iiinova — where your digital self keeps evolving

Joe Moed @mojo87.bsky.social · 06/10/2026

PQ wipes CAQ away, and flushes them out of AssNat. A joke 8 years in the making.

Screenshot of CBC election night coverage in Quebec, showing PQ (Parti Quebecois) in the ascendancy, and the governing CAQ (Coalition Avenir Quebec) with ZERO seats.

110

Joe Moed @mojo87.bsky.social · 02/10/2026

Must belong to this fellow.

A still from the 1989 British classic "The BFG", showcasing the BFG's enormous ears.

010

Joe Moed @mojo87.bsky.social · 30/09/2026

I admit I did not expect to see this particular Talmudic deep cut in our national press, but let me tell you, MY community can think of a few Jews this description applies to better than Avi Lewis. Extremely dangerous and incendiary nonsense…

Jewish law developed two character sketches to describe such scenarios. The first is known as a “mosar” — a Jewish person who informs to oppressive authorities, potentially leading to the imprisonment, torture or death of community members. The second, called a “rodef,” refers to a Jew who is actively pursuing another with the intent to harm or kill them. In both scenarios, the law is clear that a Jew is duty-bound to step in. Protecting the community and its members is the most important thing.

012

Joe Moed @mojo87.bsky.social · 29/09/2026

At such extreme values its hard to make reasonable projections, but @338canada.bsky.social reckons 270!

Output of 338Canada's simulator on a hypothetical election result of LPC 50, CPC 28, NDP 11, BQ 6, GPC 4. It results in seat counts of LPC 270, CPC 46, BQ 15, NDP 11, GPC 1, including near-total wipeout for CPC east of Manitoba.

331

Joe Moed @mojo87.bsky.social · 01/06/2026

"Why don't you go and HAVE A LOOK?"

000

Joe Moed @mojo87.bsky.social · 10/05/2026

Here's the latest GB picture - admittedly, LAB still have most councillors with ~25%, followed by CON 21%, LD 18% & REF 13%, but that's what happens when only 5k of the ~19k GB council seats are up for election. Most of these seats haven't yet been contested since REF become a factor.

221

Joe Moed @mojo87.bsky.social · 10/05/2026

Caution: these appear to be the proportions BEFORE Thursday's elections.

UK councillor standings as of 5th March 2026: LAB 5,873 (31%), CON 4,205 (22%), LD 3,213 (17%), REF 963 (5%), GRN 906 (5%)

110

Joe Moed @mojo87.bsky.social · 29/04/2026

While I sympathise w/ the sentiment (and FPTP is a dogshit electoral system), a result like this (i.e.: basically OLP +7 & PC -7 from 2025, would be around the tipping point where FPTP starts to work fine for OLP again - they'd start narrowly winning lots of GTA seats they narrowly lost last time.

Projected results from a recent Ontario poll:

Votes: PC 36%, OLP 38%, NDP 20%, GPO 4%
Seats: PCs 45, OLP 47, NDP 28, GPO 3

100

Joe Moed @mojo87.bsky.social · 28/04/2026

Something like that yeah - well done if that estimate was on gut instinct :p

Based on the Abacus poll, the 338Canada Ontario Simulator predicts 52 seats for PCs, 41 for OLP, 28 for NDP and 2 for GPO.

110

Joe Moed @mojo87.bsky.social · 02/02/2026

Checked out Chessdotcom’s new monthly themed bots for February, and on behalf of the Curling community, I am OUTRAGED. Our sport is literally known as Chess On Ice. Why is the curler a 350-ELO doofus, while a hockey goalie gets to be a Magnus Carlsen level genius?? CURLING ERASURE.

Screenshot of the February monthly bots on chess.com - the lowest rated bot is a 350 ELO curler from Canada.

Screenshot of the February monthly bots on chess.com - the highest rated bot is a 2850 ELO hockey goalie from Sweden.

030

Joe Moed @mojo87.bsky.social · 10/12/2025

That is one POWERFUL spring.

