xG

i voted

Useful info used in either way suggested

It's kinda useful, sito

I'm not even sure what it is tbh!

i get annoyed when people use it to say that the team is actually better than it is when they underperform their 'underlying numbers', as if poor finishing doesn't exist, only poor luck. it's useful when people point it out to suggest a team is overperforming or creating lots of good chances and doing little with them

Automatic thread bump. This poll is closing tomorrow.

Automatic thread bump. This poll's results are now in.

surprised how many dyches/keysys we have on this board

lets talk

Oh I forgot to vote, make that 3, altho my position is 'vaguely useful, but mostly from a defensive POV, not an attacking one'

Like, a team can go 2 up with some low XG clinicality and then see the game out without much chance creation and be comfortable winners but the opposition getting a couple of headers over the bar will have them ahead in xG and it'll look like burglary when it isn't

surprised how many dyches/keysys we have on this board

I'm not sure I voted in this but then I have no interest in stuff like assists, completed passes etc anyway.

Mine was an anti-discourse vote. Clearly all measurables are hypothetically useful but to people who work in the game rather than random cunts. The way people talk about the sport now is generally very dull and the elevation of stats-chat is to me a companion to the ifab-ification of the experience of watching a match. Game’s gone, basically

i think xg is just like a neat way of talking about a team having had a few chances. the mythologising and complication is only ever introduced by people criticising it/deliberately refusing to understand it. not so much here but there are some men of decent intelligence, like martin o'neill for example, who seem to act as if it's the most complex concept ever invented rather than a really simple way of talking about something that's always been talked about.

it's seldom you even hear it numerically dissected ime either, just briefly mentioned as a shorthand.

I guess I see it as the opposite of a neat way. If it’s not being numerically dissected then it’s just an exceptionally ugly and personality-free way to get a basic point across, so what’s the point.

The likes of MO’N either pretending not to understand it to make a point, or just stubbornly refusing to learn what it actually means, is another thing really (old men who used to be relevant, C/D)

Sympathise with crisp's mistrust of hyper-optimisation within sports (as well as fuckin everything else) tbh though

I think we'll get a better idea of how relevant MO'N still is on Sunday!

I guess I feel like the stat backing up the discussion means it's less arbitrary. Like it's not a made-up or nonsense statistic. Idk, there's a lot more annoying stuff in the forced ignorance and football man type discourse which borders on right wing at times.

The main problem with xG is that it just isn't scientific *enough*. Someone can miss a 0.8 xG chance cos they were 3 yards out with an open net and stuck it over but xG wouldn't consider if they were on the full stretch to a bouncing ball, etc. But then an accurate xG is almost impossible to measure even taking everything into account. Ultimately the best you can do is try to take into account how much was within that player's control to achieve in that moment

It uncomfortably reminds me of American sport's obsession with stats. Yet I'm a fan of the most stat obsessed sport on the planet, cricket.

Cricket feels like a sport that's fairly well-insulated against moneyballism due to its overwhelmingly psychological nature, which is why I'm actually not so annoyed by the vibes-based violence of Bazball as that feels like a reaction against the current of most modern sport. (The Hundred can fuck off forever obv)

Unless they're picking only the most psychmaxxed players obv. Shoaib Bashir sangfroid test off the charts etc

I think that's an anecdotal criticism of a statistic.

Wait, is xG based *only* on position on the pitch (in 2D plan view) relative to the goal? I thought a few things beside that were taken into account. In that case it can fuck off

It’s not one or the other though xxp

I’ve got no problem with the statistic itself but I consider it very boring to hear someone trot out stats during analysis. It also requires unpacking - did the 0.8xG come from one penalty or ten long-rangers. If you don’t unpack it then it’s empty and insight-free, if you do then you might as well not bother referencing the stat itself.

I’m perfectly fine with people working in football to optimise to their heart’s content, but I find the availability of abundant statistics to everyone only encourages the most boring people.

it's not even really that prevalent in analysis. it's definitely a case where people complaining about it is taking up more room than actual use of it, again i mean in the media rather than here.

it's used in passing like shots on goal or whatever.

I guess there’s so much content out there that there isn’t one football media experience for us all. I never hear people complaining about xG (the MO’N second captains live appearance was the first time in a very long time), but I do hear plenty of discussion of xG, xT and so on.

i think i hear people complaining about the complainers, which illustrates your point about content, lol. but there are some truly absurd angry rants about modern football and its language from dyche, allardyce, keys/gray, and an absolutely absurd to the point of amazing one from dean saunders, where he gets angry and claims 'park the bus' is some jargon complex term.

dyche literally used the words 'proper woke nonsense'

idk why I still bother with this fucking site

expected flag

it's just so noticeable the lengths you'll all go to, to talk around me, even when I'm making relevant points. still the idiot child

don't think of you that way at all, just a fair few people commenting at once.

