Opposition fans back from the game

[SouthDownsRoyal]

Would be interesting if Derby went down, I don't think they've been in the third tier in my lifetime.

Nice to have a post that doesn’t mention xG

[Zip]

Would be interesting if Derby went down, I don't think they've been in the third tier in my lifetime.

Third tier 85/86 when we won our first 13

[tidus_mi2]

Would be interesting if Derby went down, I don't think they've been in the third tier in my lifetime.

Third tier 85/86 when we won our first 13

Born in 89 so that is an affirmative to my earlier comment!

[muirinho]

If the description given on the Tilehurst End of Xg is accurate it isn’t a stat. It’s a subjective opinion wrapped up in smoke and mirrors to make it look meaningful.

Yup, seems to be just mathematical window-dressing over the subjective "fat chance, good chance, how-did-he-miss-that chance" of yesteryear. And, even then, it apparently fails to include crosses across the box that are inches away from a touch into an empty net in its calculation.

It's a stat. Like shots off target which treats a "shot" that nearly ends up a throw, or a Row Z special exactly the same as off-the-woodwork, it's a small part of the story of the game.

[Franchise FC]

The ‘stat’ that bugs me is the shots on target which exclude shots that were clearly on target but blocked by a defender rather than the goalkeeper

[Snowball]

The ‘stat’ that bugs me is the shots on target which exclude shots that were clearly on target but blocked by a defender rather than the goalkeeper

I think "shots on target" really mean "shots the keeper has to stop"

a BLOCKED shot, I always think of as a shot stopped within a couple of feet of execution

Begs the question if a defender clears off the line is it a blocked shot or a shot on target

[windermereROYAL]

Seem to remember their keeper making a good save off his own defender too, in theory that can be classed as on target shot because it was from our attack.

[WestYorksRoyal]

Okay, final thing on xG. I saw one figure from Tuesday that our xG was 0.5. How is that calculated?

If Meite had missed his, it would have essentially been missing an open goal. Olise scored from a 1 on 1 situation. Granted, Laurent's
and Joao's were great finishes, but still decent half chances.

If our xG was 0.5 on Tuesday, how can the measure have any credibility?

[Stranded]

Okay, final thing on xG. I saw one figure from Tuesday that our xG was 0.5. How is that calculated?

If Meite had missed his, it would have essentially been missing an open goal. Olise scored from a 1 on 1 situation. Granted, Laurent's
and Joao's were great finishes, but still decent half chances.

If our xG was 0.5 on Tuesday, how can the measure have any credibility?

The xG for Meite's and Armstrong's equalisers were the same as it doesn't take into account where the opposition dedenders are and if the goal is empty or not.

[Franchise FC]

Okay, final thing on xG. I saw one figure from Tuesday that our xG was 0.5. How is that calculated?

If Meite had missed his, it would have essentially been missing an open goal. Olise scored from a 1 on 1 situation. Granted, Laurent's
and Joao's were great finishes, but still decent half chances.

If our xG was 0.5 on Tuesday, how can the measure have any credibility?

The xG for Meite's and Armstrong's equalisers were the same as it doesn't take into account where the opposition dedenders are and if the goal is empty or not.

And on that particular illogical note - for xG I'm out

[Snowball]

I have an idea

We take the number of goals a team scores. Call these "Goals For"

We take the number of goals a team concedes. Call these "Goals Against"

We take the second total from the first. We could call this, say, Goal DIFFERENCE


any good?

[Franchise FC]

I have an idea

We take the number of goals a team scores. Call these "Goals For"

We take the number of goals a team concedes. Call these "Goals Against"

We take the second total from the first. We could call this, say, Goal DIFFERENCE


any good?

Definitely works for me - surprised no-ones thought of it before
Clearly doesn't work in American sport where goal difference of zero (in any match) isn't allowed

[tidus_mi2]

Okay, final thing on xG. I saw one figure from Tuesday that our xG was 0.5. How is that calculated?

If Meite had missed his, it would have essentially been missing an open goal. Olise scored from a 1 on 1 situation. Granted, Laurent's
and Joao's were great finishes, but still decent half chances.

If our xG was 0.5 on Tuesday, how can the measure have any credibility?

The xG for Meite's and Armstrong's equalisers were the same as it doesn't take into account where the opposition dedenders are and if the goal is empty or not.

If that's true then they really need to improve xG before it can be called credible. What would people call the Meite chance, a 95% chance to score, possibly higher? So really that alone should be a significant increase to our xG for the game.

[Pepe the Horseman]

Courtesy of @shakefon on twitter.

[Kitsondinho]

I have an idea

We take the number of goals a team scores. Call these "Goals For"

We take the number of goals a team concedes. Call these "Goals Against"

We take the second total from the first. We could call this, say, Goal DIFFERENCE


any good?

Perfect! xG is a load of absolute Oxford. It reminds me very much of possession stats in recent years. Barca having 75% possession was great for them, because they were deadly with it. Under Stam, we were having lots of possession, but (especially in his 2nd season) there were times when we weren’t actually doing anything with it. In the same way, having a high xG is no use if you don’t score many goals.

[windermereROYAL]

I have an idea

We take the number of goals a team scores. Call these "Goals For"

We take the number of goals a team concedes. Call these "Goals Against"

We take the second total from the first. We could call this, say, Goal DIFFERENCE


any good?

That will never catch on.

[Snowflake Royal]

I find it amusing that people look at a statistical prediction tool that says the team clear at the top of the table should be mid-table and say "The table is wrong, they're lucky and will fall away" rather than "the statistical model isn't very good is it"

[Kitsondinho]

I find it amusing that people look at a statistical prediction tool that says the team clear at the top of the table should be mid-table and say "The table is wrong, they're lucky and will fall away" rather than "the statistical model isn't very good is it"

Indeed.
Under the xG model
Greece don’t win Euro 2004
Wigan don’t win FA Cup in 2013
Leicester don’t win PL in 2016
Portugal don’t win Euro 2016
and so on...
It is just a mathematical variation of ‘on paper, their squad is the strongest in the league.’

[SWLR]

I find it amusing that people look at a statistical prediction tool that says the team clear at the top of the table should be mid-table and say "The table is wrong, they're lucky and will fall away" rather than "the statistical model isn't very good is it"

Indeed.
Under the xG model
Greece don’t win Euro 2004
Wigan don’t win FA Cup in 2013
Leicester don’t win PL in 2016
Portugal don’t win Euro 2016
and so on...
It is just a mathematical variation of ‘on paper, their squad is the strongest in the league.’

or "to good to go down"

[Hound]

in saying all that, I do have a small interest in the XG table.

Only just to see Brentford top as they have been for the past 18 years or so.