Prediction Competition - December 2018

[yellowperil]

Dec 6, 2018 13:48:03 GMT peterl said:

Dec 6, 2018 9:13:51 GMT yellowperil said:

peterl : (Dec 3rd) "not a great month for me, but I'll persevere"

Bloody politicians - can depend on them not to keep their promises

Ha, I'm not a politician, I'm a community activist.

Eh, I was out late yesterday, slept in late and forgot. C'est la vie.

I guess we are all politicians though we may prefer to call ourselves something more flattering to the ego.  And yes I've fallen foul of the "bloody politicians thing before now when circumstances beyond control make it impossible to keep to the letter of a promise

[Tony Otim]

Dec 6, 2018 13:55:18 GMT yellowperil said:

Dec 6, 2018 13:26:04 GMT Tony Otim said:

Firstly, you seem to be suggesting that voters voting "honestly" for their actual preferences, rather than having to game the system and try to work out the best tactical vote is a bad thing?

Secondly, if a voter ranks the SNP candidate last of all the candidates, then that vote can never be used to elect the SNP candidate in any meaningful sense. It would only be added to the SNP total in a redundant last round after all other candidates are eliminated and therefore the SNP candidate is already elected whether or not votes transfer to them or not.

Thirdly, some voters may assume some of those things (although some of them would be pretty darned strange things to assume). In my experience, most voters either have a roughly accurate idea of how it works or understand that they need to rank candidates in order of preference and really aren't all that interested in how the count works.

That business of the superfluous final round and adding  the final  last votes to the already successful candidate is more than just a bit of unnecessary bureaucracy or an example of stupidly programmed machines not knowing when to stop. It really does enable opponents of the system  to say the lie implicit in Carlton's post -that you finish up voting for the party you like least.As you say, its not true in any meaningful way, but that's not a good line to pacify someone who is inherently suspicious that they're being conned. It should not be that difficult to make the computers stop at the point they have a winner.

Although nobody is obliged to use all their preferences anyway.

[yellowperil]

After the three English results.....ah well, never mind, I think I will forget about the December competition!

[andrewp]

Dec 7, 2018 8:13:23 GMT yellowperil said:

After the three English results.....ah well, never mind, I think I will forget about the December competition!

Did you work out anyone else’s?

[yellowperil]

Dec 7, 2018 8:53:24 GMT andrewp said:

Dec 7, 2018 8:13:23 GMT yellowperil said:

After the three English results.....ah well, never mind, I think I will forget about the December competition!

Did you work out anyone else’s?

Only a quick rough calculation but if I'm right we have a new leader in jamesrodriguez, closely followed by (of course) Robert Waller, but I hadn't the heart to do it with sufficient accuracy to be sure. Fairly sure James did best with the Leicester vote mountain none of us saw coming.

[andrewp]

Dec 7, 2018 9:45:31 GMT yellowperil said:

Dec 7, 2018 8:53:24 GMT andrewp said:

Did you work out anyone else’s?

Only a quick rough calculation but if I'm right we have a new leader in jamesrodriguez , closely followed by (of course) Robert Waller , but I hadn't the heart to do it with sufficient accuracy to be sure. Fairly sure James did best with the Leicester vote mountain none of us saw coming.

I have a feeling that the Scottish result today will be the key result in deciding this weeks standings! Anything could happen

[Robert Waller]

I agree that the Highlands will be decisive this week.
There doesn't seem to be much in it all round after the first three results, but my rough figures have James, Right Leaning, myself and Andrew P probably in the first four places, possibly in that order.
However I'm out on a limb on the Scottish winner ... with no confidence at all.

[andrewp]

Dec 7, 2018 9:53:55 GMT Robert Waller said:

I agree that the Highlands will be decisive this week.
There doesn't seem to be much in it all round after the first three results, but my rough figures have James, Right Leaning, myself and Andrew P probably in the first four places, possibly in that order.
However I'm out on a limb on the Scottish winner ... with no confidence at all.

