Prediction Competition - October 2018

[yellowperil]

Oct 25, 2018 13:04:20 GMT yellowperil said:

Oct 25, 2018 10:54:55 GMT andrewp said:

Could be a good night for the Lib Dems. There are 3 gains that most/ all of us think they will acheive.

Yes,trouble with that is that it suddenly becomes a bad night if in the end there are "only" a couple of gains. Several of those (3 and more ?)could indeed go either way and I'm a bit wary about that convergence of opinion on this thread, bit of herding going on there I think as few of us have dared to step out of line when the competition is so tight. Has happened a number of times lately that we finish up with 100% wrong!

A lot of these seats are ones where the Lib Dems were previously very strong but where it has since gone pear-shaped. The question is whether the underlying strength will shine through, or whether the reasons why things went wrong still apply. If the latter, results might not live up to expectations.

This bit of prediction (unlike my entry in the competition per se) looks increasingly spot on, and I will confess it was the Suffolk case I had at the back of my mind as I wrote those words, also Sutton when I suggested more than 3 could be in play.

[andrewp]

Oct 26, 2018 0:31:07 GMT yellowperil said:

Oct 25, 2018 13:04:20 GMT yellowperil said:

Yes,trouble with that is that it suddenly becomes a bad night if in the end there are "only" a couple of gains. Several of those (3 and more ?)could indeed go either way and I'm a bit wary about that convergence of opinion on this thread, bit of herding going on there I think as few of us have dared to step out of line when the competition is so tight. Has happened a number of times lately that we finish up with 100% wrong!

A lot of these seats are ones where the Lib Dems were previously very strong but where it has since gone pear-shaped. The question is whether the underlying strength will shine through, or whether the reasons why things went wrong still apply. If the latter, results might not live up to expectations.

This bit of prediction (unlike my entry in the competition per se) looks increasingly spot on, and I will confess it was the Suffolk case I had at the back of my mind as I wrote those words, also Sutton when I suggested more than 3 could be in play.

Yes, still 1 or 2 Lib Dem gains ( depending on how one treats Wells). I was surprised about Suffolk. 

[andrewp]

Oct 26, 2018 0:00:03 GMT yellowperil said:

I am interested in how the wrong winner faults are distributed in the Scottish election ( this is sort of a preemptive  appeal in advance of any pronouncements by middleenglander !) Most of us are now remembering to indicate our eventual winner after redistribution, and in a number of cases it was different from the leader after first preferences only. One or two out of 15 didn't do so, but that could be taken as indicating they expected the leader on first prefs to stay ahead. I would argue in this case those of us who showed the SNP ahead (or tied )at stage 1 but indicated an eventual Lab win should not pick up wrong winner faults.

I think I make it that Robert and Hempie are pretty much exactly tied with Basingstoke and Coatbridge to go, with myself about 2 faults behind. So how wrong winner faults are treated in Coatbridge could be crucial to the outcome. 

[yellowperil]

Oct 26, 2018 7:02:28 GMT andrewp said:

Oct 26, 2018 0:00:03 GMT yellowperil said:

I am interested in how the wrong winner faults are distributed in the Scottish election ( this is sort of a preemptive  appeal in advance of any pronouncements by middleenglander !) Most of us are now remembering to indicate our eventual winner after redistribution, and in a number of cases it was different from the leader after first preferences only. One or two out of 15 didn't do so, but that could be taken as indicating they expected the leader on first prefs to stay ahead. I would argue in this case those of us who showed the SNP ahead (or tied )at stage 1 but indicated an eventual Lab win should not pick up wrong winner faults.

I think I make it that Robert and Hempie are pretty much exactly tied with Basingstoke and Coatbridge to go, with myself about 2 faults behind. So how wrong winner faults are treated in Coatbridge could be crucial to the outcome. 

Just going by ME's officially declared results, so with 5 to go. I make it there was only 4 points between the top 4, and indeed Robert and Hempie were exactly tied at that point!This could be decided by drawing lots!

[middleenglander]

I failed to include that the ultimate winner in North Lanarkshire had to be predicted. In these circumstances either I assess this particular result excluding the final winner, in which case 11 entries get 10 faults, andrewp, Right Leaning & Robert Waller get 0 with ricmk 5, or I assess the ultimate wrong winner faults, in which case 4 entrants did not predict this. I am minded to do the latter but do not believe the 4 should be overtly penalised as I bear a large part of the blame.

