In his interview on Radio Wales, Erol Bulut stated that our xG had gone down since we lost Ramsey and O'Dowda. I thought I'd have a look, not out of any personal endorsement of xG, but by mentioning it, Bulut must read something into it.

Ramsey started our first 6 league games, O'Dowda our first 4. Our average xG over our first 4 games when both started was 1.55 per game. That rose to 1.63 per game following the games at Ipswich and at home to the Jacks. Our xG since losing both has dropped to 0.85. Bulut is right.

However, this is why I don't necessarily agree with the way he blames losing Ramsey and O'Dowda for our troubles in front of goal and why I'm not sure using xG is the best in this situation.

After we lost Ramsey, we won 5 out of 8 in the league. We hit our best run of form, and in some cases, performances, without Ramsey and O'Dowda. In 6 games with both we scored 10. In the next 8 without them we scored 13.

What might be interesting here is that, for the first 6 games our xG average was 1.63. For the next 8 that was nearly halved to 0.99 per game. During our first 6 games our total xG was 9.8 and we scored 10 goals, virtually the same. For the next 8 games our total xG was 7.9 yet we scored 13. We put 3 past Coventry and had an overall xG of 0.6 and put 4 past Huddersfield with an overall xG of 1.4.

For this spell we were clearly very good at scoring goals and creating chances, more so than when Ramsey and O'Dowda played. Defensively we conceded only 6 goals during this period but had an xGA of 10.4.

Here's a comparison, going back to averages again, of each set of fixtures.

First 6 games (with Ramsey and, for 4 games, O'Dowda):

xG average: 1.63
Scoring average: 1.66

xGA average: 1.3
Conceding average: 1.66

Next 8 games:

xG average: 0.99 (-0.64)
Scoring average: 1.63 (-0.03)

xGA average: 1.3 (same)
Conceding average: 0.75 (-0.91)

Since Bristol City:

xG average: 0.76 (-0.23)
Scoring average: 0.92 (-0.71)

xGA average: 1.48 (+0.18)
Conceding average: 1.42 (+0.67)