
Nigel Marriott
@MarriottNigel
Followers
2K
Following
8K
Media
1K
Statuses
8K
Independent statistician using data to understand our world & to predict the future. Currently focused on pay gaps, diversity, elections, forecasting, surveys.
Bath, England
Joined June 2018
I now offer a newsletter you can subscribe to here. You can select from these categories to receive notifications of latest news, articles & offers. - Diversity & Pay Gaps.- Forecasting.- Elections.- Surveys & Polling.- Sport.- Weather.- #Stats Training.
0
0
8
All 32 @BBCSport pundits predict the same 4 clubs as last season making up the EPL Top 4 this season. Odds are they're all wrong. The English top flight top 4 has only ever repeated in 5 seasons. 2001, 2007, 2008, 2009, 2021. BBC groupthink?.
bbc.co.uk
We ask 32 BBC Sport football pundits to predict who will win the Premier League title, and which clubs will finish in the top four this season.
2
0
0
Some good points here from Chris about the quality of economic statistics in the UK and US. Concerning.
Trump showed how not to fire a statistics chief. but if you think the immediate departure of the UK's statistics chief ahead of a very critical report was any better, think again. My newsletter.
0
0
1
RT @__paleologo: I did not know that, for any random variable x, . | mean(x) - median(x) | <= stdev(x). Direct proof: .
0
123
0
RT @MarriottNigel: Is 2025 now the new normal for British politics?. Other parties combined (ie not CON, LAB & LD) have led the polls since….
0
1
0
High school cliques continue to rule the Netherlands as my wife would say!. A governing coalition would need a minimum of 4 parties.
Netherlands, Ipsos I&O poll:. PVV-PfE: 17% (-2).GL/PvdA-G/EFA|S&D: 15% (-3).CDA-EPP: 15% (+1).VVD-RE: 13% (-2).D66-RE: 8% (+1).JA21~ECR: 5% (+3).SP-LEFT: 4%.PvdD-LEFT: 4%.BBB-EPP: 3%.DENK-*: 3% (+1).SGP-ECR: 3%.CU-EPP: 3% (+1).Volt-G/EFA: 3% (+1).FvD-ESN: 2%.NSC-EPP: 1%. /- vs.
0
0
0
RT @salonium: I'm very proud to have won the Royal Statistical Society's award for best statistical commentary by a non-journalist this yea….
0
75
0
RT @MarriottNigel: The right & wrong way to read polls today. @BritainElects posts 2 polls with @FindoutnowUK showing a 14pt lead & @YouGo….
0
1
0
@FindoutnowUK have achieved a notable distinction in their last 2 polls which I hope @markpack can confirm from his Pollbase data. They're the 1st pollster ever to show 50% of GB voters DO NOT intend to vote for the big 3 parties of CON+LAB+LD. 6/n.
Westminster Voting Intention:. RFM: 30% (-1).LAB: 20% (-2).CON: 17% (-2).LDM: 13% (=).GRN: 12% (+3). Via @FindoutnowUK, 16 Jul. Changes w/ 9 Jul.
1
0
0
But the underlying stability of the Right & the Left is the story of 2025 polls. The difference between @FindoutnowUK & @Yougov is enough to suggest an underlying difference in their methodologies. Has someone with with more knowledge than me written anything about this?. 5/n
3
1
1
The right way is to count-. The Right = REF + CON.The Left = LAB + LD + GRN. So latest polls are-. @FindoutnowUK .Right 50%.Left 44%. @YouGov .Right 44%.Left 49%. YG & FON are the 2 ends of the R v L spectrum over the last 30 days. Avg Last 25 polls.Right 47%.Left 46%. 2/n
3
0
1
The right & wrong way to read polls today. @BritainElects posts 2 polls with @FindoutnowUK showing a 14pt lead & @YouGov a 4pt lead for REF over LAB. Looking at polls by party is the wrong way to read them. A thread on the right way to read the polls. 1/6
2
1
2
RT @lakens: The first step to publishing null results is to learn about equivalence testing, and design studies that statistically support….
0
22
0