Stat Attack

[Matt Stevenson]

Unusually, Kelly (5) played as high as Smith (15) with both as close to the halfway line as Lerma. This looks balanced and all of the substitutions (Tavernier (16) for Ouattara (11), Cook for Rothwell and Semenyo for Solanke were all like for like). It was a brave call to remove Ouattara after 38 minutes due to low energy levels and highlighting a possible mistake in the starting line-up, but it’s pleasing that O’Neil didn’t let this sway him from trying to win the game.

 

[Matt Stevenson]

The back four had the most touches of the ball (all 60 or more). In the opposition half and final third Billing had most touches closely followed by Solanke and Christie.



We had 34 touches in the opposition’s penalty area, Solanke with 13. Ouattara had 5 despite being on the pitch for only 38 minutes

 

[Matt Stevenson]

Mepham and Stephens attempted most passes with 52, with the Welsh international having a 94% success rate. Rothwell had a 91% completion rate whereas Tavernier didn’t have his best day with a 63% success rate and also attempting less passes in 52 minutes than Ouattara did in 38 minutes.

 

[Matt Stevenson]

AFCB attempted 16 crosses, being successful with 3, slightly better than against Brighton, but still not great.



Our players tried to take on a man 22 times, being successful 8 times which is lower than average. Rothwell tried three times and was successful each time, Solanke, Christie and Tavernier tried 12 times between them succeeding twice.

 

[Matt Stevenson]

We created 9 chances (one big, which was Billing to Solanke). Leicester also created 9 chances with one big (the Vardy ‘attempt’ – these data are from Opta, so this seems a consistent potential error). Billing, Cook and Rothwell all created two chances.



We had 19 attempts (2 big) noting that chances can be provided after saves, deflections or a defensive error and not created. Thank you Maddison for providing Billing’s big chance. We hit the target 7 times. Brighton had 14 attempts, 1 big, and hit the target 4 times.

 

[Matt Stevenson]

AFCB won 20 of 38 aerials duals with Christie winning 3 out of 3.



There were many fewer tackles (16) than normal, where we have been averaging over 30 a game in the last three. Billing and Tavernier both attempted 3 tackles, Billing being successful twice and Tavernier no times.

 

[Matt Stevenson]

Mepham had most defensive actions with 16. Stevens made most interceptions, Neto made most recoveries, Mepham made most clearances and Mepham and Smith made our two blocks.

 

[KingKieffer]

Interesting fact, from our top four XGs from games this season, three of them come from the last three games(Everton being the other.) It would seem that we have finally found a way to threaten the opposition goal consistently. Hopefully this will stand us in good stead.

 

[Matt Stevenson]

Interesting fact, from our top four XGs from games this season, three of them come from the last three games(Everton being the other.) It would seem that we have finally found a way to threaten the opposition goal consistently. Hopefully this will stand us in good stead.

Yes, early on in the season there were plenty of games where we significantly under 1 xG. Averaging over 2xG in the last three games is a big improvement. We had 1.51 against Brighton, with Spurs only managing 0.83 this weekend.

Other games where we got > 1.5 xG were Fulham A, Leeds A, Forest at H, Newcastle at H, Man C at H, Arsenal A, the majority of which are recent, so maybe it is all coming together.

 

[KingKieffer]

Yes, early on in the season there were plenty of games where we significantly under 1 xG. Averaging over 2xG in the last three games is a big improvement. We had 1.51 against Brighton, with Spurs only managing 0.83 this weekend.

Other games where we got > 1.5 xG were Fulham A, Leeds A, Forest at H, Newcastle at H, Man C at H, Arsenal A, the majority of which are recent, so maybe it is all coming together.

Yes, we are certainly looking a more cohesive attacking threat. Earlier in the season even when we were picking up the occasional result it never really felt as though it was something that could be sustained as we just weren’t creating chances, we’re looking a lot better now.

