Prediction Competion - June 2019

[middleenglander]

I have been wrestling with what to do with the Forest of Dean result; not only that but I am cream crackered with moving house and fell asleep on the settee for much of the evening.

If we take average votes, including the average of the 3 Independents, then the Liberal Democrats have the highest share which no-one predicted apart from John Loony who had all 5 candidates on 20%. Furthermore no-one predicted a Liberal Democrat councillor whilst everyone but John had at least one Independent. All get 8.33 additional faults apart from John who gets 9 additional faults.

If we take the top candidate for each of the parties, including a slate of 3 Independents, then the Independent has the highest share. So only iainbhx, ricmk and robbienicoll get 5 additional faults for predicting Conservatives having highest share, with John Loony again 4 faults from this aspect. As with the average method everyone gets 3.33 additional faults for having no Liberal Democrat councillors but at least one Independent apart from John who gets 5 from this aspect.

I think the top candidate method reflects the outcome better than the average method so will use that for the result.

[johnloony]

Jun 21, 2019 23:15:41 GMT middleenglander said:

I have been wrestling with what to do with the Forest of Dean result; not only that but I am cream crackered with moving house and fell asleep on the settee for much of the evening.

If we take average votes, including the average of the 3 Independents, then the Liberal Democrats have the highest share which no-one predicted apart from John Loony who had all 5 candidates on 20%. Furthermore no-one predicted a Liberal Democrat councillor whilst everyone but John had at least one Independent. All get 8.33 additional faults apart from John who gets 9 additional faults.

If we take the top candidate for each of the parties, including a slate of 3 Independents, then the Independent has the highest share. So only iainbhx, ricmk and robbienicoll get 5 additional faults for predicting Conservatives having highest share, with John Loony again 4 faults from this aspect. As with the average method everyone gets 3.33 additional faults for having no Liberal Democrat councillors but at least one Independent apart from John who gets 5 from this aspect.

I think the top candidate method reflects the outcome better than the average method so will use that for the result.

I think that you should calculate the scores:

(a) treating the three Independent candidates as one group, using the average vote method for all parties;
(b) treating the three Independents as separate parties, using the average vote method for all other parties;
(c) treating the three Independent candidates as one group, using the highest vote method;
(d) treating the three Independents as separate parties, using the highest vote method for the other parties

and then all the people who entered the prediction competition should vote by STV to select two options from (a), (b), (c) and (d)
and then all the forum members should vote in a membership ballot between the two options selected
and then you should ignore whichever option is chosen by the membership ballot, and apply whichever set of calculations provides me with the maximum relative advantage compared with the other predictors.

[middleenglander]

Week 3

Authority	Forest of Dean	Isle of Wight	Merton	Neath    Port Talbot	Salford	South Ribble	Wandsworth	Week 3
Ward	Newent & Taynton	Whippinghan   & Osborne	Cannon Hill	Pelenna	Walkden       South	Farington West	Furzedown	faults
andrewp	27.8+3.3	64.8+10	28.0+10	41.6+10	16.0	26.8	18.8	257.3
David Boothroyd	33.8+3.3	54.8+10	40.4+10	61.6+10	23.6	25.2	7.2	280.0
greenrobinhood	30.5+3.3	84.8+10	19.3+10	51.6+10	10.0	24.8	13.2	267.6
iainbhx	29.8+8.3	63.9+10	26.0+10	49.6+10	19.0	24.8	18.0	269.5
John Loony	34.6+9	54.8+10	42.0+10	43.3+10	32.6+10	19.2	26.8	302.3
Olympian95	100	100	100	100	100	100	100	700
priceofdawn	25.8+3.3	51.3+10	37.8+10	61.3+10	21.6+10	44.8	22.0+10	318.0
rickmk	33.8+8.3	39.7+10	27.8+10	67.6+10	19.6	16.8	18.0	261.7
Right Leaning	35.8+3.3	68.8+10	30.0+10	63.6+10	18.0	24.8	15.2	289.7
robbienicoll	39.8+8.3	69.9+10	20.0+10	67.3+10	21.6	44.8	3.2	305.0
Robert Waller	33.8+3.3	64.8+10	27.7+10	65.6+10	14.0	20.8	20.0	280.2
Tony Otim	26.4+3.3	86.8+10	22.4+10	61.4+10	17.3	19.0	17.2	283.9
Toylyyev	28.5+3.3	58.6+10	17.1	55.6+10	10.0	14.8	54.6	262.6
Yellow Peril	37.8+3.3	65.9+10	24.9+10	55.6+10	13.0+10	18.8	12.0	271.5
Total faults	518.2+64	929.2+130	463.5+120	846.0+130	336.3+30	425.8	346.3+10	4,349.3

