Surprise results I found as an election Count Agent 

Thanks for asking me on Matt! Lovely to meet you (and the Count) at the count on what was a very very long night for both of us. Your next challenge is trying to explain the maths of how MRP polling projections work.

"It's not good footage" just as Rishi Sunak strolls through the shot. Nice.

I do want to point out that "the only paper I have on me happens to be a hexagon" is the most on brand expected thing I've heard.

So close, if only Labour, Lib Dems, Reform and Greens hadn't split the Binface vote.

Professional counter and first-time count agent recounts his account of counting the count's counts, thus making sure every vote counts.

I love the tradition of having all the candidates gather at the count. It will never not be funny for the PM to have to stand on the same stage as Count Binface.

I knew they were running in the same constituency, but seeing the previous prime minister in the same room as someone dressed as a trash bin was quite surreal.

This has got to be the first time in Matt's life where he wandered into an event and just hung out for a couple hours before the thought occurred, "Hey, I should be doing some math." Binface must have been horrified at hearing this story

Data scientist here. 86.7% on newly drawn maps, snap election, and a lot of swing vote changes and poor turnout nationally is amazing accuracy for a model. And arguably most 2nd places were close so mathematically, the accuracy within a 95% confidence interval is likely more, it's just those ones where a few hundred swung it.

Part of why the Lib Dem vote translated into seats so efficiently is tactical voting. In general, supporters of Labour and the Lib Dems were happy to vote for each other in order for 1 of the pair to win. That means that normally, you'd ONLY VOTE FOR the Lib Dems if they had a strong chance of winning. In contrast, supporters of the Conservatives and Reform are often sick of the other party and wouldn't consider switching. Therefore they tended to compete with each other, often both losing when one of them could have won.

I've worked on several election counts, and I found the rigour and transparency of the process hugely reassuring for my faith in democracy. Just gutted that I was working in the wrong part of Yorkshire to meet you and the Count at the count!

Your explanation of Libdem vs Reform vote share is so good. Makes it clear why they got so many more seats.

9:48 Loving the footage and photos where soon-to-be-just-another-MP Rishi Sunak gets to be in the same frame as Count Binface

Excellent video. For years now I've been a candidate, election agent and counting agent for the Green Party. The tallying you describe yes is done for finding out if your candidate is winning or not, but also it's done to find out how well each candidate has done in each polling district. The verification is done ballot box by ballot box (IE polling district by polling district). That's when the tallying happens. After the verification is finished, the votes are then mixed together so we can no longer tell where each of the votes came from. It is useful for knowing where your support is and where you might want to work harder to get your vote up.

Worth mentioning that even despite the LDs' 72 seats, they still have fewer than they'd have got if seats were properly proportional

I counted 53 Counts - 50 by Matt. 0:05 0:08 0:20 0:22 0:23 1:12 1:18 1:32 1:44 1:49 1:52 1:54 2:00 2:15 2:31 2:33 2:42 2:49 3:05 3:12 3:28 3:40 3:44 3:52 3:57 4:00 4:03 4:27 4:36 4:38 5:07 5:10 5:20 5:22 5:28 6:02 6:16 6:22 7:54 7:56 7:57 9:18* 9:58 10:20 10:40 12:08 12:58 17:09 17:10 22:08 22:16* 22:18* 22:31 23:18 23:39 23:45

I was a vote agent for the labour party this year. Sampling the vote is less about estimating the overall result (we'll know that soon anyway), but more knowing the local data for each polling station. These data let us know the topology of our support and where to concentrate our efforts in future campaigns

I am not a native English speaker, and I was wondering the entire time what Count Bin Face had to do with counting and some kind of binface, whatever that is lol, eventhough he clearly looked like a robot mimic. Until I realized he is a Count that has a Bin as his face. :facepalm:

I cannot comprehend how someone could vote for the Tories over Binface.

Please send a copy of the binface leaflet to the university of Bristol elections archive!

15:33 Lol @ saying views instead of votes, classic RU-vid

I've counted votes too a couple of times. Good times all around. Mainly city council elections in Finland, so relatively small, but still thousands and thousands of votes of course.

Outside the Count Room: Everyone's vote counts Inside the Count Room: Everyone counts votes

I can see why Matt likes COUNT Binface so much

I know Matt was a teacher but seeing him in a suit and a landyard really shows it 😂

Matt I have helped at elections here in Australia. I found it to be a amazing expirence. i was proud to see how pleasent everyone was and how well the count was run. I am glad that you had the oppertunity to do the same.

