Best Of

@viewcode

I got your message on the previous thread and will respond as requested.

I note this comment from you, which is a view share by some other posters -

"I hold the Widdecombe position on trans: namely, those who have gone all the way surgically/hormonally and can reasonably function as the opposite sex should be allowed to be legally considered as the opposite sex."

I realise that people cannot be expected to follow the ins and outs of complicated areas of law. But the ECHR ruled in 2017 in the case of Garçon & Nicot v France (followed in a case against Italy in 2018 and Romania in 2021 saying the same) that countries could NOT insist on surgery (with a high risk of sterility) as a condition for recognising a change in gender in law. The ECHR has ruled that a change gender identity must in certain circumstances be recognised in law, regardless of a persons physical body. This therefore means means that you cannot link gender recognition to surgery/hormone treatment and must give it to, for instance, a man who has made no changes to his body whatsoever or to a woman trying to get pregnant, if they claim to have changed their gender (re this last, see the recent ruling by Mr Justice Hayden in the "W" case - in October 2025).

Ms Widdecombe, as a former MP and Minister in the Home Office, should have done her research before opining on this or putting this forward as some sort of solution or compromise. It isn't. It is unlawful. This debate is persistently hampered by politicians and commentators putting forward "solutions" which are unlawful. It is unfair and unkind to all concerned to offer false solutions and to mislead about what the law says and has said for some time, whatever side of the debate anyone is on.

Out of interest, if it weren't unlawful (i.e. if the law or the ECHR in this case were changed), would you personally be happy with such a condition for recognising a change in gender in law?

I ask because I suspect many would agree with such a condition.

On topic, for those with access to Economist articles:

https://www.economist.com/interactive/britain/2025/12/04/our-new-model-captures-the-lottery-of-britains-electoral-system

Our election model considers what might happen in each of Britain’s 632 constituencies if an election were held today. But rather than giving a single prediction, we fine-tuned our model to show a range of possible outcomes, based on historical trends and the latest opinion polls. As illustrated above, Britain’s electoral system is highly uncertain. We drew on data from the past 80 years of elections to quantify this uncertainty, and estimate how it might unfold in an imaginary election.

I've read it (https://www.economist.com/interactive/britain/2025/12/04/our-new-model-captures-the-lottery-of-britains-electoral-system ).

Here are my initial thoughts

"...The starting point for our model is nationwide opinion polling..."

• That's your input, but what's your output?
  

"...We measured the difference between thousands of historical nationwide polls and the subsequent election results to estimate the likely polling error if an election were held today. We found that polling errors often lean towards the last election result (perhaps because undecided voters return to the party they supported last time) and encoded that into our model..."

• OK the best predictor of the next election is the last election, but I could have told them that. And they're modelling the polling error (between the poll and the vote?)
  

"...Next we considered how each region and nation might swing..."

• Fair enough, you've added a regional component. I like that. This is presumably a swing in the vote?
  

"...We fit a regression model to consider the historical relationship between a party’s vote share in a constituency in one election and the next, given the nationwide and regional picture..."

OK, you've built a model that inputs one or more polls, the estimated polling error and regional swing, and converts that to...an election vote in a region?

"...Finally, we translate these regional results to each constituency [using MRP]...."

• OK, got that,
  

So they have three models.

MODEL 0

• Inputs: lots of historical and regional polls up to the last election
  
• Method: Simulation
  
• Output: polling error and regional swing for the next election
  

MODEL 1

• Inputs: Today's poll(s)
  
• Method: Linear regression?
  
• Output: Today's election vote in the regions and nations taking polling error and regional swing into account
  

MODEL 2

• Input: election vote in the regions and nations taking polling error and regional swing into account
  
• Method: MRP
  
• Output: Today's election result for each constituency and hence today's election result
  

Not to mention their model is based on data from as far back as the 1945 election.

On topic, for those with access to Economist articles:

https://www.economist.com/interactive/britain/2025/12/04/our-new-model-captures-the-lottery-of-britains-electoral-system

Our election model considers what might happen in each of Britain’s 632 constituencies if an election were held today. But rather than giving a single prediction, we fine-tuned our model to show a range of possible outcomes, based on historical trends and the latest opinion polls. As illustrated above, Britain’s electoral system is highly uncertain. We drew on data from the past 80 years of elections to quantify this uncertainty, and estimate how it might unfold in an imaginary election.

