 GMT Davıd Boothroyd said:

I agree with a 'wrong winner' additional fault.

I've also long been of the opinion that in Scottish byelections, predictions should also state who wins after redistribution of preferences, and the wrong winner faults distributed on that basis not on plurality.

I agree with the last point- I gave up trying to predict the eventual winner in Scottish by-elections because most other contestants did not, but at present we are regarding the winner as whoever was ahead on first preferences, and the interesting cases are those minority of elections where that is different from the eventual winner. Maybe we should make that a requirement next time we have a Scottish by-election?

[yellowperil]

Mar 26, 2018 9:35:26 GMT ricmk said:

Indeed a considered response from middleenglander - thank you. It's clear that in the rare event of a tie, returning officers have a couple of different approaches, either by declaring the scores level and separately noting who won the tiebreak, or adding the extra vote into the scores with or without a note on the declaration paper. Whatever is adopted here should be consistent for both approaches, as scores shouldn't be subject to the whim of the returning officer (it's bad enough being subject to the voters!)

However anyone who thinks that the referee is always right has clearly never been to a League One football match!

Oh, yes I have, but retain the old-fashioned view that the referee is always right even when he is clearly wrong.

[yellowperil]

It occurs to me that the next Scottish by-election where the discussion above, about predicting the final result rather than relying on first preferences only, would be the Highland election in ward 11 (Caol & Mallaig). That would be a particularly difficult one, as there are 3 Independents out of 6 candidates, and almost all of the councillors in this ward in the past have been Independents. This will also be the first by election for a good while where I am not going to post the details, though at this stage I am still spending the time on researching the election, just not posting the results! I am posting this on this thread as we haven't yet got the April thread up, but maybe we have to think whether we could be up for it?

[greenhert]

Prediction for the only local by-election of this week:

Knowsley MBC, Page Moss: Labour 72, Conservative 15, Green 10, UKIP 3.

[yellowperil]

Knowsley, Page Moss: Con 12, Lab 78, Green 6, UKIP 4