Right Leaning & Robert Waller predicted Labour both ahead on first preferences and ultimate winner so clearly get 0 wrong winner faults.
David Boothroyd, robbienicol and Tony Otim predicted SNP both ahead and ultimate winner so get 10 wrong winner faults.
europeanlefty, hempie, Luke Dickinson, priceofdawn and Yellow Peril all predicted SNP ahead and Labour ultimate winner so get 5 faults
ricmk predicted a tie Labour / SNP on first preferences with a Labour win so gets 2½ faults

That leaves the 4 who did not predict the ultimate winner where I have to use some judgement:
andrewp predicted Labour ahead on first preferences so I will award 0 faults,
greenrobinhood, Olympian95 and ThirdChill all predicted SNP ahead on first preferences, but only narrowly so (4%, 4%, 2%), which will incur 5 faults
- I award 2½ faults for the wrong ultimate winner as it is not clear what they would have predicted thereby giving 7½ faults in total.

So andrewp, Right Leaning and Robert Waller would get the same 0 faults in both possible ways.
David Boothroyd, robbienicol and Tony Otim would get the same 10 faults in both possible ways
europeanlefty, hempie, Luke Dickinson, priceofdawn and Yellow Peril get 5 faults instead of 10 the first way
ricmk gets 2½ faults as opposed to 5 faults
greenrobinhood, Olympian95 and ThirdChill gets 7½ faults rather than 10.

[yellowperil]

Oct 26, 2018 8:46:58 GMT middleenglander said:

I failed to include that the ultimate winner in North Lanarkshire had to be predicted. In these circumstances either I assess this particular result excluding the final winner, in which case 11 entries get 10 faults, andrewp, Right Leaning & Robert Waller get 0 with ricmk 5, or I assess the ultimate wrong winner faults, in which case 4 entrants did not predict this. I am minded to do the latter but do not believe the 4 should be overtly penalised as I bear a large part of the blame.

Right Leaning & Robert Waller predicted Labour both ahead on first preferences and ultimate winner so clearly get 0 wrong winner faults.
David Boothroyd, robbienicol and Tony Otim predicted SNP both ahead and ultimate winner so get 10 wrong winner faults.
europeanlefty, hempie, Luke Dickinson, priceofdawn and Yellow Peril all predicted SNP ahead and Labour ultimate winner so get 5 faults
ricmk predicted a tie Labour / SNP on first preferences with a Labour win so gets 2½ faults

That leaves the 4 who did not predict the ultimate winner where I have to use some judgement:
andrewp predicted Labour ahead on first preferences so I will award 0 faults,
greenrobinhood, Olympian95 and ThirdChill all predicted SNP ahead on first preferences, but only narrowly so (4%, 4%, 2%), which will incur 5 faults
- I award 2½ faults for the wrong ultimate winner as it is not clear what they would have predicted thereby giving 7½ faults in total.

So andrewp, Right Leaning and Robert Waller would get the same 0 faults in both possible ways.
David Boothroyd, robbienicol and Tony Otim would get the same 10 faults in both possible ways
europeanlefty, hempie, Luke Dickinson, priceofdawn and Yellow Peril get 5 faults instead of 10 the first way
ricmk gets 2½ faults as opposed to 5 faults
greenrobinhood, Olympian95 and ThirdChill gets 7½ faults rather than 10.

the wisdom of Solomon.......

[Robert Waller]

Whatever the month's results, what is absolutely clear is ricmk's terrific performance this week!

[yellowperil]

Oct 26, 2018 11:21:53 GMT Robert Waller said:

Whatever the month's results, what is absolutely clear is ricmk's terrific performance this week!

Well,quite, Ric was quite brilliant......

.. but then we all know whatever the month's results will bring, don't we? Well done yet again for such bloody dogged consistency!

[middleenglander]

