Zidane sack watch - 19/20

[_00_deathscar]

Statistical models can be very useful, but you can't put your ENTIRE faith in them.

Separate sport, but a good example is the recent India v New Zealand T20 series down in New Zealand.
After the 12th over in the final T20, New Zealand were in a comfortable chasing position (having choked two big chases in the last two games, following on from their massive cock-up vs England in a similar manner in the World Cup Final). The statistical model had them at a 95%+ probability to win that final game at that point, which anyone watching the game would have told you is a load of bollocks. Sure, they were in 'control', but that didn't mean they were going to ease home. It took just one wicket to precipiate a complete collapse.

See also, Liverpool vs Barcelona last season. Prior to the second leg, no Firmino, Salah etc sure, Barca were in a great position and no one EXPECTED Liverpool to go through - simply just fight for it.
But at 1-0 down in the second half, Barcelona were still HUGE favourites to go through. That Barcelona/Rakuten-sponsored documentary they produced later with the behind the scenes in the changing room shows Barcelona absolutely shattered and rattled (Alba was basically in tears at half time...being 3-1 up on aggregate). Barca didn't have the cojones that night and that was obvious from kick off - a mixture of the Roma result, insipid leadership (Messi really isn't the greatest for this), the Anfield factor, a 1-0 lead with 45 minutes to go playing against the Kop, and Liverpool essentially riding that momentum.
Barca were actually shattered after that first leg despite winning 3-0 because they were struggling to sometimes live with the intensity of the team - but taking that lead in, and with no Firmino/Salah, they certainly were in control of the game. That should have changed BIG time once Origi scored, and Liverpool were to attack the Kop. But it didn't in the model.

I'm not sure how statistical models account for these factors, or if they do at all - but based on how certain teams like Man City are consistently rated in Europe, they clearly don't.

On another point, domestic cup games are VERY different to European Cup knockouts (for starters, two legged ties home and away).

 

[JPRouve]

Liverpool are proving that XG is a bag of shite. Never liked XG. Just use your damn eyes to see which teams and players are good

That's a bit harsh, like most statistics it's a good complement for the eye test. Some make too much of deal out of xG/xA but it's useful to put things into context. And no Liverpool have not greatly outperformed xG.

 

[giorno]

One thing about the xG and Liverpool, am I right in thinking they have greatly outperformed it last season and this. If this is the case then we are talking about over 60 league games now, surely this is too long of a streak to be a statistical outlier and there must be something the xG is missing.

The overperformance last season didn't stand out - it was fairly standard great side overperforming their numbers stuff, if on the higher end of the scale

This season on the other hand is ridiculous. Btw yours is also the first side to put this kind of overperformance over consecutive seasons in PL using understat's database. Other teams to have pulled that off include real madrid(14-16), barcelona(17-19) atletico (17-19) monaco(16-18) juventus twice(14-16 & 17-19 - highest overperformance on record btw) and, funnily enough, Nice!

In fact, it's very similar to juventus last year, grinding out results without putting up incredible numbers and outperforming xG and xGA by not so impressive amounts, yet massively outperforming xPts. In the case of juventus though it's easy to chalk that up to individual quality - they simply have much better players than anyone else. This matched up with the eye test, too - after a great start they were unimpressive the rest of the way but kept winning games because Cristiano, Dybala, Douglas Costa, Cancelo, Chiellini, Pjanic, Matuidi, Mandzukic, etc kept making decisive plays

It's possible we might have to start looking at liverpool in the same way

PS: city overperformed in both their 17/18 and 18/19 seasons, too. However, the xPts are still ridiculous - 91 and 90!

14/15 and 15/16 barcelona however are actually even more ludicrous putting up 94 xPts. That's so absurd they ended up (kind of) underperforming both seasons

 

[Pagh Wraith]

They're most likely comparing betting odds and real results. Betting odds = probabilities of expected outcome. Over a large sample size (like 10,000 + games) I would expect odds and outcomes to correlate pretty closely. So if 33 out 100 teams priced at 3.0 win, then the odds are pretty efficient. That part I have no problem with.