Never would have expected Niko and Stand-up Maths to be in the same room

Preferential voting is the obvious first step before worrying about proportional systems.

I love that shot of Rishi Sunak doing a double-take when he spots Count Binface, hilarious.

Having worked in U.S. politics for a couple of decades now myself, this video is a delight. Thanks Matt, and congrats to Count Binface on his record night!

Next big Pi video idea Calculating Pi with bayesian ballot counting inference. Just watch at 4 am as Matt and the gang all smiling doing calculations as 6 MP hopefuls glare. "Just count them already!" 😂

The tally is also to get an idea of the vote breakdown in each ward of the constituency as this is not officially recorded. Specific wards may have been targeted with leafletting and canvassing.

9:50- Sunak's face is just priceless here, eyes of incredulity glued to that big ol binhead.

Maths aside, seeing Matt Parker walking down the same high street I drive down not too infrequently (and used to walk down on my breaks while at college), past a shop I used to work at, and standing next to a pool I used to swim in, is just... weird. Like "oh, hey, I know that place, I was there last w-... WHAT IS HE DOING THERE?!!" 🤣

You should also plot those Lib Dem results against the Labour ones. Both parties shared their votes very efficiently

Speaking as a Liberal Democrat you have described it perfectly; whereas previously we had done Proportional Representation Campaigns, but this time we were doing a FPTP Campaign. If you were in a Target Seat you saw a lot of Liberal Democrat coverage, whereas in non-Target Seats it was down to the Local Candidate to run their own campaign. This is why we achieved a much greater success than we have done previously.

Pentagonal paper for the civil service should be a policy on Count Binface's 2029 manifesto!

First-past-the-post has all sorts of problems causing people to vote against their preferred candidates and reinforces the top two across elections. It can also lead to "spoilers" causing a split vote and unpopular candidates winning.

Great to know who was behind Count Binface

I love the names on the example ballet box. I personally support dusty boots.

I think the difference of vote distribution between LibDems and Reform is not only about campaign strategy and "focusing effort" - it's likely also how the potential voters is distributed across the country. While you can focus your campaign spending to specific locations, voters generally won't move to locations where their preferred party has a chance (and can't even do so in the time the election was announced). (And to a lesser degree, your party volunteers are also mostly going to campaign near where they live.)

As a non-Brit, I was shocked to learn that your new PM took office the day after the election. I genuinely did not think that kind of speed was possible.

A cheeky little Sunak sighting in there. Did he just...deflate as the evening progressed, or did he not stay long?

The Reform vs LibDems plot is so good. Shows so clearly how the different strategies were borne out and how they affected the results.

all that Count - count jokes... reminds me at the fourth Doctor saying "You can't count, Count!" Damn, good times.

I have been a "Count Agent" in Australia several times. We call them "Scrutineers" here. Our preferential system means there's a lot more to look at. Sometimes a party's observers want to tally not vote totals but the preference flow rate from parties that will be excluded. While I think your point on "Reform wouldn't necessarily win more seats than Lib Dem under a different system" is true I would make the point that the current UK system is quite strange. It would be better if the system incentivised parties to treat all votes as equally valuable rather than votes in winnable/marginal constituencies being very high value whereas everywhere else is basically only for show. PR could help improve party politics.

Part of me doesn't want to comment because there happens to be 615 comments right now, but oh well. There's something really fitting about Count Binface's counting agent doing vote verification on random scraps of pentagon

Thank you for dedicating your time and effort to the process of democracy!

I live in Australia. We do preferential voting. You number all the boxes with your preferences 1 to 5 or however many candidates are running.

Sorry Count Binface didn't win

You answered pretty much all of the random questions I had while watching the election counting. Always good to see how the process works at the ground level.

Fascinating stuff, Matt! Would love to see you do this again next year for the Australian Federal Election too! Maybe even try get Antony Green on as a guest to talk about the formulas he uses to always call the election first?

I'm so excited that you got a Count Binface rosette to wear the night of. I wasn't expecting it, and it really made me happy in your video.

I think it is quite appropriate for an agent of the count would write on polygon shaped paper. Probabky the standard shape from the planet he is from.

Congratulations! You are the only commentator I've heard making the pointbthat you can only evaluate results in the terms under which the election was conducted.