I've read it (https://www.economist.com/interactive/britain/2025/12/04/our-new-model-captures-the-lottery-of-britains-electoral-system ).

Here are my initial thoughts

"...The starting point for our model is nationwide opinion polling..."

• That's your input, but what's your output?
  

"...We measured the difference between thousands of historical nationwide polls and the subsequent election results to estimate the likely polling error if an election were held today. We found that polling errors often lean towards the last election result (perhaps because undecided voters return to the party they supported last time) and encoded that into our model..."

• OK the best predictor of the next election is the last election, but I could have told them that. And they're modelling the polling error (between the poll and the vote?)
  

"...Next we considered how each region and nation might swing..."

• Fair enough, you've added a regional component. I like that. This is presumably a swing in the vote?
  

"...We fit a regression model to consider the historical relationship between a party’s vote share in a constituency in one election and the next, given the nationwide and regional picture..."

OK, you've built a model that inputs one or more polls, the estimated polling error and regional swing, and converts that to...an election vote in a region?

"...Finally, we translate these regional results to each constituency [using MRP]...."

• OK, got that,
  

So they have three models.

MODEL 0

• Inputs: lots of historical and regional polls up to the last election
  
• Method: Simulation
  
• Output: polling error and regional swing for the next election
  

MODEL 1

• Inputs: Today's poll(s)
  
• Method: Linear regression?
  
• Output: Today's election vote in the regions and nations taking polling error and regional swing into account
  

MODEL 2

• Input: election vote in the regions and nations taking polling error and regional swing into account
  
• Method: MRP
  
• Output: Today's election result for each constituency and hence today's election result
  

In similar vein, Waitrose has just emailed to let me know that I am the top buyer of their rosemary and garlic crackers on the island. Someone in their data-crunching department has too much time on their hands…

If you get an email like that you're not likely to rush out and buy more. You'll be thinking, do I need to see someone about my rosemary and garlic cracker dependency?

They are clearly highly addictive. Maybe that's why their name includes the word crack.

They certainly are! When, some years back, Waitrose discontinued its rosemary crackers, I was forced to prowl the island in search of a replacement, and found myself leaving the dog in the car while I went into Morrisons to buy six boxes at a time of their rosemary crackers, often attracting very suspicious glances from the staff member watching their self-checkout terminals. Once I was even driven to go into the local co-op for their similar product.

Now, Waitrose is back in the game, and the only embarrassment I have to endure is their delivery person dropping multiple boxes onto my doorstep every Thursday morning, while the dog sniffs about to see if I’ve remembered to order his favourite treats.

@viewcode

I got your message on the previous thread and will respond as requested.

I note this comment from you, which is a view share by some other posters -

"I hold the Widdecombe position on trans: namely, those who have gone all the way surgically/hormonally and can reasonably function as the opposite sex should be allowed to be legally considered as the opposite sex."

I realise that people cannot be expected to follow the ins and outs of complicated areas of law. But the ECHR ruled in 2017 in the case of Garçon & Nicot v France (followed in a case against Italy in 2018 and Romania in 2021 saying the same) that countries could NOT insist on surgery (with a high risk of sterility) as a condition for recognising a change in gender in law. The ECHR has ruled that a change gender identity must in certain circumstances be recognised in law, regardless of a persons physical body. This therefore means means that you cannot link gender recognition to surgery/hormone treatment and must give it to, for instance, a man who has made no changes to his body whatsoever or to a woman trying to get pregnant, if they claim to have changed their gender (re this last, see the recent ruling by Mr Justice Hayden in the "W" case - in October 2025).

Ms Widdecombe, as a former MP and Minister in the Home Office, should have done her research before opining on this or putting this forward as some sort of solution or compromise. It isn't. It is unlawful. This debate is persistently hampered by politicians and commentators putting forward "solutions" which are unlawful. It is unfair and unkind to all concerned to offer false solutions and to mislead about what the law says and has said for some time, whatever side of the debate anyone is on.

Joke mode on.

Only consider anyone a woman if they can multi-task - as no man can.

Only consider anyone a man if they can parallel park - an no woman can.

Joke mode off, and dives behind the sofa.

On that basis, I am neither a man nor a woman.