Being right over a large sample size however, doesn't mean the bookies are right over every single one - which was my point all along, the odds for this game seem off - not Spurs, not Bayern, not Liverpool, not Juve, not Barca... City.

@Pagh Wraith

It'd be interesting to compare opening odds vs closing odds for this matchup. My belief is that the odds for Madrid would shorten sufficiently from mid Dec to mid Feb.

You wouldn't get that perfect correlation (i.e. a favourites expected to win ~55% of the time really end up winning ~55% of the time) if you used random numbers or let's say historical 1X2 averages. You can always make the test by blindly betting the Pinnacle closing price (data is easily obtainable from football-data.co.uk). You'll end losing the exact margin of the bookmaker (about 2.5% for the Premier League for example).

Regarding opening vs closing odds: You are correct of course that odds movements do happen. Quite significantly sometimes, let's say when one of the team's star players gets injured. The other reason is, depending on the competition, that limits are quite low when the markets open and the big players don't get involved until late when they can get significant money down. Bookmakers and agents have become more conservative in that regard and right now I could only bet about 1,000€ on the handicap lines on the upcoming Premier League games at the weekend whereas a few years ago opening limits weere much larger early in the week. And this is top-tier, in League One limits are around 100€. This is due to the markets getting sharper and more efficient all the time. I could always re-bet of course but if my money is deemed as sharp, the line will have been adjusted. So for example, if I staked Wimbledon +0.25 at 2.05 for the maximum of 100€ now, and then wanted to place the same bet again, it will be down to 2.03. For the bookmakers this is a great way of getting good information cheaply if they have identified the smart money. So long story short, opening prices are quite volatile as any sort of sizeable stake moves the line. The limits for the upcoming CL fixtures aren't that high at the moment either so we can definitely expect some price changes, just not to the extent you might expect. Spiersey recently tipped some 16/1 women's team in the betting thread and they closed at 3/1 because the opening price was completely wrong. This does not happen in high-profile games with 24/7 media coverage. There is nothing the public doesn't know about Real and City at this point and any meaningful move would have to be down to injuries or similar events. So the markets are often wrong but they correct themselves until kickoff and the more mainstream a league or fixture is, the fewer error you will find in the lines.

The other thing is that the big syndicates operate on tiny margins but huge volume. They enter the markets late (often after confirmed line-ups) when the limits are high and then squeeze every last penny out of it. If the true price is 1.85 but 1.90 is available, they will stake it because they are so confident in their own prices. And that is why closing prices are so accurate.

One small amendment because you specifically referred to the City match. I don't say it is impossible for the markets to be wrong and you to be right. But you have to admit it is extremely unlikely for the price to be so completely off. You always have to ask yourself, what do you know that others don't. City having a run of bad results or losing in previous seasons is public knowledge and not valuable information you can beat the market with. Btw, City -0.25 opened at 2.24 in Asia and is now 2.23, so no movement.

 

[Stocar]

This part is incorrect. City concede significantly better chances than Liverpool do. They concede fewer shots, which is why their xGA is so similar. They are unquestionably worse at limiting chances.

You're right that the difference shouldn't be that big, that's down to variance - players shooting better against city. But it's likely that the higher quality of shots they give up plays its part in that

Maybe my language wasn't completely precise, but you're contradicting yourself here. How are Liverpool better at limiting chances (ie conceding less of them), when their xG conceded is about same as City's, who do tend to concede higher quality chances? This should mean Liverpool actually concede more chances (from what I've seen, they do - it's just that other teams mostly refuse to score, regardless of chance quality).

Anyway, as you pointed out, there shouldn't be a big difference in conceded goals, and one would even expect to have City as league leaders based on chances created/conceded. On account of Liverpool being obviously better at finishing, fair enough for them being in the lead despite not being as creative as their main opponents. But 22 points... That's where every rational explanation stops.

 

[Big Ben Foster]

They're most likely comparing betting odds and real results. Betting odds = probabilities of expected outcome. Over a large sample size (like 10,000 + games) I would expect odds and outcomes to correlate pretty closely. So if 33 out 100 teams priced at 3.0 win, then the odds are pretty efficient. That part I have no problem with.