[middleenglander]

Weeks 1 - 3

Authority	Week 1 & 2	Week 3	Week 3	Weeks 1 - 3	Weeks 1 - 3
Ward	faults	faults	position	faults	position
andrewp	170.3	257.3	1st	427.6	2nd
David Boothroyd	171.0	280.0	7th	451.1	6th
greenrobinhood	193.3	267.6	4th	460.9	7th
iainbhx	158.0	269.5	5th	427.4	1st
John Loony	228.2	302.3	11th	530.6	12th
Olympian95	187.0	700		887.0	14th
priceofdawn	261.9	318.0	13th	579.9	13th
rickmk	200.3	261.7	2nd	461.9	8th
Right Leaning	186.1	289.7	10th	475.8	9th
robbienicoll	214.5	305.0	12th	519.4	11th
Robert Waller	152.5	280.2	8th	432.7	4th
Tony Otim	211.6	283.9	9th	495.5	10th
Toylyyev	178.2	262.6	3rd	440.8	5th
Yellow Peril	156.7	271.5	6th	428.1	3rd
Total faults	2,669.5	4,349.3		7,018.8

Objections please by 6 pm Sunday.

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

[yellowperil]

Thank you middleenglander for wrestling with the intractable  problem presented by Forest of Dean and coming up with a very reasonable solution in the circumstances! At least from my point of view (politically) it was a good problem to have, even though I was going to get hammered whichever calculation you made, given that I could not get away from the fact that I had predicted the lowest LD vote share of any of the contestants. Even though it had crossed my mind that Newent was barely in the Forest proper and might behave a bit differently, say more like Ross which did well for the LDs not so long ago, but had discounted that thought as unjustifiable optimism! 

As for falling asleep on the job, I did that one  yesterday too, and in consequence was nearly an hour late retrieving Eileen from her day care. I had finished all my jobs with 5 minutes  to spare, and just thought I would take a 5 minute rest before I needed to get into the car and drive the half hour to Tenterden. An hour later I checked my watch....

[iainbhx]

Ah, thanks for doing this week middleenglander especially with the joys of moving house which is an incredibly stressful and tiring time.

That was a very tough week and I can see it is very tight at the top for the final week.

[olympian95]

Was away and completely forgot

[johnloony]

I predict
Mansfield: MIF 52% Lab 31% Con 11% UKIP 6%

[Tony Otim]

Mansfield: MIF 37.7; Lab 36.2; Con 15.9; UKIP 10.2

[Toylyyev]

Mansfield DC, Sandhurst: Mansfield Independent Forum 42, Labour 33, Conservative 16, UKIP 9.

[greenhert]

Mansfield BC, Sandhurst: MIF 38, Labour 36, Conservative 19, UKIP 7.

[Deleted]

Lab 35
MIF 36
CON 18
UKIP 7

[AdminSTB]

Jun 26, 2019 19:59:03 GMT @priceofdawn said:

Lab 35
MIF 36
CON 18
UKIP 7

A victory for miffed voters.

[robbienicoll]

Sandhurst, Mansfield: Con 21, Lab 38, UKIP 4, MIF 37

[iainbhx]

LAB 38 MIF 37 CON 17 UKIP 8

[andrewp]

Mansfield Sandhurst MIF 44, Lab 36, Con 14, UKIP 6

[Robert Waller]

Mansfield DC, Sandhurst: Mansfield Independent Forum 39, Labour 36, Conservative 15, UKIP 10.

[yellowperil]

Mansfield, Sandhurst: Con 16, Lab 37, UKIP 9, MIF 38

[Davıd Boothroyd]

MANSFIELD Sandhurst: Lab 40, MIF 29, C 20, UKIP 11

[Right Leaning]

MANSFIELD Sandhurst: MIF 39, Lab 32, C 21, UKIP 8