CBC News headline reading “Finch West LRT to be 10 minutes faster by spring, TTC says”

010

Joe Moed @mojo87.bsky.social · 02/05/2025

Feast your eyes on this FPTP monstrosity. Mind-boggling that the UK used to have Supplementary Vote (kinda like Ranked Ballot) for such elections & *actively* got rid of it! Nice of the parties to arrange a neat stepped pattern. Though I see Ian Scott not holding up his end of the bargain. Prick.

Results of the 2025 West of England Mayoral Election - LAB 25.0%, REF 22.1%, GRN 20.0%, CON 16.6%, LIB 14.0%.

030

Joe Moed @mojo87.bsky.social · 29/04/2025

I don't think they've called it yet (the dark blue colour means leading, right)?

000

Joe Moed @mojo87.bsky.social · 20/04/2025

Not at present, I believe (although still a lead in seat count, and in prior updates, Mainstreet have been showing LPC majority even when CPC had 2% lead…). The point still stands that LPC structural advantage is significant! As well as this just being one poll of course :)

Current Mainstreet projection of 162 LPC seats based on 39.1% vote and 145 CPC seats based on 43.2%.

020

Joe Moed @mojo87.bsky.social · 16/04/2025

Indeed - I believe they don't publish regionals, unfortunately, but it's true their model shows a strong Liberal majority on a 2.7% lead (in fact, as of today, the exact same majority as 338 aggregate shows on a 6% lead!). www.mainstreetresearch.ca/dashboard/ca...

338 Canada projection as of 16-Apr-25 - they project a Liberal majority of 43 seats, based on a 6% popular vote lead.

Mainstreet projection as of 16-Apr-25 - they project a Liberal majority of 43 seats, based on a 2.7% popular vote lead.

200

Joe Moed @mojo87.bsky.social · 14/04/2025

A tragedy in 2 acts over on chess dot com. THIS is why they say "Never Resign", especially against pissy opponents so concerned about losing "a minute of their life" while playing *A 3-MIN A SIDE CHESS GAME*... And, dare I say it, especially against entitled USians who bring Canada into it 🤣

Chess opponent was pissed at me for not resigning.

Press the resign button
1 minute of my life gone
because someone in Canada... is tripping
bro you really think you getting a stalemate
seriously

Oops turns out I could salvage a stalemate after all...

GAME OVER
Game drawn by stalmate (3 | 2 Rated)

030

Joe Moed @mojo87.bsky.social · 11/04/2025

They were a bit further off in 2015 (e.g.: see below from Eric Grenier's ThreeHundredEight before he joined CBC!) But that's largely down to the 3pt polling miss for LPC - that is to say, arguably polls were to blame, not the models :p

Final projection of 2015 Canadian Federal Election by Eric Grenier - LPC 146, CPC 118, NDP 66 BQ 7, GPC 1

010

Joe Moed @mojo87.bsky.social · 11/04/2025

Pretty good - here were CBC Poll Tracker's final predictions in the last two elections (other models would probably have similar results): 2021 projection: L 155, C 119, N 32, B 31 2021 result: L 160, C 119, N 24, B32 2019 projection: L 141, C 121, B 39, N34 2019 result: L 157, C121, B 32, N 24

Final CBC Poll Tracker projection of the 2021 Federal Election

Final results of the 2021 Federal Election

Final CBC Poll Tracker projection of the 2019 Federal Election

Final results of the 2019 Federal Election

130

Joe Moed @mojo87.bsky.social · 08/04/2025

The Innovative poll is on the CBC Poll Tracker, you just have to expand the list a bit.

List of recent polls in CBC poll tracker

110

Joe Moed @mojo87.bsky.social · 05/04/2025

There isn’t a specific breakdown by age and gender, but given this is the split for 18-34s, if you overlay the gender split on that, I would imagine it is more acute among young men, yes…

Breakdown of 18-34 voting intentions from the latest Nanos poll: CPC 42, LPC 32, NDP 13, GPC 6, PPC 3, BQ 3

160

Joe Moed @mojo87.bsky.social · 04/04/2025

Can’t decide if this is better or worse! Dadaist Google Translate…

110

Joe Moed @mojo87.bsky.social · 04/04/2025

Not saying I love it, but worth bearing in mind that, depending on regional splits, even a 39-39 tie could be very good for the Liberals. Here's a projection based on Abacus' regionals - basically CPC only tied due to a 36% lead in AB, which masks the fact they trail by 5% in the rest of Canada!