Being right over a large sample size however, doesn't mean the bookies are right over every single one - which was my point all along, the odds for this game seem off - not Spurs, not Bayern, not Liverpool, not Juve, not Barca... City.

@Pagh Wraith

It'd be interesting to compare opening odds vs closing odds for this matchup. My belief is that the odds for Madrid would shorten sufficiently from mid Dec to mid Feb.

Not if the model is bullshit to begin with.

10,000 shots taken from 10 yards out by Ruud van Nistelrooy would yield you significantly different results than the exact same shots taken by Emile Heskey. xG treats them all the same. Garbage in, garbage out.

 

[Kevin]

United fans have a zidane sack watch?

You really cant make this up

Alot of these geniuses were trying to make a point about how they were right it was stupid of us to go for Zidane as our new manager when the rumors were about. They were actually triumphant when they thought their point was proven correct.

Gotten embarrassing now though.

 

[Paul_Scholes18]

The overperformance last season didn't stand out - it was fairly standard great side overperforming their numbers stuff, if on the higher end of the scale

This season on the other hand is ridiculous. Btw yours is also the first side to put this kind of overperformance over consecutive seasons in PL using understat's database. Other teams to have pulled that off include real madrid(14-16), barcelona(17-19) atletico (17-19) monaco(16-18) juventus twice(14-16 & 17-19 - highest overperformance on record btw) and, funnily enough, Nice!

In fact, it's very similar to juventus last year, grinding out results without putting up incredible numbers and outperforming xG and xGA by not so impressive amounts, yet massively outperforming xPts. In the case of juventus though it's easy to chalk that up to individual quality - they simply have much better players than anyone else. This matched up with the eye test, too - after a great start they were unimpressive the rest of the way but kept winning games because Cristiano, Dybala, Douglas Costa, Cancelo, Chiellini, Pjanic, Matuidi, Mandzukic, etc kept making decisive plays

It's possible we might have to start looking at liverpool in the same way

PS: city overperformed in both their 17/18 and 18/19 seasons, too. However, the xPts are still ridiculous - 91 and 90!

14/15 and 15/16 barcelona however are actually even more ludicrous putting up 94 xPts. That's so absurd they ended up (kind of) underperforming both seasons

I think you certainly should not just take these numbers at the surface. Actually watching the games tells a better story about how things go.
I think for rating strikers it can be good to look at. Which players can score better than expected and so on.
Thus it could indicate form if someone takes his chances better. Same with keepers potentially although xG is given even if it is off target right.
So you would need to look up xG for shoot on target to maybe rate the keepers. Still watching the saves a keeper makes can probably tell a much better picture about how good they have been in.

If a team waste chances you could see that on xG too. Need to ask why though and you need to watch the game to see that.
With City they seem to waste many easy chances in games recently. Hard to explain beyond low confidence in the squad and not being happy about Liverpool being so far ahead. With Pool I would claim the opposite. Being so far ahead just gives them more confidence.
Still Aguero has been in form despite that so give him chances and I expect him to score. Sterling not so much right now.

 

[Paul_Scholes18]

Not if the model is bullshit to begin with.

10,000 shots taken from 10 yards out by Ruud van Nistelrooy would yield you significantly different results than the exact same shots taken by Emile Heskey. xG treats them all the same. Garbage in, garbage out.

Yeah the models do not tell the full picture. Since they do not account for who gets the chance and the quality that player got.
Do not account for Van Dijk being a defender for Pool and Allison in goal too compared to Stones and Ederson (I do rate Ederson).
Also do not account when it happens and how it fully happens. They treat similar chances all the same even though all chances are different in many ways.
They have to statistically, but it also avoids telling the full story. I know they take into account many variables like keeper position, time factors, defenders positions, but so many they do not. If the use say 1-10 million goals then I guess they need to group them into around 1000-10000 different types of goals. I guess positions are very fluid too and not sure if they use symmetry or not. More right footed players could potentially impact things. Many good players with left foots though who often are very clinical.

