But these keepers prevent more shots than ones that stay on their line.

This is where a "possible xG prevented" model can help. What would the xG have been if the keeper hadn't come out and made an interception?

With those three concepts I think you get a pretty good overview of goal prevention.

Only issue is that more aggressive keepers can often rank poorly in a positioning model (vs the optimal position to suppress xG).

Coming out and not quite getting to the ball first can have a big impact, especially if an attacker ends up rounding the keeper or can head into a near-empty net.

This is why I think a goalkeeper positioning model is so important, like the one produced by StatsBomb. Can calculate an overall shot-stopping performance that's split into the impact of positioning & of diving/reflexes/etc...

Also, positioning seems more repeatable than goals conceded vs xGOT

And 90% of set pieces are flagged for holding, a blindside block or an illegal pick

▫️By isolating the 1st pass location & last reception in a pass & move combo, we can visualize some off-ball movement with event data

▫️Metrics could be created to see which players make the combos more often, whilst visualizations show trends

▫️Could also analyze who the players tend to connect with

⚽ Shot creation

👇 These graphs show off-ball runs, in each pass combo, which led to a shot (min 0.1 xG)

🔷 Wall passes in central areas & the half spaces seem an effective method of generating dangerous shots

🔷 Shot creating ABCA's tended to start from the right, involving right-footed players

🔷 Heat maps to show pass & move trends, where each combo type most frequently starts & ends

1️⃣ Progressive wall passes more common out wide, to move inside or up the wing

2️⃣ Sideways 1-2s often used to find space wide in the midfield- left side preference

3️⃣ ABCA's generated progression into the box

Luka Modric was, understandably, less direct in his movement in the previous World Cups, providing an option for midfield circulation.

His main trend is a preference to move right when playing a wall pass, and there are examples of forward runs down the right wing.

Kylian Mbappé shows much more specific movement after playing a pass

1️⃣ More direct than Messi, with the majority of his pass & move combos generating progression

2️⃣ Focus on wall passes, which help Mbappé to move inside from the left wing or progress in the left half space

There are clear trends in Messi's movement after making a pass

1️⃣ Uses wall passes to penetrate in the center of the pitch

2️⃣ Normally moves laterally to receive a 1-2 when occupying the right flank

3️⃣ Longer forwards runs, to create an ABCA combo are more common when Messi drifts to the left

Good thought, the wide and progressive passes could be separated into ground or high passes. Maybe adding switches of play as well.

Would be awesome with more advanced data to split up the triggers caused by possession team "errors", like loose touches, poor body shape, receiving on weak foot.

Also John, I'd be interested to get your thoughts on this. Are there any different pressing triggers that you would add to this analysis?

Thank you Rob! Glad you liked the idea.

Out of interest, what did you think of the pressing triggers that I chose to analyze? Any that you'd add?

Thank you John, glad you enjoyed it! Your piece on The Athletic was the inspiration behind the "non-progressive wall pass" trigger.

▫️Coaches can exercise a fair amount of control over pressing triggers & orientation

▫️Therefore, they can be analyzed & prepared for

▫️Such analysis supports video reports & helps coaches design specific sessions

✍️ Full article available here 👇

cambiodejuegoenglish.beehiiv.com/p/pressing-t...

5️⃣ How the trigger impacts decision making

▫️Creating pass clusters & filtering out the 4 most common attempted passes, following the 8 defined pressing tiggers 👇

▫️Keepers more frequently go right (open body shape for right-footer)

▫️Wide pressures often lead to backwards passes or balls up the line

4️⃣ Positioning whilst pressing

▫️With @statsbomb.com 360 data, I visualized how Sweden were organized when they generated pressure from the left or right side of the opposing keeper

▫️Close marking of central passing options, but at times leaving a short, sideways pass open 👇

3️⃣ Oriented pressures

▫️I then split specific pressing triggers into different types depending on their orientation

▫️Analysis of how the 8 Women's Euros quarter-finalists jumped to press a keeper 👇

▫️Norway jumped straight to press the keeper, Sweden & Spain more often came from an angle

2️⃣ How the trigger impacts press efficiency

▫️Pressing the keeper is risky, inevitably leaves a free player

▫️Jumping out wide & on non-progressive wall passes seems more efficient (considering ratio of recoveries to opposition progression)

▫️As expected, more recoveries after progressive passes