Regional breakdowns of the tied 39-39 Abacus poll - CPC leads by huge margins in MB/SK/AB, but conversely, the LPC vote efficiently converts to seats in QC, ON and Atlantic.

Possible projected results, based on the regional breakdowns of the tied 39-39 Abacus poll - LPC wins a large majority of seats in Atlantic, QC & ON, resulting in a small overall majority despite tied headline numbers.

240

Joe Moed @mojo87.bsky.social · 01/04/2025

Not an in-depth study or anything, but this chart from Wikipedia gives a good overview! At a high level: - Polls were fairly stable (no party's average was ever more than 4% from their final result) - Final poll average was quite close to actual result (+/-2% in most cases)

Chart of the polling averages in the run up to the 2021 Canadian Federal Election (key points in post)

120

Joe Moed @mojo87.bsky.social · 31/03/2025

Lower than their recent highs of 6.78% & 6.55% in 2008/2019 respectively, but not materially different than 2021, where they managed 2.33%.

Chart of Green Party of Canada votes & seats in each Canadian Federal Election since 1984 (data relevant to the question at hand is in post)

110

Joe Moed @mojo87.bsky.social · 30/03/2025

I've already shared the data showing pollsters (and projections models based on them) did not materially get it wrong in 2019. And if we're going down the "they got Brexit wrong" route, I don't think this is going to be a productive conversation.

100

Joe Moed @mojo87.bsky.social · 30/03/2025

That article is from two weeks before the election though - LAB declined 2-3% in final weeks of campaign and CON recovered slightly by 1-2%! Point still stands that the FINAL projections, the only ones anyone's hanging their hat on, got quite close - one or two outliers, mostly CONs 100ish.

200

Joe Moed @mojo87.bsky.social · 30/03/2025

This is directly from what I originally shared. Nine major models were active during the election, and those numbers I gave are the median of these projections. Same thing for 2019 election. Although I miscalculated slightly, the median projection is: CON 343, LAB 223, LD 16, SNP 43

Final projections from nine major projection models ahead of the 2024 UK General Election.

Final predictions of major projection models for the 2019 UK General Election.

100

Joe Moed @mojo87.bsky.social · 30/03/2025

Sorry, but that's simply not true, at least recently in Canada: 2021 final projection: L 155, C 119, N 32, B 31 2021 final result: L 160, C 119, N 24, B32 2019 final projection: L 141, C 121, B 39, N34 2019 final result: L 157, C121, B 32, N 24

100

Joe Moed @mojo87.bsky.social · 30/03/2025

Sure they’re flawed, and by necessity have to make assumptions about turnout & demographics. But the proof is in the pudding - the point I’m making is that, *despite* being imperfect, in Canada at least, recent polls managed to do very well!

2011 election polling and final results - average poll error is approx 2% for the three major parties.

2015 election polling and final results - average poll error is approx 1.5% for the three major parties.

2019 election polling and final results - average poll error is approx 2% for the three major parties.

2021 election polling and final results - average poll error is approx 2% for the three major parties.

110

Joe Moed @mojo87.bsky.social · 30/03/2025

Indeed - FPTP is wack, and there’s no better example of its wackness than the UK 2024 General Election! Which makes it tricky to accurately model results as a side effect, and can make small polling errors translate to large seat projection errors.

Chart of % votes vs % seats in the U.K. 2024 general election, illustrating the distortions of FPTP.

100

Joe Moed @mojo87.bsky.social · 30/03/2025

But it might also be due to shifts in how LPC vote is distributed, and who they've gained from? For example, in ON, latest projection has L 47%, C 38%, N 9%, giving 81/39/1 seats. But in last election, it was 39/35/18%, giving 78/37/5 seats.