I think when you compare a player to himself it can give very good info. Like compare Messis xG this season to last one and so on.
Thus it could tell do he get better service? Is he more clinical? Does he create more chances for himself potentially? Is he getting older/slower/worse? Confidence factors?
Looking at Lingard as an example. Is he shit in front of goal recently? Terrible service? Terrible movement/dribbling to get into chances? All of the above?
(answer yes from watching him and us play and the stats).
I guess you need to watch games to answer these things fully, but it could give good clues about it.

 

[giorno]

Maybe my language wasn't completely precise, but you're contradicting yourself here. How are Liverpool better at limiting chances (ie conceding less of them), when their xG conceded is about same as City's, who do tend to concede higher quality chances? This should mean Liverpool actually concede more chances (from what I've seen, they do - it's just that other teams mostly refuse to score, regardless of chance quality).

shots-based xG models will naturally favour teams that concede fewer shots. Give up 10 shots with an xG of .1 and they add up to 1xGA. Give up 3 shots with an xG of .3 and it adds up to .9 xGA

However, the .3 shots are 3 times more likely to go in, by logic being much better chances there's a greater chance of great finishing on those shots

It's likely a large part of the reason why Liverpool go from 24.6 xGA to 18.3 PS-xGA while City go from 23.6 xGA to 24.4 PS-xGA

(Statsbomb via Fbref)

But 22 points... That's where every rational explanation stops.

Yep

 

[Paul_Scholes18]

shots-based xG models will naturally favour teams that concede fewer shots. Give up 10 shots with an xG of .1 and they add up to 1xGA. Give up 3 shots with an xG of .3 and it adds up to .9 xGA

However, the .3 shots are 3 times more likely to go in, by logic being much better chances there's a greater chance of great finishing on those shots

It's likely a large part of the reason why Liverpool go from 24.6 xGA to 18.3 PS-xGA while City go from 23.6 xGA to 24.4 PS-xGA

(Statsbomb via Fbref)


Yep

xPT is very flawed though. I like xG, but do think giving it points is a bit silly. Just as saying the teams with most shoots in a game or most possesion deserves to win.
Football is about getting wins at the end of the day. You do that by scoring more than the opponent. Not by having great stats.
3-0 and 1-0 give the same amount of points. Although in xPT you do not get the points if you win the xG in the game, but rather they give you odds for how many goals you might get from each shoots probability of being a goal. If you have one poor chance and takes it with a super goal and then defends for 85 minutes without giving away barely any chances at all you deserve those 3 points. Even if xG might say it should be a 0-0 draw 90% of the time. Still watching the game you might feel the side is just keeping the lead knowing they will get the points. I remember us against Wigan under SAF at home one season. Scored early and then didn't do much after Rooney got injured I think, but they did nothing. It was only 1-0, but they never came close to get a draw that game. Not sure they got xG data for that game. Might be 1 vs 0.2 or something like that. Do not fully remember it beyond it being one easy win, but dull to watch. It was after a 3-0 win over Chelsea so I hoped we would go for the kill.

Not saying this is fully the case with Pool though, but they often do enough to win the games rather than going for big wins.
It might be boring and not fun to watch unlike City. Although 3 points is 3 points even if it is ugly.

 

[Siorac]

Not if the model is bullshit to begin with.

10,000 shots taken from 10 yards out by Ruud van Nistelrooy would yield you significantly different results than the exact same shots taken by Emile Heskey. xG treats them all the same. Garbage in, garbage out.

But then the team would underperform its xG and analysis would show that Heskey significantly underperforms, too, compared to someone like Van Nistelrooy. So the manager could say that "OK, we're actually playing well and creating chances but we need a better striker whose movement is at least as good as Heskey's but his finishing is far better".

It would be harder to actually find a use case for xG if it somehow weighted the chances according to player ability - a team underperforming their xG in that case would mean... what? That the players are performing below par, relative to their ability that was statistically calculated... how again? It'd be murky as feck.

 

[Paul_Scholes18]

But then the team would underperform its xG and analysis would show that Heskey significantly underperforms, too, compared to someone like Van Nistelrooy. So the manager could say that "OK, we're actually playing well and creating chances but we need a better striker whose movement is at least as good as Heskey's but his finishing is far better".

