Liverpool v Real Madrid

[Mr Pink]

It's not it's a 1 in 10 shot

A 1 in 10 shot is a massive task, and quite unlikely…

All I said was it’s closer to 1 in 10, than to 1 in 100… and certainly not 1 in 1000 as someone suggested.

Out of interest, what odds do you think it is??

[Mr Pink]

A 1 in 10 shot is a massive task, and quite unlikely…

All I said was it’s closer to 1 in 10, than to 1 in 100… and certainly not 1 in 1000 as someone suggested.

Out of interest, what odds do you think it is??

Sorry just realized you are the one who suggested 1 in 1000!

You should open a bookies mate!

[Doc_Piptorious]

I read that only 8 times in the history of the European Cup and CL has a side lost the first leg at home and gone through, which is a remarkable stat.

Only once in the CL era has a side gone into an away leg with a 2 goal deficit and gone through, no one has ever done it 3 goals down. It's a massive ask

Those are the sort of stats that should have competive people chmping at the bit.


We are good at those never been done before scenarios.


What have we to lose by really having a crack at them? Nothing as we are expected to go out anyway.

Go at them, rattle them and then see if something gives. If they come back at us as hard as they did at Anfield then fair dues to them, but if they wobble then go for the knock out.

[Rich]

Sorry just realized you are the one who suggested 1 in 1000!

You should open a bookies mate!

I'm not interested in gambling and what I'm suggesting is that if the game is played next week 1000 times we don't get the result we need in 999 of them.

And given the history of Real it has yet to happen I'd say that stacks up.

[red g]

I'm not interested in gambling and what I'm suggesting is that if the game is played next week 1000 times we don't get the result we need in 999 of them.

And given in the history of Real it has yet to happen I'd say that stacks up.

How many times have they been 3 goals up on an away le, probably not many? and if so the quality of team would probably be pretty poor.

Look the likelyhood is we won't but I think its a lot greater than a 1000/1 shot.

if we get an early goal anyhting can happen

[Exiled_red]

Those are the sort of stats that should have competive people chmping at the bit.


We are good at those never been done before scenarios.


What have we to lose by really having a crack at them? Nothing as we are expected to go out anyway.

Go at them, rattle them and then see if something gives. If they come back at us as hard as they did at Anfield then fair dues to them, but if they wobble then go for the knock out.

I'm not saying that we shouldn't have a go, just saying that it's incredibly difficult and would be one of the greatest achievements in the history of European football, we can hope, but we shouldn't be expecting. An early goal and anything can happen

[Mr Pink]

I'm not interested in gambling and what I'm suggesting is that if the game is played next week 1000 times we don't get the result we need in 999 of them.

And given the history of Real it has yet to happen I'd say that stacks up.

It’s definitely a very long shot- but definitely not 1 in 1000. Real Madrid hasn’t even played 1000 European games, in their history!

In any case there’s no point looking at their history- man city have never won the champions league in their limited history- it doesn’t mean that they have a 1 in 1000 chance of winning it this year.

I stand by what I said- I reckon somewhere between 1 in 10 and 1 in 20.

[Mark79]

FiveThrityEight, who use actual analytics and maths for their probabilities, give us a 3 in 100 chance of getting the result we need to progress. Or, 1 in 33.

Or 30 in 1000

https://projects.fivethirtyeight.com/soccer-predictions/champions-league/

[little dave hedgehog]

hate those other sites where the jerks use fake analytics.

[leftpeg]

I read that only 8 times in the history of the European Cup and CL has a side lost the first leg at home and gone through, which is a remarkable stat.

Only once in the CL era has a side gone into an away leg with a 2 goal deficit and gone through, no one has ever done it 3 goals down. It's a massive ask

I imagine that most of those were subject to the away goals rule, so the circumstances are a bit different. It's still a 3 goal deficit though, so a massive undertaking. That being said, I'm an eternal optimist when it comes to football, so I'm looking forward to the boys stepping up to the challenge.

[Mark79]

hate those other sites where the jerks use fake analytics.

I was being facetious, but I mean as opposed to the finger in the air numbers on here.

[Tatterdemalion]

538 are pretty well respected in terms of political polling - not that that says much about their ability to predict football scores. Evenso, 1 in 33 doesn't seem so long to me. It's only double Pink's assessment of 1/10 to 1/20. Whatever it is, it's going to be tough and really it doesn't matter, because the only choice we have is to believe we will do it (and then to believe it didn't matter if we don't...).

[Mr Pink]

FiveThrityEight, who use actual analytics and maths for their probabilities, give us a 3 in 100 chance of getting the result we need to progress. Or, 1 in 33.

Or 30 in 1000

https://projects.fivethirtyeight.com...mpions-league/

I like fivethirtyeight, but the below is a huge flaw in their model.


A poisson process, means it’s assuming that scoring of goals are independent from eachother. Think about the Barcelona game- where gini’s 2 goals came within a few minutes of eachother- they were not independent events! We scored, they get rattled, the atmosphere in the stadium ramps up, we get momentum, push and score again- not independent events!! ie, the fact that we scored the second goal, made it more likely we are going to score another.

Their model also isn’t accounting for the fact that we need to win by 3. It is just looking at both teams ratings and saying if they were to play what is the odds that Liverpool would win by at least 3. In a normal game if we’re 2 to the good against Madrid, we’re not gonna be going all out to get a 3rd, but their model doesn’t see this, as different circumstances.

So while fivethirtyeight is good, their model is way too conservative here.

[Mark79]

I like fivethirtyeight, but the below is a huge flaw in their model.


A poisson process, means it’s assuming that scoring of goals are independent from eachother. Think about the Barcelona game- where gini’s 2 goals came within a few minutes of eachother- they were not independent events! We scored, they get rattled, the atmosphere in the stadium ramps up, we get momentum, push and score again- not independent events!! ie, the fact that we scored the second goal, made it more likely we are going to score another.

Their model also isn’t accounting for the fact that we need to win by 3. It is just looking at both teams ratings and saying if they were to play what is the odds that Liverpool would win by at least 3. In a normal game if we’re 2 to the good against Madrid, we’re not gonna be going all out to get a 3rd, but their model doesn’t see this, as different circumstances.

So while fivethirtyeight is good, their model is way too conservative here.

OK, so what's your superior process?

Seriously though, they're coming far closer to your 1 in 10 shout than 1 in 100 (nevermind 1 in 1000!), so I'd take it if I were you

[Tatterdemalion]

I like fivethirtyeight, but the below is a huge flaw in their model.


A poisson process, means it’s assuming that scoring of goals are independent from eachother. Think about the Barcelona game- where gini’s 2 goals came within a few minutes of eachother- they were not independent events! We scored, they get rattled, the atmosphere in the stadium ramps up, we get momentum, push and score again- not independent events!! ie, the fact that we scored the second goal, made it more likely we are going to score another.

Their model also isn’t accounting for the fact that we need to win by 3. It is just looking at both teams ratings and saying if they were to play what is the odds that Liverpool would win by at least 3. In a normal game if we’re 2 to the good against Madrid, we’re not gonna be going all out to get a 3rd, but their model doesn’t see this, as different circumstances.

So while fivethirtyeight is good, their model is way too conservative here.

Ok good points well made. You're making the point that the next goal is influenced by the prior. Yes, I guess so. But definitely not normally distributed. Evenso - how often do RM ship 3 goals at home without reply? It probably is about 3 in 100, if that.