Yeah, also I don’t know anything about cricket

ok good talk

firstly it is used a lot ime

secondly, there's no such thing as an xG. it has no place in any discussion about what did happen in a game, and it has a lot less place in a discussion about what could/should have happened in a game in a meaningful way than actual events that already were in use - shots on target, big chance

thirdly- i also understand it as a location only and depersonalised statistic, as if harry kane at the end of any given pass is comparable to richarlison at the same juncture, furthermore without reference to the quality of the chance ito marking, ball position, ability to set oneself

its trying to measure something but in a flat and irrelevant way, it measures a tabletop representation of football and it is presented during games as a neutral aspect as if entire tactics aren't set up around "quinny will take that chance that we set up specifically to unlock a weakness in centre halves aerially in this iteration of the epl" (for instance)

the writing off of o'neill in his lazily contemptuous handwaving away of it (he is an intelligent and articulate man generally, let alone for a footballing man) doesn't sit well with me.

dyche/pulis and yeah latter o'neill tactics were effective enough to an extent precisely because they dealt in realpolitikitaka and while the students of the game may advance through and into backroom ipad roles its actually quite rare that a purely dogmatic user of a concept (as opposed to an actual tactic, to be clear) actually leads men to success in an oppositional and fractured (ito long and unpredictable phases of play) team contact sport. youse boys might progress over a season when ye are playing this game on paper against each other, these guys say, but today, against us, you have to account for quinny.

and treating that level of applied inarguable as less worthy than something that doesn't strictly speaking exist other than as emperor's new kit is imo simply incorrect.

MON or Dyche don't "misunderstand" xG they just think it's nothing to do with how you prepare to compete in a game of football.

tactics is a nonsense bill, says gilesy, and he's right for 90% of football matches.

xG isn't even tactics, it's a unit for people that can't function without a taylorite measurement to see what any illiterate headbutt contest loser from burnley knows instinctively when watching through the stupor of ramsweat and bitter- "got away with that one", or "they couldve played all day and not scored"

except those statements are true from actual events to be intepreted with an understanding of the game and the actual difficulty of any one chance, whereas xG itself is some aggregate of less than satisfactory characteristics and therefore not even representative of what the hnwellactuaries of the modern game claim in order to give those in need of certainty and reassurance the comfort they crave that the human element, the baser element, of the game can be eradicated by time and processing power

cunts

It's honestly more useful in leagues of poor quality that have parity.

It don't tell you as much in leagues where there's a humongous gulf between top half and bottom.

This thread should just be about do you like statistics

Statistics only show one thing!!!

xP

I'll deal with this thread properly tomorrow.

secondly, there's no such thing as an xG. it has no place in any discussion about what did happen in a game, and it has a lot less place in a discussion about what could/should have happened in a game in a meaningful way than actual events that already were in use - shots on target, big chance

it is a way of measuring what was a big chance or not. shots on target is a less specific metric.

thirdly- i also understand it as a location only and depersonalised statistic, as if harry kane at the end of any given pass is comparable to richarlison at the same juncture, furthermore without reference to the quality of the chance ito marking, ball position, ability to set oneself

not true really, it's applied both to teams and to players in terms of their performance, so if you wanted to you could consider both together. marking and ball position etc are considered in most versions of it, that's what it is, the chance based factors like that.

its trying to measure something but in a flat and irrelevant way, it measures a tabletop representation of football and it is presented during games as a neutral aspect as if entire tactics aren't set up around "quinny will take that chance that we set up specifically to unlock a weakness in centre halves aerially in this iteration of the epl" (for instance)

it could be presented in a deeper way, though i assume you don't want that? but i did already say that at the moment it is sort of just mentioned in passing. not sure how the quality of chances a team created is flat or irrelevant, unless again, the flatness is about how it's used by people on tv. i feel like we are straying into anger at one simple metric, but the fact nobody gets annoyed at shots on target or possession or whatever is telling here. half of the annoyance is because it is a metric with a modern-sounding abbreviation.