Week 4
amended for South Derbyshire, Linton revised Labour vote

Authority	Ashford	Basingstoke  & Deane	Dorset	East Dorset	Hertfordshire	Mendip	North  Lanarkshire	South  Derbyshire	Suffolk	Sutton
Ward	Kennington	Norden	Ferndown	Ferndown    Central	Three Rivers Rural	Wells  St Thomas	Coatbridge South	Linton	Bosmere	Belmont
andrewp	28.4	10.1	15.2	12.8	10.7	8.5	15.5	26.2	15.5+10	17.3
David Boothroyd	14.3	20.3	19.8	19.9	5.7	20.5	23.2+10	30.3	13.5+10	9.3
europeanlefty	31.0	28.8	11.2	8.8	22.8	8.5	15.5+5	26.2	5.5+10	25.6+3
greenrobinhood	37.0	18.1	17.2	12.8	17.7	5.0	25.3+7.5	22.2	7.5+10	27.2
hempie	28.4	18.1	19.8	16.8	13.7	11.5	10.2+5	26.2	7.5+10	6.8
Luke Dickinson	34.4	6.1	17.8	32.7	15.7	6.5	27.5+5	36.2	11.0+10	16.6+4
olympian95	48.3	12.1	22.2	23.3	14.7	7.5	25.3+7.5	16.2	11.5+10	7.9
priceofdawn	54.9	25.6+5	7.8	2.8	32.8	28.5	45.3+5	26.2	29.5+10	35.3
ricmk	33.0	22.1	7.8	2.8	9.7	3.0	5.5+2.5	6.2	3.0	10.8
Right Leaning	16.4	13.2	21.2	20.8	13.7	8.5	10.3	16.2	11.5+10	4.8
robbienicoll	28.4	16.3	29.8	30.7	5.7	21.0	13.2+10	26.2	7.0+10	15.3
Robert Waller	30.4	12.3	17.2	10.8	16.7	11.5	5.5	22.2	3.5+10	8.5
Third Chill	46.3	18.8	29.8	28.7	17.7	12.5	17.2+7.5	38.2	21.5+10	15.3
Tony Otim	38.2	12.1	11.5+0.3	15.9	14.7	9.5	18.6+10	17.0	4.9+10	20.8
Yellow Peril	34.4	12.3	15.2	16.1	13.7	3.0	13.5+5	20.2	5.0+10	14.5
Total	503.7	246.5+5	263.6+0.3	255.7	225.9	165.4	271.5+80	356.4	158.2+140	236.3+7

[middleenglander]

Month of November October

Authority	Week 1 - 3	Week 4	Week 4	Month	Month
Ward	faults	faults	position	faults	position
andrewp	211.5	170.3	5th	381.8	3rd
David Boothroyd	282.5	196.8	8th	479.3	10th
europeanlefty	288.9	201.9	9th	490.8	11th
greenrobinhood	331.4	207.6	11th	538.9	14th
hempie	194.5	174.1	6th	368.6	2nd
Luke Dickinson	286.8	223.5	13th	510.3	12th
olympian95	251.5	206.7	10th	458.2	8th
priceofdawn	505.0	308.7	15th	813.7	15th
ricmk	319.9	106.4	1st	426.3	6th
Right Leaning	310.9	146.7	2nd	457.6	7th
robbienicoll	248.7	213.6	12th	462.3	9th
Robert Waller	198.1	148.8	3rd	346.9	1st
Third Chill	255.2	263.6	14th	518.8	13th
Tony Otim	206.4	183.6	7th	390.0	5th
Yellow Peril	226.8	163.1	4th	389.9	4th
Total	4,118.0	2,915.4		7,033.4

Objections please by 5.00 pm Sunday.

New thread for November.

[hempie]

Congratulations to Robert for the month and to ricmk for a tremendous final week.

[Tony Otim]

What he said ^

[yellowperil]

I have to say I was quite pleased with 4th place both for the week and month but I do note my worst election prediction for the week was in Ashford where you might think I'd know what was going on, but obviously I didn't. In fact I always seem less able to make a balanced assessment when I am close by,and I don't think I am unique in that.

[ricmk]

Oct 26, 2018 11:21:53 GMT Robert Waller said:

Whatever the month's results, what is absolutely clear is ricmk's terrific performance this week!

Goodness me! Talk about a stopped clock being right twice a day and all that - thanks for your kind words.

but I confess I'm giggling at being congratulated by Robert Waller - I almost didn't post this week and had 2 kids to get to nursery on Thursday morning.  So in a panic at 8:30 I copied Robert Waller's predictions, tweaked them all a bit, just enough so no-one would notice, and posted. To be complimented by the man himself is like Mary Berry praising you for eating one of her cakes....

i thought there would be huge sympathy for the Tories in South Derbyshire given the circumstances, and know that 3 Rivers and Sutton Lib Dems have stellar reputations as campaigners - Suffolk was a bit of a punt, but I thought that a death causing the by-election, a very rural ward and a Lib Dem team not quite on the same level made it loseable.

im sure I'll be back in the tailenders next time.  Thanks middleenglander as always.

[yellowperil]

It takes real talent to take Robert's predictions and then improve on them!

Do like the notion of Robert as the Mary Berry of the psephological world.

[yellowperil]

Have only just noticed that the final results table for the month is headed November rather than October!