I’ve just done rough figures too and concur with your top four, with about 2 points covering the top three.  Anyone could be top after Highland. I suspect many of us could get more faults in that than the other three elections combined, 

[yellowperil]

Dec 7, 2018 9:53:55 GMT Robert Waller said:

I agree that the Highlands will be decisive this week.
There doesn't seem to be much in it all round after the first three results, but my rough figures have James, Right Leaning, myself and Andrew P probably in the first four places, possibly in that order.
However I'm out on a limb on the Scottish winner ... with no confidence at all.

more or less agree with all of that, except my quick calculation had you just pipping RL for second place, but well within the margin of error. Highlands could be not only decisive but a massacre, and like you I am out on a limb, just a different limb I share with PoD, which doesn't overfill me with confidence.

[andrewp]

Dec 7, 2018 10:58:41 GMT yellowperil said:

Dec 7, 2018 9:53:55 GMT Robert Waller said:

I agree that the Highlands will be decisive this week.
There doesn't seem to be much in it all round after the first three results, but my rough figures have James, Right Leaning, myself and Andrew P probably in the first four places, possibly in that order.
However I'm out on a limb on the Scottish winner ... with no confidence at all.

more or less agree with all of that, except my quick calculation had you just pipping RL for second place, but well within the margin of error. Highlands could be not only decisive but a massacre, and like you I am out on a limb, just a different limb I share with PoD, which doesn't overfill me with confidence.

Looking better for you now!

[middleenglander]

Have not got the time to do a full verification and posting of Week 1 but assuming SNP ultimately beats Conservative to the Highland seat it looks like Right leaning 2.7 faults ahead of David Boothroyd, andrewp, Robert Waller, greenrobinhood and priceofdawn. However there are just 37.3 faults separatin top from bottom with only 17 the top 10.

[Deleted]

Dec 7, 2018 10:58:41 GMT yellowperil said:

Dec 7, 2018 9:53:55 GMT Robert Waller said:

I agree that the Highlands will be decisive this week.
There doesn't seem to be much in it all round after the first three results, but my rough figures have James, Right Leaning, myself and Andrew P probably in the first four places, possibly in that order.
However I'm out on a limb on the Scottish winner ... with no confidence at all.

more or less agree with all of that, except my quick calculation had you just pipping RL for second place, but well within the margin of error. Highlands could be not only decisive but a massacre, and like you I am out on a limb, just a different limb I share with PoD, which doesn't overfill me with confidence.

worked out quite well after all

[yellowperil]

Now we know the Highland result was indeed a gain for SNP ,@priceofdawn and I stuck our necks out for that result -and won! And for that we gain a magnificent 5 points on the rest of you! Not complaining-those were the rules of engagement we all accepted, but it does seem to me that rather underplays the issue of final winner in those Scottish by-election. I wonder if middleenglander could consider raising the bar in future, so say it was 10 points for forecasting the leader on first prefs and a further 10 for predicting the winner at the final stage, rather than 5+5?

[Robert Waller]

I feel that YP's suggestion would privilege Scottish elections by making their 'winner' worth twice as many faults as others.
Surely every election should be 'equal'?
There is a case for having the 10 points for the eventual winner, and nothing for first preferences in an AV election (whatever some people's objections, it is the system in use).
But speaking for myself, I'm happy with ME's current system.

On the weekly results, I think it's clear that Right Leaning is in first place - very well done to him! - and also that everyone's still in it.

[Tony Otim]

Dec 7, 2018 14:29:36 GMT yellowperil said:

Now we know the Highland result was indeed a gain for SNP ,@priceofdawn and I stuck our necks out for that result -and won! And for that we gain a magnificent 5 points on the rest of you! Not complaining-those were the rules of engagement we all accepted, but it does seem to me that rather underplays the issue of final winner in those Scottish by-election. I wonder if middleenglander could consider raising the bar in future, so say it was 10 points for forecasting the leader on first prefs and a further 10 for predicting the winner at the final stage, rather than 5+5?

