Transitions, Transitions, Transitions!

[Tannochsidebear 2,322]

Who decides if it is a good chance (and therefore increases the Xg) or a not so good chance. Surely that is subjective, as with most sport?

 

I can (and usually do!!) discuss/argue with my mates about a sitter someone missed that they think is a hard chance or vice versa. I also think defensively some are shouting great save when I think it is routine, and vice versa.

 

If you put 6 football fans in charge of collating these stats you will get 6 different answers as to how big the Xg should be on both sides, making the whole process completely subjective, and therefore, useless?

 

Some of the stats coming out now in football are completely laughable and are being created by the FIFA computer game generation, where a player is known by a code (DM, AM) and a number between 1-100, which in most cases bears absolutely no comparison to their ability to influence football matches or their actual footballing ability, and takes no count of team tactics and how that fits in with the players talents, team/player moral and form. It seems to me this Xg stuff, while a decent debating tool to make interesting threads such as this, are taken far too literally and seriously by some. Watch the game closely, you can tell if the team should have scored 3 or 4 goals without the need of computer graphics.

 

I didnt see the game on Saturday so cant comment on how the Xg graphic relates to the real game that took place last weekend. Perhaps others who watched it live can comment.

[Rousseau 9,508]

Who decides if it is a good chance (and therefore increases the Xg) or a not so good chance. Surely that is subjective, as with most sport?

 

I can (and usually do!!) discuss/argue with my mates about a sitter someone missed that they think is a hard chance or vice versa. I also think defensively some are shouting great save when I think it is routine, and vice versa.

 

If you put 6 football fans in charge of collating these stats you will get 6 different answers as to how big the Xg should be on both sides, making the whole process completely subjective, and therefore, useless?

 

Some of the stats coming out now in football are completely laughable and are being created by the FIFA computer game generation, where a player is known by a code (DM, AM) and a number between 1-100, which in most cases bears absolutely no comparison to their ability to influence football matches or their actual footballing ability, and takes no count of team tactics and how that fits in with the players talents, team/player moral and form. It seems to me this Xg stuff, while a decent debating tool to make interesting threads such as this, are taken far too literally and seriously by some. Watch the game closely, you can tell if the team should have scored 3 or 4 goals without the need of computer graphics.

 

I didnt see the game on Saturday so cant comment on how the Xg graphic relates to the real game that took place last weekend. Perhaps others who watched it live can comment.

 

That's the point: it is a very subjective element of the game. The xG stat aims to be as objective as possible, taking into account objective historical data. Some of the data that is used to determine xG is:

 

"Angle of the shot

 

This should be fairly obvious. The more central a shot is taken, the more of the goal the shooter will have to aim at, and the higher chance he’ll have of scoring.

 

Distance of the shooter from goal

Fairly self-explanatory as well. Shots closer to the goal tend to find the net more frequently than those from outside the area.

 

Part of the body used to shoot

 

In terms of feet, this should be obvious. A player with two identical opportunities is more likely to score with his stronger foot, so this is incorporated into xG calculations. However, xG does not like headers, because they’re far harder to direct and generate power on, meaning that a shot with feet is almost always going to have a higher xG than a header from the same location.

 

Speed of the attack

 

This is partly included because of football’s lack of off the ball tracking data. The general theory behind this is that faster attacks are more likely to result in goals because the opposition’s defence will be unable to get back into defence in a strong defensive shape. Expected goals loves chances coming from counter attacks.

 

Type of assist

 

Perhaps a less obvious factor. Crosses (especially those in the air, leading to headers) are not a great way of generating high-quality chances. Instead, throughballs (because they lead to one-on-ones) and passes from the danger zone (because these leave the defence and goalkeeper out of position for an easy chance) generally result in high quality chances with little defensive pressures, so a chance assisted by one of these passes is given greater xG. Assists (or shots) following a successful dribble give the chance a higher xG because it is assumed (quite fairly) that there is less defensive pressure on the player, giving them room to get a shot away.

 

A few others

 

Individual errors often give sides an easy chance with most of the team out of position, a player rounding the keeper ramps up his shot’s xG and rebounds often give a free shot to an attacker in a good area, resulting in a higher xG for that chance."

