After producing some stats on the first 11 games;
- Passing Stats https://www.celtictrends.com/post/passing-stats-first-11-games
- Defensive Stats https://www.celtictrends.com/post/defensive-stats-first-11-games
- Shooting Stats https://www.celtictrends.com/post/shooting-stats-first-11-games
Below is a table on NP xGD and the expected goals differential between expected goals scored and expected goals against (minus penalties).
The table is ranked on NP xGD and Celtic lead the way on 24.2 (2.2 per game).
Also listed is the amount of games won (and lost) based on NP xG. Celtic have had more NP xG than their opponents in every game and Rangers have had more NP xG than their opponents in 9 of their 11 games. The two games where Rangers had less NP xG than their opponents were against Celtic (0.8 v 1.56) and St Johnstone (0.43 v 0.86).
Livingston are the team with the lowest amount of games where they had more NP xG than their opponents with only 2 games. Versus Aberdeen (1.56 v 0.59) and St Mirren (1.65 v 0.7).
The table also looks at games where teams have had more than 0.25 NP xG than their opponents, more than 0.5 NP xG than their opponents and more than 1 NP xG than their opponents.
In 8 of Celtic's 11 games Celtic have had more than 1 NP xG than their opponents. The 3 that weren't were against Hearts (1.29 v 1.2), Rangers (1.56 v 0.8) and Livingston (0.91 v 0.69).
Dundee Utd are 2nd best in regards to amount of games having more than 0.5 less NP xG than their opponents with only 1 game in that category. Versus Celtic (0.95 v 3.74).
NP xGD has been used in preference to xGD as penalties can sometimes not give a true reflection on chance creation and a team's probability to win.
However it has also to be taken with a bit of caution given that teams scoring from a penalty maybe less likely to need a goal.
In general, game state should be taken into account when using xGD as teams may not create the same amount of chances when they are winning (or losing) for example.
For the first 11 games of the season there was an average of;
- 13.9 shots per game for Celtic when Celtic were winning
- 5.8 shots per game for Celtic when Celtic were losing
- 17.3 shots per game for Celtic when Celtic were drawing
- all taken from shots per minute (in game state) multiplied by 90
For Rangers it was;
- 16.2 when winning
- 7.8 when losing
- 16.3 when drawing
For Celtic's opponents, when playing against Celtic they had an average of;
- 2.2 when winning
- 8.5 when losing
- 6.5 when drawing
For Rangers' opponents it was;
- 0.9 when winning
- 6 when losing
- 9.7 when drawing
For Dundee Utd it was;
- 2.6 when winning
- 7 when losing
- 9.7 when drawing
What I don't have though the information to say whether or no xG per shot varies for game state (ie are shots more desperate when losing?)
Celtic Trends 🍀📈⚽️