xG isn't even tactics, it's a unit for people that can't function without a taylorite measurement to see what any illiterate headbutt contest loser from burnley knows instinctively when watching through the stupor of ramsweat and bitter- "got away with that one", or "they couldve played all day and not scored"

except those statements are true from actual events to be intepreted with an understanding of the game and the actual difficulty of any one chance, whereas xG itself is some aggregate of less than satisfactory characteristics and therefore not even representative of what the hnwellactuaries of the modern game claim in order to give those in need of certainty and reassurance the comfort they crave that the human element, the baser element, of the game can be eradicated by time and processing power

right, so you agree it is just a simple metric. however you believe it has some huge cultural significance about how some group of people, somewhere, in the game or outside it, or both, think of football or coach teams. i disagree but that is very clearly why dyche and o'neill etc hate it.

and the point about their hatred that is ludicrous is that they act as if they cannot understand what it means. or that it has no meaning. when it has a very simple meaning. in dyche's case it's even more farcical because he probably does have data analysts etc in his work. allardyce was one of the first people to start using prozone and data, yet he will always be heard ranting away about xg.

it's about presenting a version of yourself and of football which is false but pleases some higher god of englishness.

mon on the other hand is just fairly bitter.

i take on board that it may (or may not, as i understand it) take into account more than location- can we fact check

that's kind of key to whether it is or isnt even akin to a simple metric imo, because even if it does take other factors into account then it cannot ever actually be called complete, it is the output of a model.

a shot on target is a shot on target. that's a completely different category of metric, it's a thing that happened. an Xg isnt. that's kind of core to my problem with it. it's an inherently parodiable aspect of any analysis whatsoever, any mention of it is cringe, basically.

possession is another actual metric of something that happened, but ofc i will note that obsession with it absent the full acknowledgement that there is useless possession, dangerous/useful possession and counterproductive/risky possession is another direction of analysis of the game that os absolutely ripe for the contempt of all right thinking football men tm

mon and dyche acting as if they cannot understand it is good not bad, it's not a demonstration of the complexity of the idea nor their inability to comprehend it, it's their showing the relevant contempt for it from their perspective and tbh regardless of whether you agree with them or not it's a valid way to treat something you think is stupid imo, good punditry that for me trevor

doesn't "shots on target" include every feeble scuff that trickles gently into the goalie's waiting arms?

it does

theres no interpretation in it, being the point

and i hardly need remind you about what a punt from 30 yards meant against paul robinson or wayne hennessey

It's like an assist can be somebody ballooning an attempted shot so badly it can end up as a cross that someone gets on the end of.

i guess my point is a shot on target is only some kind of thing until you start looking at what comes under that metric, at which point surely it's just a woolier version of xG?

not at all imo

a shot on target is a thing

an xG is an idea of a thing

It's like an assist can be somebody ballooning an attempted shot so badly it can end up as a cross that someone gets on the end of.

― Massage Attack (Tom D.), 02 November 2025 11:52 (one hour ago) bookmarkflaglink

assists are a load of shit and should not be recorded

prob not surprising that the counter argument here is mainly vibes

not so

again, im failing to see where xG measures anything but the output of a model

the other statistics measure an event or an aggregate time of actual events

it's quite a clear distinction to me

my point is not only is xG not useful for what i see it used for, it doesn't in any satisfactory way actually represent what it says it does

an xG is an idea of a thing

"an xg" is your idea of this thing. nobody talks about 'an expected goal'. it's expected goals, a cumulative tally based on analysis of chances.

as for a shot on target, as nv says that's an even woolier thing. not least given someone hitting the post prob doesn't count, or does it?

furthermore, it might be good to actually know what expected goals are rather than asking me to fact check them for you. i already know, i don't have to google it. i know the metric used by sky is opta's one cos i looked it up a few years ago. it's not solely position on the pitch, there are about nine or ten other factors, at least.

as for dyche/allardyce, and separately mon, i've already explained why i think their views are hypocritical posturing for the first two, and bitter posturing for the third. the first two don't even think it's bollocks, they use it in their jobs and one of them pioneered it, but that is not convenient for the argument you wish to make in their favour.

my point is not only is xG not useful for what i see it used for,

if this is just it's not talked about v well nobody prob disagrees. if it's this is a nonsensical/highfalutin/idiotic metric then i disagree for reasons already stated.

google is actually telling me dyche is kind of a proponent of it tbf to him. i've wronged him here.

i guess his rants about modern football are more vague.

I think the ‘woke nonsense’ stuff from Dyche is more about wearing gloves or having a good haircut or whatever

yeah, tho he does kinda do the whole 'it's not pressing, it's running around' or whatever.

a shot that hits the post is off target, regardless of xG