 

No one is suggesting it's perfect. The elite sides all take into account data analysis because it cuts out as much of the subjectivity as possible, helping them make better decisions. It's certainly not perfect, but it is a useful tool. Surely you would trust a mathematical/scientific model which takes into account the most objective data over someone's opinion?

 

The irony is that everyone uses xG in their opinions after every game when they say, “We created the best chances in the game, but I just didn’t think we got the result that we deserved.” That is xG. That difference is that the above is a subjective judgement, whereas the xG stat takes into account objective data.

Edited March 26, 2017 by Rousseau

[Tannochsidebear 2,322]

"No one is suggesting it's perfect. The elite sides all take into account data analysis because it cuts out as much of the subjectivity as possible, helping them make better decisions. It's certainly not perfect, but it is a useful tool. Surely you would trust a mathematical/scientific model which takes into account the most objective data over someone's opinion? "

 

No, never, not in football!! Football is all about opinions, not emotionless stats.

 

Thanks very much for the explanation, very detailed and fairly common sense for most of it, but it is still very subjective and cannot be taken as gospel. Is each component part of a move that leads to a chance given a score, say out of 100, and then an average, or sum, of all the component parts taken to declare what overall mark a chance is given?

 

Take Miller's last minute miss in the Hogmanay OF game at ibrox that denied us a point. To me that was an absolute sitter, others told me it was a hard chance. Or Holt's chance in the cup semi-final at Hampden? Does the player the chance falls to come into it?

 

Take current players out of it to be less emotional. If a great chance falls to McCoist in his pomp, it mostly goes in, but if it falls to Sebo it mostly doesnt? It is still the same chance, and therefore the same Xg rating? But everyone knows that Ally will score and be amazed if Filip scores? So therefore as supporters we declare the Ally chance as a sitter and the Sebo one as a difficult chance because we have filtered in the player involved.

 

If both me and my mate, sitting side by side with a great view of the Copland goal can have opposing views on the quality of a chance, how on earth can a computer score the chance just using some variables that dont take into account the emotion of it all.

 

Also, a great chance in the 20th minute is not the same as a great chance at 1-1 in the 90th minute. But to Xg it will be? When commentators, managers, players, at the end of the game (without seeing Xg stats) say they had by far the better chances, it comes from a biased emotional viewpoint that on some occasions will be accurate and on others be wrong, but they will believe it no matter what Xg stats say a few days later.

 

MW constantly used stats throughout his tenure to defend a performance that was awful to watch. 70% possession, 20 shots at goal he would cry, but the game and performance was awful. Stats dont tell everything, that is why the game is an emotional rollercoaster that fans have loved for all their lives, and will pass on to the next generation.

 

The only stats that count in games are goals scored/conceded. And overall is 54 titles, 33 Scottish cups, 27 league cups, 1 ECWC, and 4 minor titles/cups.

 

Most fans would say that Murty did a good job steadying the ship for 6 games while the board got round to appointing a replacement to MW. However we dropped 8 points out of 12 in the league under GM, winning only one league game in 4 attempts, leaving us 8 points behind Dolly. Taking stats on their own, and given only one of the 4 games was against a top 6 side, he was an unmitigated disaster, but it doesn't tell the whole story.

[Gaffer 1,665]

I get the sense you're more than a little sceptical at the use of numbers in football. Whether you believe it or not, football is about stats and more and more clubs are realising this. It's funny that you mention the computer game mentality because the databases created by people at SI Games are used by many of the top clubs in the world. They use these databases to find players with particular attributes they are looking for and they're surprisingly accurate.

 

You can say it's subjective but that's just another way of saying that understanding the complexity of the numbers is too difficult so you'll use your own judgement instead. You can afford to do that because you're a supporter but clubs can't afford to so they use whatever they can to mitigate their risk of bad decisions.

 

One example of this was when Japp Stam was sold by United. For a couple of seasons the supporters and press had noticed that he was making fewer tackles than previous seasons and suggested he was past his best. When Fergie consulted his team they agreed it was time for him to go. This was a mistake and it was caused by people making a subjective decision based on simple data such as number of tackles, as well as just watching him. When the real data systems were introduced just two years later, they were able to demonstrate that he didn't need to make as any tackles because he was reading the runs better and marking his man better than others were able to. United made a very expensive mistake and Fergie now regrets it. You'll hear him say sometimes that you shouldn't listen to the stats, but what he meant was the simple stats. He, like just about every modern day manager, now understands the importance of match and player analysis like Roussea is sharing.

