WCQ: Third Round - Window 4 (January 27- February 2, 2022)

[Califax]

5 minutes ago, CanadaFan123 said:

You realize that we've been letting guys slide in all qualifying right?

Johnston, Buchanan, Eustaquio, Kennedy, Crepeau... 

Before 2021 those guys were either uncapped, or had not played significant minutes for Canada in the case of Stephen and Max. 

Some of those guys we were all pumped to get on the field. and to your point I think Ugbo probably fits in with a lot of these guys as someone we will be happy to give a run , maybe because of our situation, maybe irrespective of it. For me, ideal circumstances, it would probably only be if we take points in games 1 and 2, not cause I think he's coming it better. But that's certainly subjective.  

A lot of other guys we will be forced to play, and though it is what it is,  it's hard to force myself to be positive about that. 

[Obinna]

Just now, QBCS said:

Confusing, But wheeler said it would drop around 3? Am I being stupid or is there conflicting answers.

No conflicting answers. The roster will be at 3pm eastern, barring any last minute hold ups. They probably won't know the roster before they film today's show. 

[DeRo_Is_King]

18 minutes ago, Approve My Account Pls said:

I don't understand why this has turned into a Millar vs. Corbeanu debate. Both kids have been having fantastic seasons, both kids deserve the call up. Millar is slightly ahead of Corbeanu imo, I'd be much more comfortable starting him for a full 90, but Theo also brings a spark off the bench if we needed it. 

It isn't. I am very supportive of Millar. He is a key piece for Canada. 

[RS]

1 minute ago, Obinna said:

No conflicting answers. The roster will be at 3pm eastern, barring any last minute hold ups. They probably won't know the roster before they film today's show. 

They will almost certainly be given an embargoed copy of the roster at least an hour before 3pm, if not earlier. Not sure why they couldn't just delay filming the show until they receive that info.

[QBCS]

2 minutes ago, Obinna said:

No conflicting answers. The roster will be at 3pm eastern, barring any last minute hold ups. They probably won't know the roster before they film today's show. 

Thanks for the info

[Xavier.]

30 minutes ago, DeRo_Is_King said:

Corbeanu has more goals in League One (3) in 16 appearances (many of those apps were as a sub for Wednesday). Millar finished the campaign with two goals in 27 apps, many of which he started. 

I totally agree that Millar is playing well, but there is no denying Corbeanu's pedigree. He is not a typical 19-year-old player and can be injected now. He has proven it at both Wednesday and MK. 

What is this argument, this is Millar's breakout season why are you using his last year stats. This isn't an argument on who will be better in the future, its an argument on who can help the team right now. Millar has been better on a better team in a better league.

[DeRo_Is_King]

8 minutes ago, Xavier. said:

What is this argument, this is Millar's breakout season why are you using his last year stats. This isn't an argument on who will be better in the future, its an argument on who can help the team right now. Millar has been better on a better team in a better league.

I'm not saying Millar shouldn't start. I did not bring him up initially. See my other posts. 

[sstackho]

All right all right all right.  It's nervous energy time.  Time to fire up that old Elo-based simulation model from el_noido on Reddit.

Here's where we last left off.

Please note that this model does not give a damn about the loss of Davies and Eustachio.  It simulates results based on current Elo ratings, which is backward-looking.  But it does help to get a sense for how other matches will impact our chances. 

Below is the approximate impact of the results of Matchday 9 on Canada's chance of finishing in the Top 3. 

For example, if Costa Rica beat Panama, then Canada's chances of finishing Top 3 would increase from 88.1% to 91.1%. 

 

 

Edited January 24, 2022 by sstackho

[TOCanada115]

48 minutes ago, gator said:

Isn't Cornelius more of a left sided player and Johnston right sided?

This. Cornelius doesn’t play on the right side. If he did, would be an interesting conversation. I agree with Califax though. I think the difference between Kennedy to Miller is closer than Johnston to Cornelius. 
 

Id start Kennedy Vitoria Johnston against Honduras if we run a back 3. 
 

I think everyones most memorable Kennedy match during this campaign was the giveaway for the US goal, but he has so much quality when I watch Jahn. Don’t forget he made a huge sliding challenge on Quioto in stoppage time in the 1st match against Honduras. 

[Kadenge]

1 hour ago, TOCanada115 said:

I might be in the minority here, but I’d start an in form Scott Kennedy, over Kamal Millar who hasn’t played in over 2 months. 

This. The one player on Honduras that is in season and red hot is Albert Ellis at Bordeaux and he has wheels. Davies would have helped to keep him in check on our left flank. Maybe Liam Millar can do the job on left wing/mid. I'm not sure Kamal Miller can keep up with him, but Kennedy can. Plus Scott is  good on set pieces. The only concern is he has not played for us in a while.

[sstackho]

7 minutes ago, sstackho said:

Below is the approximate impact of the results of Matchday 9 on Canada's chance of finishing in the Top 3. 

For example, if Costa Rica beat Panama, then Canada's chances of finishing Top 3 would increase from 88.1% to 91.1%.

It's interesting that we've reached a turning point.  The odds now slightly improve if Mexico and the US lose.  Whereas in the past, we wanted them to run over everyone else. 