Genuinely believe it has a lot to do with Solanke. GON mentioned in an interview a couple of weeks back that they were encouraging him not to drop deep, a lot of the threat we provide comes from him playing in the shoulder of the defender imo. Be interesting to see if there are any stats to back this up (average position etc.)

 

[Matt Stevenson]

Yes, we are certainly looking a more cohesive attacking threat. Earlier in the season even when we were picking up the occasional result it never really felt as though it was something that could be sustained as we just weren’t creating chances, we’re looking a lot better now.

Genuinely believe it has a lot to do with Solanke. GON mentioned in an interview a couple of weeks back that they were encouraging him not to drop deep, a lot of the threat we provide comes from him playing in the shoulder of the defender imo. Be interesting to see if there are any stats to back this up (average position etc.)

Here's a screenshot of Solanke's positioning over the season - note this is by Fantasy Game Week (so there are two in GW29, where we played Fulham and Brighton). Data is from Fantasy Football Scout. Green are home games. In GW28-30 Solanke has averaged nearer the opposition's goal (all being outside of the centre circle). GW 27 was Liverpool at home, so you'd probably expect our players to be deeper.



But, we could just have been playing easier teams too in GW28-30. I've eyeballed the average position in the games to see how Solanke compares with the next highest forward. It would take too long to try and put into graphics, so you'll need to trust me (or look back through the average position graphs).

With some exceptions (Fulham A, Chelsea A, Man U A) Solanke was typically not much farther advanced (if at all) than others. This was particularly when Moore was playing as we'd expect, but even when not (against W Ham A, Tavernier and Billing were higher, and Solanke was pretty much level with Fredericks, Christie, Cook and Lerma).

In the last run of games, Solanke isn't that much more advanced: Ouattara was as high at Villa and Traore was higher against Brighton. At Leicester, both Christie and Ouattara / Tavernier weren't much deeper.

This is a long-winded way of saying I don't know. It could equally be that the team is gelling more and thus having more attacks (and hence all players are further up the pitch) than Solanke being told to stay on the last man.

Hopefully this is useful and people can interpret the data better than me!

 

[SlowDownDerek]

Here's a screenshot of Solanke's positioning over the season - note this is by Fantasy Game Week (so there are two in GW29, where we played Fulham and Brighton). Data is from Fantasy Football Scout. Green are home games. In GW28-30 Solanke has averaged nearer the opposition's goal (all being outside of the centre circle). GW 27 was Liverpool at home, so you'd probably expect our players to be deeper.

View attachment 10316

But, we could just have been playing easier teams too in GW28-30. I've eyeballed the average position in the games to see how Solanke compares with the next highest forward. It would take too long to try and put into graphics, so you'll need to trust me (or look back through the average position graphs).

With some exceptions (Fulham A, Chelsea A, Man U A) Solanke was typically not much farther advanced (if at all) than others. This was particularly when Moore was playing as we'd expect, but even when not (against W Ham A, Tavernier and Billing were higher, and Solanke was pretty much level with Fredericks, Christie, Cook and Lerma).

In the last run of games, Solanke isn't that much more advanced: Ouattara was as high at Villa and Traore was higher against Brighton. At Leicester, both Christie and Ouattara / Tavernier weren't much deeper.

This is a long-winded way of saying I don't know. It could equally be that the team is gelling more and thus having more attacks (and hence all players are further up the pitch) than Solanke being told to stay on the last man.

Hopefully this is useful and people can interpret the data better than me!

Has our game-by-game xG improved with these seemingly more advanced tactics. Seems that way to me but would be interesting to see the stats.

 

[KingKieffer]

Has our game-by-game xG improved with these seemingly more advanced tactics. Seems that way to me but would be interesting to see the stats.

If you rank the top four games in which we have produced our highest XG this season, three of those are from the last three games.

 

[Matt Stevenson]

Has our game-by-game xG improved with these seemingly more advanced tactics. Seems that way to me but would be interesting to see the stats.