 

Embrace it TB, it's a good thing.

[Tannochsidebear 2,322]

I get the sense you're more than a little sceptical at the use of numbers in football. Whether you believe it or not, football is about stats and more and more clubs are realising this. It's funny that you mention the computer game mentality because the databases created by people at SI Games are used by many of the top clubs in the world. They use these databases to find players with particular attributes they are looking for and they're surprisingly accurate.

 

You can say it's subjective but that's just another way of saying that understanding the complexity of the numbers is too difficult so you'll use your own judgement instead. You can afford to do that because you're a supporter but clubs can't afford to so they use whatever they can to mitigate their risk of bad decisions.

 

One example of this was when Japp Stam was sold by United. For a couple of seasons the supporters and press had noticed that he was making fewer tackles than previous seasons and suggested he was past his best. When Fergie consulted his team they agreed it was time for him to go. This was a mistake and it was caused by people making a subjective decision based on simple data such as number of tackles, as well as just watching him. When the real data systems were introduced just two years later, they were able to demonstrate that he didn't need to make as any tackles because he was reading the runs better and marking his man better than others were able to. United made a very expensive mistake and Fergie now regrets it. You'll hear him say sometimes that you shouldn't listen to the stats, but what he meant was the simple stats. He, like just about every modern day manager, now understands the importance of match and player analysis like Roussea is sharing.

 

Embrace it TB, it's a good thing.

 

Its only a good thing if you can understand what your reading, which in effect is what you are saying with your Stam story. I have no problem with stats, (6 goals for Miller from 28 league games is one I have used often!) it makes discussions like this all the more interesting, but when people use them to defend the indefensible, or out of context, they cause more confusion than they are worth.

 

Watching a player with your eye to determine if they are good enough will never get old, no matter what their stats say. Same for teams as a whole. A team with 70% possession and 20 shots at goal must have put in a fantastic performance, right? It happened a lot with MW's Rangers side and those same performances were often utterly toothless and not pleasant on the eye.

 

I would add that people watching games on tv dont get the whole picture of what is going on, you can only tell that from being there and watching all the off the ball work done by teams/players. These stats sites are only going from tv pictures, so again they're only ball-watching which doesnt always give you the best information about that how players played in that game.

 

Let me give you an example -a team has a breakaway and the player with the ball makes a forward pass right to a team-mate to keep the move going, who then goes on and shoots on target but the shot is easily saved by the keeper. In stats-world, all is good. The passing player is marked up for good play, the shooting player has a shot on target. What was missed was the player out of shot who was in acres of space and if he had been the recipient of the pass a better chance should have been created. Stats-world is oblivious to this. I leave the match fuming with said player for failing to see the better option, but stats-world has him down as a plus.

 

Stats should only ever be as a side-dish, not the main event. My feeling is that too many people are using these computer program stats as gospel, when they are simply not.

Edited March 26, 2017 by Tannochsidebear

[Rousseau 9,508]

"No one is suggesting it's perfect. The elite sides all take into account data analysis because it cuts out as much of the subjectivity as possible, helping them make better decisions. It's certainly not perfect, but it is a useful tool. Surely you would trust a mathematical/scientific model which takes into account the most objective data over someone's opinion? "

 

No, never, not in football!! Football is all about opinions, not emotionless stats.

 

Thanks very much for the explanation, very detailed and fairly common sense for most of it, but it is still very subjective and cannot be taken as gospel. Is each component part of a move that leads to a chance given a score, say out of 100, and then an average, or sum, of all the component parts taken to declare what overall mark a chance is given?

 

Take Miller's last minute miss in the Hogmanay OF game at ibrox that denied us a point. To me that was an absolute sitter, others told me it was a hard chance. Or Holt's chance in the cup semi-final at Hampden? Does the player the chance falls to come into it?