110

Joe Moed @mojo87.bsky.social · 30/03/2025

Needless to say, a polling miss involving overestimating LPC by 4.3% and underestimaating CPC by 2.4% could have a FAR larger impact on the makeup of Parliament in this election, where things are much closer to start with!

338Canada simulator based on LPC at 43% and CPC at 36% - they win 188 and 133 seats respectively, a firm majority for LPC.

The same 338Canada simulator assuming LPC instead get 38.7% and CPC get 38.4% - this time, CPC get 169 seats, just short of a majority, to LPC's 149.

100

Joe Moed @mojo87.bsky.social · 30/03/2025

Having said that, the actual national poll numbers were off by a fair amount: LAB averaged 39ish, they ended up on 34.7% (GB only); CONs averaged 22ish, they ended up on 24.4%. But LAB were so far ahead and vote so efficient that the modellers were able to do quite well anyway!

Chart of UK polling leading up to the 2024 UK General Election

130

Joe Moed @mojo87.bsky.social · 30/03/2025

Wait what, which polls predicted Tories would be reduced to 20 seats? These were the final calls from the various reputable groups making projections - they ranged from 64-130 seats, with an average of 102, so the Tories were on the high end, but well within the range...

Final projections of the 2024 UK General Election. The Tory predicted seat count ranged from 64 to 130, and averaged 102.

240

Joe Moed @mojo87.bsky.social · 27/03/2025

I dunno, it doesn't seem that bad! Here's the latest from each pollster per @338canada.bsky.social: Mainstreet: L +4 Research Co.: L+4 Liaison: L+4 Abacus: L+1 EKOS: L+15 Angus Reid: L+8 Leger: L+6 Pallas: L+4 OK, EKOS is doing EKOS things, but the rest all have Ls between 38-46% and Cs 36-41%.

List of all recent polls for the Federal Canadian Election from the last week

210

Joe Moed @mojo87.bsky.social · 27/03/2025

SK is a bit harder, as most polls lump it in with MB, but you can ballpark non-AB numbers using regional breakdowns, if you really want. AB is (roughly) 1/8th of Canada, so take the 42% lead there & divide by 8 - it impacts national margin by approx. 5%. Excluding AB, it's roughly LIB 41%, CON 34%.

Regional breakdown of the latest Abacus poll for the Federal Canadian Election. In Alberta, the Conservative Party is at 61% and the Liberals are at 19%.

100

Joe Moed @mojo87.bsky.social · 12/03/2025

PHEW. You had me going for a few questions there! #YiddishPride

Another question from the Duolingo Yiddish course, asking to write a Yiddish sentence in English. This time, the correct English answer reads "Yiddish is not a dialect of German".

010

Joe Moed @mojo87.bsky.social · 12/03/2025

Hey look, a self-hating Duolingo!

A question from the Duolingo Yiddish course, asking to write a Yiddish sentence in English. The correct English answer reads "Is this German or Yiddish?".

130

Joe Moed @mojo87.bsky.social · 12/02/2025

4. There’s also the other extreme, where they try to make a SV recommendation in ridings such as Lambton / Kent / Middlesex. I’m sorry, but saying anything other than "It doesn’t really matter; soz, OLP/NDP voters, pick who you want 🤷” is laughable.

Votewell.ca recommendation for the riding of Lambton / Kent / Middlesex, showing PCs on 66%, OLP on 15% and NDP on 13%, with a recommendation to strategically vote for OLP.

Votewell.ca recommendation for the riding of Lambton / Kent / Middlesex, showing PCs on 65%, OLP on 15% and NDP on 13%, with a recommendation to strategically vote for NDP.

110

Joe Moed @mojo87.bsky.social · 12/02/2025

3. For example: trying to pick the right SV in Humber River / Black Creek, based solely on this projection, is essentially impossible, and could backfire! There may very well be a theoretical “right” SV in this riding – but a more honest take here would be “insufficient data for meaningful answer”.