It would be harder to actually find a use case for xG if it somehow weighted the chances according to player ability - a team underperforming their xG in that case would mean... what? That the players are performing below par, relative to their ability that was statistically calculated... how again? It'd be murky as feck.

I agree it is great to compare strikers with xG. Not ideal, but a very good tool.
To judge a team though is harder. Saying someone is good because they got terrible strikers doesn't really hold.

 

[Siorac]

I agree it is great to compare strikers with xG. Not ideal, but a very good tool.
To judge a team though is harder. Saying someone is good because they got terrible strikers doesn't really hold.

You can, however, say that a team is creating chances but their finishing is poor.

 

[giorno]

I agree it is great to compare strikers with xG. Not ideal, but a very good tool.
To judge a team though is harder. Saying someone is good because they got terrible strikers doesn't really hold.

Other way around. xG are great at telling us how good a team is at generating chances

Finishing is something of a crapshoot, even the best show a lot of variance from season to season, what stays the same is they put up high xG - meaning they get a lot of chances

The outlier as always is one Leo Messi...

 

[Paul_Scholes18]

Impressive we turned this into a betting and xG thread . If City wins the xG battle do Zidane deserve the sack?

 

[Vault Dweller]

I feel quite confident.
They are not only defending well, it is the best defensive record in the history of la liga at this stage of competition (22 games-13 goals). There are also several players in a very good form.
I could name Casemiro,Courtois,Valverde,Mendy or Kroos,but actually everyone is playing good minutes when they have a chance.
Besides the midfielders have increased their offensive contribution a bit and Hazard will be back the next week,with plenty of time to be fit.
The fact that it is the only competition they have left may be an advantage or disadvantage.
We will see how they handle the competitive level. Madrid always move well in these scenarios, we will see how they do it.

To be honest, I think you will knock them out.

 

[Paul_Scholes18]

Other way around. xG are great at telling us how good a team is at generating chances

Finishing is something of a crapshoot, even the best show a lot of variance from season to season, what stays the same is they put up high xG - meaning they get a lot of chances

The outlier as always is one Leo Messi...

Lots of variation in chance creation too though and performances in many sides. Maybe not in the other big leagues at the top with the giants with similar players that dominate each year. PSG, Bayern, Juve, Barca/Real. City has obviously been great under Pep too, but they are worse this season when they are under pressure to get the goals.
In PL we have seen massive variation over the last few years.
The big players put up great numbers regulary though, but we see players with great variation too. Sterling might be the best obvious example. Cavani has had period missing chance after chance. Lukaku can be clinical on his day and miss chances for fun on his worst days. Rashford certainly not clinical either and that applies to this season too. Salah has been great in Pool all years though, but not as good as his first season.
Kane is probably the most consistent guy for getting the goals when in form. Lewandowski too, but he do not score as many hard goals as Kane. Aguero always gets injuries and other things stopping him recently, but great when in form.

 

[Suedesi]

Maybe my language wasn't completely precise, but you're contradicting yourself here. How are Liverpool better at limiting chances (ie conceding less of them), when their xG conceded is about same as City's, who do tend to concede higher quality chances? This should mean Liverpool actually concede more chances (from what I've seen, they do - it's just that other teams mostly refuse to score, regardless of chance quality).

Anyway, as you pointed out, there shouldn't be a big difference in conceded goals, and one would even expect to have City as league leaders based on chances created/conceded. On account of Liverpool being obviously better at finishing, fair enough for them being in the lead despite not being as creative as their main opponents. But 22 points... That's where every rational explanation stops.

Or you know, maybe it's not a good-fitting model after all.

 

[Paul_Scholes18]

Or you know, maybe it's not a good-fitting model after all.

The model is fine for what it does in terms of rating quality of chances. Just xPt and real points are totally different things.
In xG the difference is not that big. It just shows Liverpool are slightly more clinical than average. What you would expect from the leaders.
Goal against may underrate the quality of the goalkeeper in Pool and the defenders ability to make it harder than the stats suggest.
Blocked shoots etc not sure how much that effects things. Pool has got away with things like the goal against Palace etc and some close offside calls not sure if they count for xG or not since they got removed though. Thus I say they have maybe had 4-5 goals saved by luck and VAR.
I think Van Dijk has been superb this season and that improves those defensive stats.