Certainly over the last 4 - I'm using the same source across the comparisons - in the last 4 we have an xG of 7.90 and an xG against of 7.86 (being comfortable ahead in 2 (Fulham and Leicester) and comfortably behind in 2 (Villa and Brighton). In the previous 26, we had an xG of 22.77 and an xG against of 44.84.

The averages are:

Last 4 xG 1.98; xG against 1.97

Other 26 xG 0.88; xG against 1.72

Having played an inform Aston Villa and Brighton it is probably unlikely that we have coincidentally had 4 easy fixtures. In all 4 games we created more chances than the average of the other 26 (1,36 at Villa being the lowest)

 

[SlowDownDerek]

If you rank the top four games in which we have produced our highest XG this season, three of those are from the last three games.

Who would have thought having a bigger squad full of better players would lead to better performances?

 

[KingKieffer]

Who would have thought having a bigger squad full of better players would lead to better performances?

I imagine everybody thought that.

Not a surprise that we see an upturn after taking out championship level players like Lowe and Marcondes, and replacing them with players who are proven at the top level in France and Italy.

 

[KingKieffer]

Certainly over the last 4 - I'm using the same source across the comparisons - in the last 4 we have an xG of 7.90 and an xG against of 7.86 (being comfortable ahead in 2 (Fulham and Leicester) and comfortably behind in 2 (Villa and Brighton). In the previous 26, we had an xG of 22.77 and an xG against of 44.84.

The averages are:

Last 4 xG 1.98; xG against 1.97

Other 26 xG 0.88; xG against 1.72

Having played an inform Aston Villa and Brighton it is probably unlikely that we have coincidentally had 4 easy fixtures. In all 4 games we created more chances than the average of the other 26 (1,36 at Villa being the lowest)

Also worth noting that the XG against for those four games is massively skewed by the crazy last ten minutes at Villa in which our strange substitutions completely handed them the game.

 

[Kicks_On_66]

This discussion on xG prompted me to gather our xG for this season from here: https://fbref.com/en/squads/4ba7cbe...ournemouth-Scores-and-Fixtures-Premier-League
and make some plots.
The first is a running record of our xG across the 30 matches:

That appears to show a trend up, and outliers against Everton at home and then Fulham at home (both about 2.5 xG).
I like to put data like these on something called a Shewhart chart (or control chart) that helps confirm whether what looks like a trend or shift really is. Here is a Shewhart chart for these figures:


I interpret this chart to show that xG began to improve after the first 8 matches. It confirms the home games against Everton and Fulham as unusually high xG given the history within this season. Finally, the last 3 matches also signal a shift up: they are varying around about 2 xG instead of the season average (also the central line on the chart) of 0.99.

 

[Kicks_On_66]

The match log from FBref has a couple other columns that I decided to practice my charting skills on.
Expected Goals Against (xGA):

This plot looks noisier and flatter. The corresponding Shewhart chart confirms that:

This chart displays no signals of a trend or shift during the season. The figures vary randomly around 1.7.
The FBref data also include a column named "Formation". I hope those are derived from average positions, but I don't know. I combine four less-frequent formations into "3-back" and "5-back".
Here is a plot that shows both xG and xGA within panels for each Formation. The lines connect the two points for each match, and the colors again correspond to the result:

The most interesting part of this chart for me is that 5 at the back was indeed fairly disastrous: 3 matches, all lost. Three at the back (arguably similar to 5 at the back for us) is not much better: 1 win from 5. Not a single draw from either of those two groups. The other Formations are associated with 27 of our 30 points. All the matches with 1.5 xG or more came from matches with those Formations.

 

[Kicks_On_66]

Three "bonus points" at Spurs today leave AFCB about two points ahead of the 40-point pace modeled by this tracker:

The 13-point "away bogey" for the season has been achieved. In my view, AFCB are now playing with house money on the road. Still some work to do, but the goal is in sight and well within our grasp.