I'm not sure how comfortable I am with that hypothesis. 

[Obinna]

9 minutes ago, TOCanada115 said:

This. Cornelius doesn’t play on the right side. If he did, would be an interesting conversation. I agree with Califax though. I think the difference between Kennedy to Miller is closer than Johnston to Cornelius. 
 

Id start Kennedy Vitoria Johnston against Honduras if we run a back 3. 
 

I think everyones most memorable Kennedy match during this campaign was the giveaway for the US goal, but he has so much quality when I watch Jahn. Don’t forget he made a huge sliding challenge on Quioto in stoppage time in the 1st match against Honduras. 

Moreover, I think you have to play your guys in-season. Kennedy is back and playing (and scoring) for Jahn. 

The fact he played against Honduras already is a nice bonus. 

[frmr]

3 minutes ago, sstackho said:

It's interesting that we've reached a turning point.  The odds now slightly improve if Mexico and the US lose.  Whereas in the past, we wanted them to run over everyone else. 

I'm not sure how comfortable I am with that hypothesis. 

That's what happens when you're King of the Castle!

[CanadaFan123]

2 minutes ago, sstackho said:

It's interesting that we've reached a turning point.  The odds now slightly improve if Mexico and the US lose.  Whereas in the past, we wanted them to run over everyone else. 

I'm not sure how comfortable I am with that hypothesis. 

I understand it's formula based but it's flawed.

If the USA and Mexico both win Thursday, we draw and Panama loses our odds would go down by 0.4% to make the top 3.. yet we would have gained a point on Panama with 5 matches remaining. 

[sstackho]

3 minutes ago, CanadaFan123 said:

I understand it's formula based but it's flawed.

If the USA and Mexico both win Thursday, we draw and Panama loses our odds would go down by 0.4% to make the top 3.. yet we would have gained a point on Panama with 5 matches remaining. 

Mathematically, you shouldn't just add up the individual match impacts to get to the cumulative impact - though it will usually get you close. 

I ran this particular set of scenarios, and Canada's odds of Top 3 would go up by about 1%. 

Edited January 24, 2022 by sstackho

[Vasi]

 

https://mobile.twitter.com/PaulTenorio/status/1485662071521959937

And it begins!

 

Edited January 24, 2022 by Vasi
Not sure how to post a tweet with picture :(

[king1010]

embedding Vasi's post

 

 

Edited January 24, 2022 by king1010

[El Hombre]

9 minutes ago, Vasi said:

 

https://mobile.twitter.com/PaulTenorio/status/1485662071521959937

And it begins!

 

Not sure of the veracity of that.  There are a bunch of replies saying that the Panamanian soccer federation president said that there are no new cases earlier today.

[theaub]

The math is fine on the ELO model (obviously) - the bigger issue to me is that the ELO isn't even close to lining up with the betting odds.

[Free kick]

1 hour ago, jonovision said:

I hate to be that guy, but in a country with 4.5 time zones, you really ought to indicate that it is 3 pm ET, if you work for a national broadcaster.

Where have you gone Brian Williams? 😀

Edited January 24, 2022 by Free kick

[Shway]

I gotta feeling that this will be Millars best camp yet. There will be his first goal and assist!

Lets goooooo!!

[sstackho]

11 minutes ago, theaub said:

The math is fine on the ELO model (obviously) - the bigger issue to me is that the ELO isn't even close to lining up with the betting odds.

Betting odds should be much more realistic than Elo-based.  

I took a quick look for betting odds on World Cup qualification, but didn't immediately find any.  Can you please point me to these odds if publicly available online?

[maplebanana]

31 minutes ago, sstackho said:

Mathematically, you shouldn't just add up the individual match impacts to get to the cumulative impact - though it will usually get you close. 

I ran this particular set of scenarios, and Canada's odds of Top 3 would go up by about 1%. 

Yeah, I think based on this model, you can treat each match more like a perturbation, and simulate based on what will happen all else being equal. But if you take multiple matches together then the non-linearity is broken. These numbers would be like the diagonal of a tensor matrix, but the off-diagonal is certainly non-zero. Disclaimer: nearly 20 years since I took my last undergrad math class.

[sstackho]

5 minutes ago, maplebanana said:

Yeah, I think based on this model, you can treat each match more like a perturbation

 

Quote

per·tur·ba·tion

noun: perturbation

1. anxiety; mental uneasiness.

Seems about right. 

Quote

2.  a deviation of a system, moving object, or process from its regular or normal state or path, caused by an outside influence.

Ah. 

Edited January 24, 2022 by sstackho

[theaub]

6 minutes ago, sstackho said:

Betting odds should be much more realistic than Elo-based.  

I took a quick look for betting odds on World Cup qualification, but didn't immediately find any.  Can you please point me to these odds if publicly available online?

Tough to find those...for the Honduras/Canada game for example though the ELO puts the win expectancy at 25/75 Canada, whereas the betting odds are around 35/65.  

Where it then gets more difficult is how much of that variance is injuries etc and how much of it is ELO ignoring our historical issues in Central America?  Because maybe the ELO is just off for the first match (and likely the US one which puts gives us a 54 win expectancy) and then will be OK.