I'm not sure you were sticking your necks out that much. I always thought the SNP were favourites, but decided to have a little gamble on the Indy and then everybody else seemed to follow me...

[yellowperil]

Dec 7, 2018 16:37:04 GMT Robert Waller said:

I feel that YP's suggestion would privilege Scottish elections by making their 'winner' worth twice as many faults as others.
Surely every election should be 'equal'?
There is a case for having the 10 points for the eventual winner, and nothing for first preferences in an AV election (whatever some people's objections, it is the system in use).
But speaking for myself, I'm happy with ME's current system.

On the weekly results, I think it's clear that Right Leaning is in first place - very well done to him! - and also that everyone's still in it.

I understand the point you are making about privileging Scottish elections, though I might counter that be saying it merely reflects the additional amount of work predictors have to undertake. Anyway, the wrong winner points are nearly always rather peripheral to those which come from the vote share predictions. I would not be opposed to your compromise proposal of loading all the wrong winner points on the eventual winner. Congratulations to Right Leaning  for coming through relatively unscathed.

[Deleted]

Thanks, it has only taken me 11 months of taking part to win a week, and don't worry I am sure next week I will go back to my normal 9th place :-)

[middleenglander]

Week 1

Authority	Highland	Leicester	Oxford	Surrey	Week 1	Week 1
Ward / Division	Wester Ross,     Strathpeffer Lochalsh	Belgrave	Wolvercote	The Byfleets	faults	position
andrewp	34.2+5	33.2	9.0	17.9	99.3	4th
Casual Observer	42.2+5	33.2	9.0	33.2+10	132.6	14th
David Boothroyd	21.3+5	25.4	13.0	23.2+10	98.1	3rd
electionmapsUK	42.2+5	37.2	9.0	19.9	113.3	11th
greenrobinhood	30.3+5	41.2	7.9	19.2	103.6	5th
hempie	32.2+5	35.2	11.4	25.9+10	119.7	12th
jamesrodriguez	56.2+10	19.2	13.0	21.9	120.3	13th
Olympian95	36.2+5	29.2	13.4	23.9	107.6	8th
priceofdawn	31.4+2	43.2	9.9	17.2	103.7	6th
Right Leaning	34.2+5	29.4	9.4	17.2	95.3	1st
robbienicoll	36.2+5	39.2	5.9	21.9	108.3	9th
Robert Waller	35.4+5	29.2	7.0	19.9	96.6	2nd
Tony Otim	33.0+5	40.4	11.2	17.6	107.2	7th
Yellow Peril	29.3	29.2	12.6	31.2+10	112.3	10th
Total faults	494.4+2+65	464.0	141.9	310.7+40	1,517.9

I mistakenly gave Robert Waller 10 faults on Friday for a wrong winner in the Highlands whereas it should have been 5 as he correctly predicted SNP would be ahead on first preferences.

Objections please by 9.00 am Tuesday.

There are 5 by-elections next week, 1 on Wednesday and 4 on Thursday.  Predictions on this thread by 9.00 am on day of by-election.

[middleenglander]

Until recently Scottish by-elections had 10 faults for predicting the wrong candidate ahead on first preferences and no cognisance of the one ultimately elected. By popular demand it was changed to recognise both who was ahead on first preferences and who was eventually elected. I think we need time to asses how this works - otherwise we may be in the syndrome of the Dangerous Dogs Act!!

I think awarding up to 20 faults for incorrect predictions in Scotland is wrong when the penalty for such wrong predictions in England and Wales is up to 10 faults. The Government may consider this another reason for amending the system of election that Carlton43 so clearly thinks needs to be changed!!

[andrewp]

I agree that we shouldn’t be able to incur more wrong winner faults on a Scottish election than any other, so would prefer it to be limited to a maximum of 10.

I personally would base the faults entirely on the first preference result, but am happy with the current 5/5 split as a compromise.