Jon Bromstein

Analyst, Dad, Utility Infielder.

[Jon Bromstein @JonBromstein]

@311Toronto the road at the intersection of Geoffrey Street and Parkside Drive is disintegrating. The damage here is actually deeper than the photo makes it look.

[Jon Bromstein @JonBromstein]

@Cpt_RDubya @oolsen50 @CanadianPolling The wording is a bit confusing. This is the size of their seat LEAD. Lots of polls showing them over 230 seats now.

[Jon Bromstein @JonBromstein]

@CraigKovatch @CanadianPolling @338Canada This is getting into territory that's very hard to model. Best guess 275 CPC seats. Bloc maybe 40. Liberals and NDP both under 15.

Important to note this is the best poll for the CPC so far, so it's a bit cherry-picked, but it'd be an absolute rout.

[Jon Bromstein @JonBromstein]

@aintnosatchmo @acoyne The polls are accurate in terms of predicting the popular vote...where the system makes things fun is projecting how that vote will translate into seats.

This is where polling ends and modelling begins, and it's a whole other challenge (albeit a fun one).

[Jon Bromstein @JonBromstein]

@aintnosatchmo @acoyne It's counterintuitive, but it's often easier to get an accurate read on a large population. Sample size and the ability to get a representative sample are what matter.

National polls over the past couple of elections have been extremely accurate, which is not by accident.

[Jon Bromstein @JonBromstein]

@aintnosatchmo @acoyne Some do. A lot have moved to panels...a large group of people that look like a scaled-down Canada. They then poll random samples from that group.

This is actually easier to do with a large population than a small one, which is partly why national polls are often more accurate.

[Jon Bromstein @JonBromstein]

@UPexpress Why are there ever 2 car trains at rush hour though? Every train is overcrowded.

[Jon Bromstein @JonBromstein]

@Metrolinx @UPexpress

When will UP return to running exclusively 3 car trains? 2 car trains are causing dangerous overcrowding and increasingly preventing large numbers of people from boarding at Bloor.

[Jon Bromstein @JonBromstein]

@AustonMatt51041 @KPapetti You could spend Bertuzzi's money on someone like Chandler Stephenson in free agency. But that would mean little cap space to invest in defense.

[Jon Bromstein @JonBromstein]

@Wilnerness This move makes me sad (Pearson, not BlueSky). Hopefully Manoah figures things out, Rodriguez goes to the pen, and Pearson comes back up instead of Francis.

[Jon Bromstein @JonBromstein]

@cselley So this is being communicated poorly, but apparently GO will be moving to 30 minute service and those departures will be offset from UP so that 15 minute service to Union is roughly maintained with increased capacity.

It will be confusing for people though.

[Jon Bromstein @JonBromstein]

@GurkiratGill04 @donnelly_b @bilal_akh @UPexpress @Metrolinx @RM_Transit Either way the GO trains need to be 10 cars at least to deal with all the e-bikes. If they go with 6, this entire change will create an absolute disaster at Bloor.

[Jon Bromstein @JonBromstein]

@GurkiratGill04 @donnelly_b @bilal_akh @UPexpress @Metrolinx @RM_Transit Yes. But if you take away 2 UPs and replace it with a single 6-car GO, the benefits are likely to be negligible given how much busier the GO trains are by the time they get to Bloor due to all the traffic to/from Brampton/Baramalea.

[Jon Bromstein @JonBromstein]

@donnelly_b @bilal_akh @UPexpress @Metrolinx @RM_Transit Pretty much. Goes from 4 UP and 1 GO per hour to 2 and 2. If the GO trains are 10 or 12 cars, it's definitely more capacity. If they're 6 cars, not so much.

[Jon Bromstein @JonBromstein]

@donnelly_b @bilal_akh @UPexpress @Metrolinx @RM_Transit The plan appears to be alternating GO and UP every 15 minutes. Eastbound, GO on 14/44 and UP on 29/59.

This is more capacity than it has now. The downside is the inevitable confusion and the 50% cut to airport service.

[Jon Bromstein @JonBromstein]

@hblakingfan @MoneyPuckdotcom It doesn't. West is favoured 52-48.

[Jon Bromstein @JonBromstein]

@JoeBenjamin_ @MoneyPuckdotcom Correct, it's weighted in favour of more recent performance.

[Jon Bromstein @JonBromstein]

@SamuelAdamel @MoneyPuckdotcom Could be a few things

-They won but played badly (analytically)
-Their likely opponents had a great night
-They won but it increased their odds of playing a hotter/analytically stronger team

Etc

[Jon Bromstein @JonBromstein]

@harthebar13 @MoneyPuckdotcom Factor in a player's long-term spread between Goals and xGoals. If a player scores more than he should over a big enough sample, it starts demonstrating talent, not luck.

Take Pastrnak: Career xGoals 238, actual goals 321.

Over 10 seasons, that's just proven skill.