Nothing crazy with the stats City got. Although they have had some games missing insane chances. Both games against Spurs mainly. Two totally insane games.

 

[Suedesi]

The model is fine for what it does in terms of rating quality of chances. Just xPt and real points are totally different things.
In xG the difference is not that big. It just shows Liverpool are slightly more clinical than average. What you would expect from the leaders.
Goal against may underrate the quality of the goalkeeper in Pool and the defenders ability to make it harder than the stats suggest.
Blocked shoots etc not sure how much that effects things. Pool has got away with things like the goal against Palace etc and some close offside calls not sure if they count for xG or not since they got removed though. Thus I say they have maybe had 4-5 goals saved by luck and VAR.
I think Van Dijk has been superb this season and that improves those defensive stats.

Nothing crazy with the stats City got. Although they have had some games missing insane chances. Both games against Spurs mainly. Two totally insane games.

By the same token though, last year and the year before every fecking thing was going their way. Goal after goal in the 95th minute, opponents were missing penalties in injury time etc. I just remember Sterling scoring winning goal after winning in injury time, it was so bloody annoying! The Crystal Palace guy who never misses a PK missing against Ederson, etc, etc.

 

[Stocar]

By the same token though, last year and the year before every fecking thing was going their way. Goal after goal in the 95th minute, opponents were missing penalties in injury time etc. I just remember Sterling scoring winning goal after winning in injury time, it was so bloody annoying! The Crystal Palace guy who never misses a PK missing against Ederson, etc, etc.

They were never overachieving or grinding out wins at the rate Liverpool are doing for last two seasons. Anyone who actually watched their games would know that. Stats confirm that, both advanced and basic ones.

 

[Paul_Scholes18]

By the same token though, last year and the year before every fecking thing was going their way. Goal after goal in the 95th minute, opponents were missing penalties in injury time etc. I just remember Sterling scoring winning goal after winning in injury time, it was so bloody annoying! The Crystal Palace guy who never misses a PK missing against Ederson, etc, etc.

That Palace game was annoying. Had a bet on a Palace win . They had the Bournemouth game when they played like crap and got a late winner. It turned that season around and they played well after that. Thought we had a chance to win it being top of the league at that time. Although very early in the season.
The game against Moyes was annoying, but I knew a Moyes side would let in a late goal. Huddersfield had a lead too around that time and lost it near the end. Just before the derby and once they beat us they had won it. Didn't bother as much after that.

 

[Suedesi]

You wouldn't get that perfect correlation (i.e. a favourites expected to win ~55% of the time really end up winning ~55% of the time) if you used random numbers or let's say historical 1X2 averages. You can always make the test by blindly betting the Pinnacle closing price (data is easily obtainable from football-data.co.uk). You'll end losing the exact margin of the bookmaker (about 2.5% for the Premier League for example).

Regarding opening vs closing odds: You are correct of course that odds movements do happen. Quite significantly sometimes, let's say when one of the team's star players gets injured. The other reason is, depending on the competition, that limits are quite low when the markets open and the big players don't get involved until late when they can get significant money down. Bookmakers and agents have become more conservative in that regard and right now I could only bet about 1,000€ on the handicap lines on the upcoming Premier League games at the weekend whereas a few years ago opening limits were much larger early in the week. And this is top-tier, in League One limits are around 100€. This is due to the markets getting sharper and more efficient all the time. I could always re-bet of course but if my money is deemed as sharp, the line will have been adjusted. So for example, if I staked Wimbledon +0.25 at 2.05 for the maximum of 100€ now, and then wanted to place the same bet again, it will be down to 2.03. For the bookmakers this is a great way of getting good information cheaply if they have identified the smart money. So long story short, opening prices are quite volatile as any sort of sizeable stake moves the line. The limits for the upcoming CL fixtures aren't that high at the moment either so we can definitely expect some price changes, just not to the extent you might expect. Spiersey recently tipped some 16/1 women's team in the betting thread and they closed at 3/1 because the opening price was completely wrong. This does not happen in high-profile games with 24/7 media coverage. There is nothing the public doesn't know about Real and City at this point and any meaningful move would have to be down to injuries or similar events. So the markets are often wrong but they correct themselves until kickoff and the more mainstream a league or fixture is, the fewer error you will find in the lines.