It's worth saying that as a longer-standing party with a local infrastructure in place, the lib dems were much more equipped to fight elections on a constituency basis than Reform did - Reform did put a lot of effort into the constituencies they had a chance of winning, but because of the way their support is distributed that wasn't very many

So pleased you learnt to pronounce Northallerton properly before you got there! 😀 Fascinating video, as ever, always wondered about the details of how this worked.

Thankyou, for both a peak behind the curtain into (arguably) the most important part of our election process and a measured, sensible response to the proportional representation question

This was the most fun politics related video Ive watched in a long time, thank you so much Matt! All Hail Count Binface~ Its also made me feel way more confident in how our elections work, this does feel like a very good system~ I do wonder why we vote with pencils and not pens, though having count officers for every candidate and a checking of every spoiled ballot seems to do a great job of detering tempering

One of your most interesting videos ever (and they're all interesting). Thanks.

Even if this is a result of campaigns accounting for regional representation instead of proportional representation, this is still strong evidence that regional rep is garbage and should be replaced by proportional rep for future elections.

count binface sounds like it could be a villain for a kids show

Hey, Matt! You clean up pretty good! Great video, thanks!

Matt I got my copy of Love Triangle just last week (and having a signed book from you is awesome and so is the sydler print!) and I must say, you're just a treasure. Just please keep making stuff as long as you can stand it. I love your writing, I love your videos. Keep it up.

The reason that parties like to get people in to observe the count is (as I understand it) less to do with predicting the outcome before it is announced and more to do with obtaining data at a more granular level than that which is released to the public. By having observers tally the number of votes for their candidate from each ballot box they can obtain data which is useful for campaigning at the next election - achieving electoral success is a long game. Unfortunately this makes things even more unfair for new small parties because they find it hard to get enough people to observe all the counting tables so they are generally working with less data than their bigger/richer/more established rivals.

Very lucky prediction, given the small sample size. Presumably (although after the fact) you could calculate a “confidence level” for your “back of the pentagon” estimate - in fact, I’m surprised that you haven’t… surely an ideal opportunity to get normal distribution curves and standard deviations out of the toolbox :-)

Absolutely fantastic video. Thanks, Matt

amazing content as always

Presenting the difference in vote share between libdems and reform as just party strategy ignores that audiences have different patterns all by themselves. While of course we do not have those stats it wouldn't surprise me that large amounts of libdem voters voted for labor just because they themselves know the odds in their district. Reform voters on the other hand might not know about the win chances, not care or a bunch of their votes may have come from people who only voted for them because they considered it a lost seat.

"I'm a professional counter" sounds more expressional than to say "I'm a mathematician".

I've been a counting agent at for a number of years, and I really enjoy it. Glad you had a good time. And thanks for the total count numbers, I couldn't find it yesterday when I wanted to calculate the short money.

Great couple of videos

9:50 - Sunak's double take on seeing Count Binface is one for the archives

In my country we don't have a centralised counting facility per constituency. Each constituency is subdivided into several polling stations (a neighbourhood or, in rural areas, one village or a collection of smaller villages) and the counting is done in each individual polling station. And if the polling station has enough registered voters, it may even be subdivided into two or more polling booths, each with their officials (3 people, plus observers nominated by the candidates). So, in the end, each team of counting officials has a maximum of votes to count that should not surpass 1,000 and, due to absenteeism, is usually under 700. It's really quick, with a few exceptions here and there.

Thanks Matt, helps a lot!

Great content.

“I didn’t have a clipboard and the only paper I had on me was some spare pentagons” classic

17:30 wtf is the sound effect you used for the numbers swooshing on/off the screen? I didn't realize the bass in my headphones was so good, it really sounded like my whole house shook from something massive being moved around downstairs.

Haha I was just thinking about when we'd get the follow-up this morning...

I love the randomness of this channel! I never know what Matt is going to be doing 😁

I recently had the chance to supervise a recount in a local election. My side had won by 37 votes out of over 6000. That's close, but not nearly close enough for the expense of a hand recount to make sense. But the other side had the right to demand it, so I oversaw the process as an observer from my side while the other side had their observers as well. It was interesting watching how the votes were stored, tabulated, tracked, etc. In the end, there was a stack of about 16 votes that accidentally never went through the machine. Those split along the same trend as the rest of that precinct, so my side picked up a few votes. WAY too few to change anything even if they had gone the other way. Only a couple of other votes wete changed. 1 ballot was challenged (we won that, they double voted but wrote a note explaining it and the explanation made it clearly a yes). 1 ballot somehow got lost. Otherwise, only 1 or 2 that had been read blank by the machine were counted by hand. Now if they did a machine recount with a hand count of only the ones read blank by the machine, it would have had the same result at a fraction of the cost. Modern machines are way too accurate for anything more than that to make sense unless you're down to just a handful of votes different in a election of that size. In the end, we increased to 44 votes, meaning we get to fully find the schools. We increased from a large plurality of the electorate (after accounting for blanks) to a small majority. And the town wasted 10s of thousands of dollars.