 

Take current players out of it to be less emotional. If a great chance falls to McCoist in his pomp, it mostly goes in, but if it falls to Sebo it mostly doesnt? It is still the same chance, and therefore the same Xg rating? But everyone knows that Ally will score and be amazed if Filip scores? So therefore as supporters we declare the Ally chance as a sitter and the Sebo one as a difficult chance because we have filtered in the player involved.

 

If both me and my mate, sitting side by side with a great view of the Copland goal can have opposing views on the quality of a chance, how on earth can a computer score the chance just using some variables that dont take into account the emotion of it all.

 

Also, a great chance in the 20th minute is not the same as a great chance at 1-1 in the 90th minute. But to Xg it will be? When commentators, managers, players, at the end of the game (without seeing Xg stats) say they had by far the better chances, it comes from a biased emotional viewpoint that on some occasions will be accurate and on others be wrong, but they will believe it no matter what Xg stats say a few days later.

 

MW constantly used stats throughout his tenure to defend a performance that was awful to watch. 70% possession, 20 shots at goal he would cry, but the game and performance was awful. Stats dont tell everything, that is why the game is an emotional rollercoaster that fans have loved for all their lives, and will pass on to the next generation.

 

The only stats that count in games are goals scored/conceded. And overall is 54 titles, 33 Scottish cups, 27 league cups, 1 ECWC, and 4 minor titles/cups.

 

Most fans would say that Murty did a good job steadying the ship for 6 games while the board got round to appointing a replacement to MW. However we dropped 8 points out of 12 in the league under GM, winning only one league game in 4 attempts, leaving us 8 points behind Dolly. Taking stats on their own, and given only one of the 4 games was against a top 6 side, he was an unmitigated disaster, but it doesn't tell the whole story.

 

Ha! Theoretically, if Sebo and McCoist had exactly the same chance their xG would be the same, yes. However, like I said, this stat is taking the average historical probability; it's not taken in isolation. Firstly, McCoist will have more shots, from better positioning; whereas Sebo would have fewer shots, in worse positions. Then their xG would be compared to their actual non-penalty goals, probably -- for I do not have the data -- showing that McCoist would have done better than his xG, Sebo worse. xG would show that McCoist was a better finisher than Sebo. Similarly, Messi tends to out-perform his xG whereas a defender will under-perform. You're not the first to bring up finishing skill.

 

xG has to be used in context. I was simply using the team xG to show that we created a good few chances, better than average for us, and took more chances than we usually do.

 

Football is all about opinions; that will never change. But for me, opinions need to be based on evidence. On the one hand you have guys like Sutton who blow hot air based on nothing but their own ego, then you have guys like Gary Neville and Jamie Carragher who similarly have strong opinions but they are based on fact; the MNF guys always use a variety of stats to defend and back up their opinion. That's the way it should be done.

Edited March 26, 2017 by Rousseau

[Rousseau 9,508]

Let me give you an example -a team has a breakaway and the player with the ball makes a forward pass right to a team-mate to keep the move going, who then goes on and shoots on target but the shot is easily saved by the keeper. In stats-world, all is good. The passing player is marked up for good play, the shooting player has a shot on target. What was missed was the player out of shot who was in acres of space and if he had been the recipient of the pass a better chance should have been created. Stats-world is oblivious to this. I leave the match fuming with said player for failing to see the better option, but stats-world has him down as a plus.

 

You keep taking stats in isolation. Yes, the stat will show the player made a pass; great. But the stats will also show, through the historical data, what you see with your eye in one instance: the stats will show that he's not very productive with his passes if he keeps making the wrong decisions, for his pass completion will be down, and crucially, he'll have fewer assists.

 

It's not gospel; it never has been. It is simply a tool to broaden our understanding.

[boabie 230]

I must admit, the minute statistics are mentioned I switch off.

No statistic is ever going to change my mind that somebody is actually an important cog in a team if I think they are crap.

"Assists" is probably the most misleading thing that can ever be applied to a football player imo.

But then I've never played FIFA football manager on computer.

[colinstein 294]

you get the ball to the player who is furthest forward who's got the most space

mind you I was a tit at the game

[MacK1950 2,251]

Just remember there are lies,damned lies and statistics(these can be bent and twisted to give any end result you want ).

As I see it people put a lot of time and effort to prepare stats but what I also see is these things becoming more and more Americanised.