The other thing is that the big syndicates operate on tiny margins but huge volume. They enter the markets late (often after confirmed line-ups) when the limits are high and then squeeze every last penny out of it. If the true price is 1.85 but 1.90 is available, they will stake it because they are so confident in their own prices. And that is why closing prices are so accurate.

One small amendment because you specifically referred to the City match. I don't say it is impossible for the markets to be wrong and you to be right. But you have to admit it is extremely unlikely for the price to be so completely off. You always have to ask yourself, what do you know that others don't. City having a run of bad results or losing in previous seasons is public knowledge and not valuable information you can beat the market with. Btw, City -0.25 opened at 2.24 in Asia and is now 2.23, so no movement.

Good post, thanks for the insight.

As for the last paragraph, I don't claim to have any insider knowledge, I simply place more emphasis on things like tradition and past performance than current goalscoring form. I also have a bias against Pep G and City in Europe and acknowledge that Real Madrid are European royalty. It's the intangibles like Madrid have pedigree, they know what it takes to navigate these types of games, in fact they will relish the challenge. Whereas City might freeze.

Further, I think Madrid match up well with City: a tight defensive unit and a press resistant midfield that will not be bothered by City having the ball. Upfront the pace of Bale and Hazard's trickery might cause City's defense problems. I could see a draw at Bernabeu and Madrid winning at Etihad.

 

[Acheron]

Or you know, maybe it's not a good-fitting model after all.

Yeah, it isn't a perfect model so I don't know why is talking about it as the only and ultimate metric. Markov models can be handy for discrete time situations with discrete outcomes but also Bernoulli process. In other situations (like continous time situations with discrete outcomes) there would be more suitable stochastic process and models. So without getting too technical there are several limitations to that model that limit their application. Also I thought this was a Manchester United forum so I'm also baffled about how some of you rate City so much.

 

[Stocar]

Yeah, it isn't a perfect model so I don't know why is talking about it as the only and ultimate metric. Markov models can be handy for discrete time situations with discrete outcomes but also Bernoulli process. In other situations (like continous time situations with discrete outcomes) there would be more suitable stochastic process and models. So without getting too technical there are several limitations to that model that limit their application. Also I thought this was a Manchester United forum so I'm also baffled about how some of you rate City so much.

There will never be a perfect statistical model for any sport, especially football with its inherent variance and randomness. But this expected goals/points model is quite good, and I'm not familiar with any better. According to some nebulous indications in your post, there might be alternative models, can you be more specific? I rate City because they play brilliant attacking football, in a style that is more proactive and artistic than reactive and industrial, in contrast with other top teams. I like that style of football.

 

[Mb194dc]

Think chances of Zidane getting sacked looking rather slim! Title needs changing.

Something of a turn around, especially defensively, since conceding 5 in one half to Atletico pre season. I'm wondering if defensive issues won't re-emerge with Hazard and / or Bale in the team though.

Surely la Liga favourites now though given Barca issues and got a good chance in cups too.

 

[Paul_Scholes18]

There will never be a perfect statistical model for any sport, especially football with its inherent variance and randomness. But this expected goals/points model is quite good, and I'm not familiar with any better. Based by some nebulous indications im your post, there might be alternative models, can you be more specific. I rate City because they play brilliant attacking football, in style which is more artistic than industrial compared with other top teams. I like that style of football.

Think stats might be much easier to look at in individual sports like Tennis etc.
Sports with higher numbers of goals/points is probably easier to look at too. There will be more chances/oppertunnities to gain points
and sports like Basketball and Handball are very repetive in what is going to happen. The best side more than often win.
Although in handball a good keeper can make massive difference.

