Stat Attack

Matt Stevenson said:
The average positional map shows how high Zemura (33) played with Kelly and Stephens being more to the left than typical. This may be the enigma of Zemura, who set up the two big chances, but may not be defensively-minded enough (or have enough quality team mates) to be selected against the top teams in the division. Christie (10) Rothwell (14) and Moore had quite similar average positions, with Ouattara the most advanced of the starting line up.

View attachment 9629
Click to expand...

Personally I would say it shows that we just had more possession. In possession, Zemura pushes on more like a left wing back. Smith doesn’t play like this and covers more creating a back three.

Christie then supported O.Dango down the right flank, like Zemura supports Anthony down the left.

Out of possession it’s a more traditional back four.
 
Those stats back up the impression that Lerma was very solid, seemed to be blocking a lot, clearing a lot of headers etc
Disappointing on the 2 missed headers, would have thought Moore's would be more than 45% in the Premier League, losser likelihood the lower down the leagues you go
 
I didn't really get why people cared about some of these stats, but I'm finding these posts more and more interesting and I'm actually looking forward to them after a game now!

Thanks Matt S for opening up a new view of football I didn't realise mattered (although I'm still not completely certain it does) :)
 
TrevorMorganFanClub said:
I QUOTE="TrevorMorganFanClub, post: 702612, member: 428"]I didn't really get why people cared about some of these stats, but I'm finding these posts more and more interesting and I'm actually looking forward to them after a game now!

Thanks Matt S for opening up a new view of football I didn't realise mattered (although I'm still not completely certain it does) :)
Click to expand...

I’m not certain either!
 
dragonm8 said:
Stats are hindsight.

Easy to analyse after the fact.

It's like complaining about a meal, after you've eaten it.
Click to expand...

I agree, but it does help interpret what happened. Otherwise, we may as well not have discussions once the match ended.

However, I'll see if I can put something (it won't be that long) statsy-based prior to the Brighton game and see the reception it gets.

Although stats can hide so many things, that it will only be a crude summary to spark discussion (and I'm certainly no tactical expert). For example, stats could show that winger X had a ineffectual game, but that might be because a team double-marked him all game creating lots of space for other players to flourish. Another example would be if a team was fearful of an attacker's pace so that they dropped deeper, allowing more room for creative players to play than if the speedy player wasn't on the pitch. So trying to use previous stats to foresee how our game will play out will be highly uncertain.
 
Kudos said:
Just out of interest how are the rankings determined? Is it a single variable like a relative team performance metric or a combination of things?
Click to expand...
These ranks came from FiveThirryEight. I may have included a link on the previous chart. I haven’t paid any attention to their methodology, but what they have seems reasonable to me. I have been curious about how the various leagues compare.
 
dragonm8 said:
Stats are hindsight.

Easy to analyse after the fact.

It's like complaining about a meal, after you've eaten it.
Click to expand...
You have pointed out that records necessarily describe events that have already occurred. That is so true. Beyond that, I don’t get your point. Have you thought about why I might have picked those four particular clubs to highlight?
 
Kicks_On_66 said:
These ranks came from FiveThirryEight. I may have included a link on the previous chart. I haven’t paid any attention to their methodology, but what they have seems reasonable to me. I have been curious about how the various leagues compare.
Click to expand...

There was a link on the page you linked that takes you to a description of their methodology. I was being lazy:
How Our Club Soccer Predictions Work | FiveThirtyEight

It does look about right but I just wanted to take a look out of curiosity.
 
Amos Alonzo Stagg first said that after the University of Chicago's game against Wabash in 1905.” “Statistics are for losers and assistant coaches. Head coaches worry about wins and losses.”

But the ability to apply metrics to sports has definitely come a long way in 120 years!
 
More late sadness in the Brighton game where we nearly took a point that I’d suspect most would have taken at the start of the day, but which looked less valuable after Everton had beaten Arsenal and Wolves were three up against Liverpool. I’ve given two timelines because of the craziness of the 13th minute. Neto kicking the ball to Undav which resulted in him having 3 shots (one which was excellently blocked by Mepham) with an added xG of 1.47. The second chart shows the multiplicative impact (where it is assumed no additional xG were a goal scored which reduced the xG to 0.87. Apart from this minute, there were only two big chances (greater than 1 in 5) which both fell to Welbeck. The first was a header (3 in 10) that went straight at Neto, the second was a cross slightly too far in front of him that he grazed wide (44%). AFCB’s best chance fell to Lerma who slid into to meet Anthony’s cross but put it inches wide under pressure. Although, there was anther strong penalty appeal - Sky’s live commentary said AFCB were ‘unlucky not to get a penalty’, which along with ‘unlucky to give away a penalty’ are the phrases that partially summarise this season so far. The xG at half time was Brighton 2.61 (or 1.97 depending on how calculated) and AFCB 0.30. The second half was better for Bournemouth, as it was a drab affair with no big chances, the best being Semenyo’s blocked shot and a total of Brighton 0.44 and AFCB 0.67. If we hadn’t conceded late, we’d probably had looked back on a hard fought point, but we did with Mitoma sneaking into space between three Bournemouth players to head in well (1 in 12). Certainly, the decision to substitute Anthony with Vina is a discussion point, (see later) with the xG between Rothwell’s coming on and Vina coming on being Brighton 0.02 and AFCB 0.44. After it was Brighton 0.27 and AFCB 0.08.

xG.JPG

xG - mult.JPG
 
Last edited:
In general, the average positional map looks unremarkable, with two central defenders, Mepham (6) and Senesi (25), two advanced full backs Smith (15) and Zemura (33) alongside Lerma (8) and Billing (29) who sat deep, with the front four more or less similarly advanced. Christie for Traore seemed a pre-planned sub as we ease Traore in, but Billings injury meant another substitution slot used with Rothwell coming on, but averaging a position behind the centre backs, this showed that we were playing deeper, but looking to hit on the counter, and the xG was in our favour up until the Vina (18) substitution. Vina’s positioning showed how far we retreated with this change and is the only time I can recall an outfield player playing over 10 minutes having an average position in our box. Withdrawing an attacking winger, Anthony (32) for Vina may have changed the momentum of the game.

Av Posit.JPG
 
Neto had most touches in the game, followed by Zemura and Senesi; Billing looked likely to have most if he had not been injured two-thirds of the way through. Anthony and Traore had most touches in Brighton’s half, which is impressive for Traore as he was withdrawn around the hour mark. Outtarra had most touches in the penalty area (6) followed by Anthony (5).

Touches.JPG

Touches Pen Area.JPG
 
Billing and Senesi attempted most passes (31) followed by Neto. In Brighton’s half most passes were played by Anthony and Lerma. Assuming the reported stats are correct, the 264 passes attempted was much lower than normal, for example there was 359 against Forest and 435 versus Brentford. I don’t know whether this is due to a feature of Brighton’s play. Ignoring Vina’s 1 pass and 1 completion, Billing, Lerma and Anthony all had completion rates of 85% or higher. Traore’s was the lowest with 59% (10 passes completed out of 17)


All Passes.JPG

Pass Opp Half.JPG
 
You must log in or register to reply here.

;