Unless they're using some really advanced editing and camera tricks, there goes my theory that Matt and Count Binface are the same person.

This is awesome; thanks Matt! Thanks for the video, thanks for participating in the election!

Thank you for your service 🎉

I covered a General Election count and let me tell you, they are bloody long. But necessary. Filmed something on it on my RU-vid channel. It's also amazing to see the distain in counters' faces when one of the candidates disputes how the ballot papers are stacked.

I just wanted it to be a massive stack of paper with the count from sesame street... "1 vote.. ah ha ha ha... 2 Votes.... Ah ha ha ha!.!....

So sad I missed my chance to grab tickets to your show in NY! Praying to Erdös for you to add a second night 🙏

I always wonder what happened to the Monster Raving Looney Party. Or possibly the Black Adder party.

Worth pointing out that the LDs using council election results for their tactical vote messaging goes a bit deeper. Their entire strategy at this election was target constituencies where they already had a strong council presence, as this correlates with robust local party infrastructure. This contrasts with their 2019 strategy, where they chased gains suggested by MRP polling and which ended in disaster.

As much as you can rightly argue the result of the vote is a result of the rules under which it happened in many ways that doesn't negate the problem the current system clearly has - yes the larger parties are going to focus on where they think they really can get first past the post, and it will pay off for them to some extent. But at the same time people don't vote with how many flyer you are bombarded with... Tactical voting is all well and good and clearly a few parties benefited there, but the whole point of democracy is supposed to be representing the will of the people, which when one party has 2/3rd the seats with only 1/3rd of the vote... Also as a new party Reform not focusing anywhere really makes sense for them under the current system - this the market research phase in effect. Looking for all the places where the population actually likes them a bit to put more effort in next time, though with how flat your graph for them was I don't suppose it is going to help them that much.

I saw Matt at the count on the BBC election night coverage (ever so briefly!) 😀

It's good to see that you were down for the Count.

Count Binface for US President!

Given the number of LibDem fliers you got, I predict you have a Gail’s bakery in your hometown.

Oh my gosh I love the scraping

Excellent insights into how the voting system and count works.

Important political context for the lib dem result: 1. The lib dems are not over-represented in parliament. They got 11% of the votes and 12% of the seats, which is the closest vote to seat percentage of any party. This is also the first election ever where they have been correctly represented. Up until now the lib dems have always been under-represented in parliament because as a smaller party they also suffer under first past the post. The parties over-represented in Westminster are Labour and the Conservatives. 2. It isn't possible to know how many votes labour, the lib dems, and the greens, would have got if we had proportional representation. All three parties are to the left of centre, and have overlaps and similarities in their policies. They are different parties in terms of political philosophy, and if you're a member of one of them you care about the differences, but most voters see them as similar. There are some areas of the country where the lib dems historically do much better than labour. People in areas where the lib dems do better than labour often vote lib dem even if they actually want a labour government, because under first past the post the lib dems have a better chance of winning the seat. Similarly in most of the country, labour have a much better chance of winning than the lib dems, so people who prefer the lib dems still vote labour because labour is close enough to what they want and has a much better chance of winning. However, as you can see from Matt's data, some people still vote for the party they actually want even when that party is very unlikely to win. The Greens have very little chance of winning in most seats, and yet people continue to vote for them, even under first past the post, because they believe in the Greens policies, or because they want to protest about labour not being far enough left. If we had proportional representation, people would be free to vote for the party they actually want most. The results could be very different. The lib dems might get more votes if we actually had PR, or they might get fewer votes. We don't know.

With such an amazing percentage increase in votes compared to other parties, I would have thought that even though Count Binface didn’t get a seat, he should have been given a pouffe.

YES!

Hi Matt. In North Herefordshire we had a strong tactical vote for the Green party, this ousting the Tory. So unsurprising that they got this one wrong.