 

[Stocar]

Think stats might be much easier to look at in individual sports like Tennis etc.
Sports with higher numbers of goals/points is probably easier to look at too. There will be more chances/oppertunnities to gain points
and sports like Basketball and Handball are very repetive in what is going to happen. The best side more than often win.
Although in handball a good keeper can make massive difference.

Football can be infuriating, but also magical due to variance and randomness. I don't like sports where everything is too mechanical.

 

[Paul_Scholes18]

Football can be infuriating, but also magical due to variance and randomness. I don't like sports where everything is too mechanical.

Totally agree with you. I like to personally play Tennis, Table tennis, Basketball, but rarely watch it beyond Olympics.
It is nice when your team can beat the odds. Terrible when the opposite happens.

 

[GatoLoco]

Think chances of Zidane getting sacked looking rather slim! Title needs changing.

Something of a turn around, especially defensively, since conceding 5 in one half to Atletico pre season. I'm wondering if defensive issues won't re-emerge with Hazard and / or Bale in the team though.

Surely la Liga favourites now though given Barca issues and got a good chance in cups too.

I think pre-season games shouldn't be a serious reference due to fitness reasons. Even once the season has started some clubs are particularly strong in September/October and face others who only start to get their best shape from November onwards which leads fans to misleading conclusions. So imagine what the situation is like in August. Big clubs shouldn't be playing each other so early, but it's just because of economical factors they do.

However, there has been a big turn around no matter that game because the change with respect to previous seasons has been massive.

Regarding Hazard, he was working very hard off the ball before his injury, and he contributes defensively on the ball too, so I don't predict a big issue here.

 

[nore1975]

In a sense Real are ahead of Barcelona not just in the league table. In moving on Ronaldo they have taken the hit of losing their talisman. They had the added benefit of getting 100m euro for him as well. This season has seen Real truly move forward. Lopetegui was unlucky with how he left the national job. His successor at the Bernabeu Solari proved to be a stop gap before Zidane returned to the hot seat.
Real have spent heavily last summer and are in the process of building a new side. Hard to tell where Barcelona are. At some point Messi will leave. It will be interesting to see how they cope.
I fancy Real to prevail over Man City given how poor City are at defending. They are in the driving seat re La Liga.

 

[Skills]

On course to once again beat Lionel Messi and Barcelona to a league title. Was a fraud redcafes footballing experts.

 

[GatoLoco]

This thread is like the stock market with lots of ups and downs

 

[Mb194dc]

On course to once again beat Lionel Messi and Barcelona to a league title. Was a fraud redcafes footballing experts.

I thought sacked by Christmas after watching pre season, especially when they conceded 5 in first half v Atletico.

Hazard been a disaster, they're better without him. Bottom line is Zidane tore up 352 and went back to what worked before. Defensively they were very solid before last few games.

I like the RM squad, lot of quality, even with players out and suspended.

 

[giorno]

Hazard been a disaster, they're better without him.

Bitter much

And no, we are so not better without Hazard. Like, not even close

 

[BalanceUnAutreJoint]

This thread is actually hilarious in hindsight.
Pre-season takes really are something uh?

 

[Eddy_JukeZ]

I did post that I don't think he should have returned there, but he has had a good season despite Hazard's poor form and injures. He's made their defense a rock.

The people calling him a fraud though

Said it before and I'll say it again: Frauds don't win the CL 3 straight times.

 

[charlenefan]

I did post that I don't think he should have returned there, but he has had a good season despite Hazard's poor form and injures. He's made their defense a rock.

The people calling him a fraud though

Said it before and I'll say it again: Frauds don't win the CL 3 straight times.

The Man City game said otherwise

It's amazing he's returned and is on the brink of yet more trophies but he's been helped by having a really poor Barcelona side be their nearest challengers. This is a poor Real Madrid side

 

[Eddy_JukeZ]

The Man City game said otherwise

It's amazing he's returned and is on the brink of yet more trophies but he's been helped by having a really poor Barcelona side be their nearest challengers. This is a poor Real Madrid side

Even in that game, they did look fairly comfortable until they blew it late.

Barcelona have been poor, but you can only beat who's in front of you.