Votewell.ca recommendation for the riding of Humber River / Black Creek, showing PCs on 36%, OLP on 33% and NDP on 27%, with a recommendation to strategically vote for OLP.

Smartvoting.ca recommendation for the riding of Humber River / Black Creek, showing PCs on 36%, OLP on 33% and NDP on 27%, with a recommendation to strategically vote for OLP.

110

Joe Moed @mojo87.bsky.social · 17/12/2024

My French may be a bit rusty, but I’m *pretty* sure that’s not how you say “smooth”…

Samsung Galaxy Tab S6 Lite packaging - the French translation of “precise & smooth” is given as “précis et smoot”.

240

Joe Moed @mojo87.bsky.social · 15/12/2024

Ha good Q! I don't "automatically" do it, but my brain does work a bit like that - was looking at "List of French PMs", to read up on recent officeholders, noticed 99/240 = 5/12ths ratio, then thought "huh, isn't 1 year 238 days about 600 days?" before doing more precise calcs for joke purposes :)

List of the last four French Prime Ministers on Wikipedia, showing their time in office:

Elisabeth Borne = 1 year, 238 days
Gabriel Attal = 240 days
Michel Barnier = 99 days
Francois Bayrou = 2 days

110

Joe Moed @mojo87.bsky.social · 13/12/2024

Élisabeth Borne: 16-May-22 to 9-Jan-24 = 603 days Gabriel Attal: 9-Jan-24 to 5-Sep-24 = 240 days Michel Barnier: 5-Sep-24 to 13-Dec-24 = 99 days 99/240=0.41; 240/603=0.4. Hmmm... Is Macron making a geometric progression of PMs with a ratio of 0.4? 🤔 Ready for the PM Singularity on 16-Feb-25?

Excel calculation showing that the time in office of the last three French Prime Ministers is following a geometric progression, and calculating that at this rate, there will be infinitely many PMs appointed on 16-Feb-25.

210

Joe Moed @mojo87.bsky.social · 21/11/2024

Chess? Completed it, mate.

A picture of a smothered mate with the knight, delivered after sacrificing my Queen, a dream checkmate.

140

Joe Moed @mojo87.bsky.social · 05/11/2024

Impressive, and well done that guy indeed, as it could quite plausibly actually matter! 😟

000

Joe Moed @mojo87.bsky.social · 02/11/2024

Done! I may have gone a bit overboard, but I guess it can't hurt - it probably can't help either, but I'm trying to be less defeatist about that sort of thing these days :)

In a recent Toronto Star Op-Ed, the Minister of Transportation contends that "what we need is a common sense, evidence-based bike lane policy".
I agree with the Minister; fortunately, that's what we already have right now.
There is ample evidence, both from Toronto itself and from around the world, that the complaints regularly levelled against bike lanes are overblown (at best), or completely bogus (at worst).
The linked fact-checks from CBC, and the linked study, cover all the main points, but in summary:
(a) Bike lanes do not cause gridlock - in fact, a safe, well-connected and well-maintained bike lane network can actually reduce it, by encouraging people out of their cars
(b) In general, building more lanes *anywhere* does not reduce gridlock, due to induced demand - more lanes encourages more cars, which fill the available space
(c) Bike lanes do not negatively impact emergency services, and the first responders themselves are bemused by these repeated claims
(d) Bike lanes are well-used - cyclists make up around 30% of vehicles entering/leaving Downtown along Richmond & Adelaide, but it doesn't always look that way as they're so much more space-efficient.
(e) Bike lanes are a boon to business - far from harming the local economy, businesses along bike lanes tend to report more customers, staying longer, and spending more.
This legislation is ill-thought-out, ideologically driven, an appalling overreach of Provincial authority, and a massive step backwards for our City. I oppose it thoroughly.

010

Joe Moed @mojo87.bsky.social · 26/10/2024

Thanks for the clarification - I knew your friend had led an interesting life, but this seemed a bit far-fetched!

020