Sandy is a gender-neutral name?

On topic, for those with access to Economist articles:

https://www.economist.com/interactive/britain/2025/12/04/our-new-model-captures-the-lottery-of-britains-electoral-system

Our election model considers what might happen in each of Britain’s 632 constituencies if an election were held today. But rather than giving a single prediction, we fine-tuned our model to show a range of possible outcomes, based on historical trends and the latest opinion polls. As illustrated above, Britain’s electoral system is highly uncertain. We drew on data from the past 80 years of elections to quantify this uncertainty, and estimate how it might unfold in an imaginary election.

I've read it (https://www.economist.com/interactive/britain/2025/12/04/our-new-model-captures-the-lottery-of-britains-electoral-system ).

Here are my initial thoughts

"...The starting point for our model is nationwide opinion polling..."

• That's your input, but what's your output?
  

"...We measured the difference between thousands of historical nationwide polls and the subsequent election results to estimate the likely polling error if an election were held today. We found that polling errors often lean towards the last election result (perhaps because undecided voters return to the party they supported last time) and encoded that into our model..."

• OK the best predictor of the next election is the last election, but I could have told them that. And they're modelling the polling error (between the poll and the vote?)
  

"...Next we considered how each region and nation might swing..."

• Fair enough, you've added a regional component. I like that. This is presumably a swing in the vote?
  

"...We fit a regression model to consider the historical relationship between a party’s vote share in a constituency in one election and the next, given the nationwide and regional picture..."

OK, you've built a model that inputs one or more polls, the estimated polling error and regional swing, and converts that to...an election vote in a region?

"...Finally, we translate these regional results to each constituency [using MRP]...."

• OK, got that,
  

So they have three models.

MODEL 0

• Inputs: lots of historical and regional polls up to the last election
  
• Method: Simulation
  
• Output: polling error and regional swing for the next election
  

MODEL 1

• Inputs: Today's poll(s)
  
• Method: Linear regression?
  
• Output: Today's election vote in the regions and nations taking polling error and regional swing into account
  

MODEL 2

• Input: election vote in the regions and nations taking polling error and regional swing into account
  
• Method: MRP
  
• Output: Today's election result for each constituency and hence today's election result
  

In similar vein, Waitrose has just emailed to let me know that I am the top buyer of their rosemary and garlic crackers on the island. Someone in their data-crunching department has too much time on their hands…

If you get an email like that you're not likely to rush out and buy more. You'll be thinking, do I need to see someone about my rosemary and garlic cracker dependency?

They are clearly highly addictive. Maybe that's why their name includes the word crack.

@viewcode

I got your message on the previous thread and will respond as requested.

I note this comment from you, which is a view share by some other posters -

"I hold the Widdecombe position on trans: namely, those who have gone all the way surgically/hormonally and can reasonably function as the opposite sex should be allowed to be legally considered as the opposite sex."

I realise that people cannot be expected to follow the ins and outs of complicated areas of law. But the ECHR ruled in 2017 in the case of Garçon & Nicot v France (followed in a case against Italy in 2018 and Romania in 2021 saying the same) that countries could NOT insist on surgery (with a high risk of sterility) as a condition for recognising a change in gender in law. The ECHR has ruled that a change gender identity must in certain circumstances be recognised in law, regardless of a persons physical body. This therefore means means that you cannot link gender recognition to surgery/hormone treatment and must give it to, for instance, a man who has made no changes to his body whatsoever or to a woman trying to get pregnant, if they claim to have changed their gender (re this last, see the recent ruling by Mr Justice Hayden in the "W" case - in October 2025).

Ms Widdecombe, as a former MP and Minister in the Home Office, should have done her research before opining on this or putting this forward as some sort of solution or compromise. It isn't. It is unlawful. This debate is persistently hampered by politicians and commentators putting forward "solutions" which are unlawful. It is unfair and unkind to all concerned to offer false solutions and to mislead about what the law says and has said for some time, whatever side of the debate anyone is on.

Joke mode on.

Only consider anyone a woman if they can multi-task - as no man can.

Only consider anyone a man if they can parallel park - an no woman can.

Joke mode off, and dives behind the sofa.

I'm so bad at parallel parking - I'm clearly a women...

There you go. Make sure you use the appropriate toilet.

I always use the Laddies' toilet