Xg....Question

I’ve never been a fan of this particular stat as I think there are too many variables in what is classed as a goal scoring chance.
If anyone is an expert of this stat can they please tell me how Coventry scored 7 goals on Saturday yet had an Xg of only 1.1?
I for one am baffled.

 

Poor defending.

 

Utterly bonkers. Someone will be along soon to tell us it’s simple…..

 

Simply put, they use data from across the world to determine the liklihood of a player scoring a similar chance.

So for Coventry, the goals they scored were more difficult to convert so resulted in a low expected goal, but high actual goal.

I lost my **** with Sormaz when he tried to use XG to defend Watters. If his expected goals we're higher than actual goals, it means he lacks the quality and composure required to put easy chances away.

 

The same way that the odds of winning the lottery are 1 in 45 million yet people still win.

 

Just a new obscure stat that doesn't add to any entertainment factor, so Garth Crooks & co can keep their job.

 

Old Mladen got plenty wrong but don't think you can lay that one at his door. I'm sure that was Neerav, now i think he's decent bloke but what he knows about professional football you could write on the back of a postage stamp. It was the infamous top 3 squad and at least 5 championship standard players car crash interview. I hope all concerned have learnt the lesson a little knowledge can be a bad thing and dont ever put Neerav in front of microphone talking about on field matters.

 

Non binary makes more sense or is easier to understand

 

I know Nareev said it in that 'golden' top 4 squad interview of his, but then I read or watched Sormaz justifying that comment and that's what boiled my piss. Nareev isn't the statatition. Mladen was meant to be and that's why it irked me to comment at the time.

 

Garth Crooks? Lol. Not seen him on owt for years, thankfully.

 

That's cos it's ballox.

 

Fair do's

 

Utterly utterly pointless data point created to make people who are not fit to judge a players ability to be parachuted into a position they should be nowhere near.

Using your eyes is a far simpler way of gaining better metrics.

For example. Give our mate Clement Rodrigues 30 simple chances to score and you’ll have zero goals and 27 injuries to supporters in the back row of the Ponty End. Give DKD 30 chances and he’ll convert far more.

Oh for the old days of common sense and watching a game and not just taking note of a screen of stats.

 

Does he not do Final Score anymore? I still have PTSD from his amdram ravings

 

Basically just quality of chance created mate. If they take all their shots from 35 yards out, xg will be incredible low. If they take a penalty, the xg is around 0.75. I think the free kicks declan rice scored against real madrid was 0.02 and 0.04 xg..

 

It's because you're judging an imperfect model from a match when its supposed to be useful over the long term.

Football is high variance and low scoring, so xg is often 'wrong' with most games especially the Coventry one, all the shots seem to fly in the top corner or find a way in, in the same way as some games you create a load of big chances and none seem to go in or only one.

xG is useful for the long term as it provides a figure for chance quality for and against after a longer run of games. Its going off the priniciple that if a team consitently gets a high xG and concedes a low xG, provided your keeper isn't crap and you finishers are sh*te in the long run results will reflect your perfomance.

To judge xG as nonsense when basing it off one game is like flipping a coin 5 times and it landing heads each time, claiming its a load of boll*x that its supposed to land heads only around 50% of the time. That's how short term variance works, the short term is just noise but over a longer period it tends to even out.

In the same aspect context isnt accounted for if just using xG as it doesnt account for if a team was winning or losing after 20 minutes.Or for every single detail that we can see with our own eyes played an impact or if you have max watters with low confidence in your team. Which is why its a descriptive stat rather than prescriptive, its telling you what kind of chances were created and conceded rather than saying you did brilliantly. Which is why just like every stat it's a tool so depends on how you use it rather than being the be all and end all like so many claim it to be.

With the Coventry game im sure a better model would get a higher figure, but wouldnt capture any where near 7 xG. But the principle is if you played that game 100 times over the outcome would rarely be 7, it would take an outlier in terms of clinical finishing that would very rarely occur again in a season.

Its a metric to assess process rather than the outcome. Which is why many fans understand find it pointless and i don't blame them. It's just when say they it's useless when its been used for ages at the best clubs and by gambling sites, when its used in a long term context its proven to be at least very useful.

It doesn't help that is misused by nearly everyone including the media who show it after single games (even in cup competitions) but never over the long term. And treated as gospel by Neerav who claims having the highest xG justifies keeping a player who clearly had a dip in confidence and needed a change in enviroment.

xG isn't about judging a single result. Its just a better way of understanding performance beyond the luck and randomness of goals.

 

The only stat that matters is 7-0 and the only people who need to analyse are the QPR backroom team.

 

+1

It's like the stat that shows teams scoring first win approximately 70% of all games, but then picking out a particular weeks results, or that of a certain team over the course of the season who come back to win on a semi regular basis - then saying the stat is worthless. Stats are descriptive, not prescriptive.

 

Thanks mate, and well done for trying to educate poor souls like myself ha. I am one for if theres a stat then please let it be a stat for “what actually happened “ as opposed to a stat for “what could have/maybe happens”, many an ex pro footballer thinks same I reckon. Thanks again for interpretation even though its not for me mate.

 

If this was the case every team in the country wouldn’t be using xG as part of their analysis and I guarantee they all do.

It’s a useful statistic alongside other stats when analysed over longer timeframes.