For some reason I always cringe when pundits and fans etc wax lyrical about a player’s assist stats.
I personally think that they can overstate a player’s contribution. I mean, for a player to just pass a ball sideways to another player who then pulls off a great strike or bit of individual brilliance and to then get credited with an assist seems a bit off to me.
I know that it’s difficult to differentiate between the quality of an “assist" when compiling stats, so the easy option is to describe the last person to touch the ball (before the player that scores) as providing an assist.
The opposite can be true also. A player can make a very creative pass to another player, giving them a golden opportunity to score, but then they fluff their lines…..and then the “creator" of the chance does not get an assist.
Does anyone else think that the assist stats is a bit misleading?
Assist Stats
posted on 25/2/25
comment by Edinspur (U1109)
posted 32 minutes ago
comment by Darren The King Fletcher (U10026)
posted 11 minutes ago
Yeah, expecting an assist when passing to Nunez is not the same as when passing to Haaland.
----------------------------------------------------------------------
But it’s not strictly relevant. It’s not adjusting for player quality as that’s not the point of the metric and to get quantifiable data would be impossible. It’s just creating the value across the average of the whole dataset.
It’s the point in post-processing to determine what the data is telling.
----------------------------------------------------------------------
I won’t things like that deter me from insulting Nunez.
posted on 25/2/25
comment by Calum Ferrie (U1734)
posted 7 minutes ago
comment by Bãleș left boot (U22081)
posted 12 minutes ago
comment by Edinspur (U1109)
posted 5 minutes ago
comment by Darren The King Fletcher (U10026)
posted 11 minutes ago
Yeah, expecting an assist when passing to Nunez is not the same as when passing to Haaland.
----------------------------------------------------------------------
But it’s not strictly relevant. It’s not adjusting for player quality as that’s not the point of the metric and to get quantifiable data would be impossible. It’s just creating the value across the average of the whole dataset.
It’s the point in post-processing to determine what the data is telling.
----------------------------------------------------------------------
So an xA is just a player putting the ball into a particular space on the football pitch regardless of who their teammates are. Presumably there has to be a player in the vicinity of the pass at least.. what about if the ball is spinning, or played to their weak foot? It soon becomes a bit of a nonsense trying to apply it to the real world and mostly is relegated to something that only stat nerds will care about. All of which to say, Brennan Johnson clearly understands it and ensures he is in the right place at the right time to convert high xA opportunities.
----------------------------------------------------------------------
Those are rookie "what ifs"
If stats tell us that gay keepers have a slightly lower save percentage than straight keepers and being gay is a spectrum, then what if a keeper is 51% gay and another keeper is 49% gay, should this be accounted for in the xA stat? And then what if the striker is 12% Danish on his ancestry stats and the Danish have a higher finishing rate than the Equatorial Guinean, and another striker is 15% that but also 9% Danish, but has a white mum and black dad when stats say that those with a white dad and black mum have a higher finishing rate. Does it take that into account?
----------------------------------------------------------------------
It doesn't really matter if we know a player actually has a cocck in their bum at the time the pass is played, as long as we know within a decent margin of error that they like a bum. xB stats will only ever get you so far these days, especially with these foreign sounding players that like big butts and DO lie about it.
posted on 25/2/25
comment by Darren The King Fletcher (U10026)
posted 10 minutes ago
comment by Edinspur (U1109)
posted 32 minutes ago
comment by Darren The King Fletcher (U10026)
posted 11 minutes ago
Yeah, expecting an assist when passing to Nunez is not the same as when passing to Haaland.
----------------------------------------------------------------------
But it’s not strictly relevant. It’s not adjusting for player quality as that’s not the point of the metric and to get quantifiable data would be impossible. It’s just creating the value across the average of the whole dataset.
It’s the point in post-processing to determine what the data is telling.
----------------------------------------------------------------------
I won’t things like that deter me from insulting Nunez.
----------------------------------------------------------------------
If anything its brilliant for highlighting how much of a clown that guy is:
Each season he has underperformed his xG, which is very remarkable: 0.76 so far this season, 8.19 last season and 5.37 the season before. Meaning that had he been an average finished he'd have had 15 more goals in his three seasons at Liverpool facking donkey
posted on 25/2/25
I definitely got triggered when he kept hitting the post or blasting shots narrowly wide for a while and Klopp and the Liverpool fans were convinced like he'd somehow been cheated and deserved to be considered a top striker still despite missing the goal frequently.
posted on 25/2/25
Agent of chaos.
posted on 25/2/25
I personally think that they can overstate a player’s contribution. I mean, for a player to just pass a ball sideways to another player who then pulls off a great strike or bit of individual brilliance and to then get credited with an assist seems a bit off to me.
_______
Think u can understate this too, theres something in picking the right pass however simple or being in the right position to do so etc.
posted on 25/2/25
I don't understand when people say this because surely it's equally true for goals
like yes they can be total flukes that don't tell you much. but the idea is that the volume with which they happen tells you something. nobody has gotten 20 assists in a season from knocking it sideways to players scoring screamers
posted on 25/2/25
comment by Christopher (U20930)
posted 14 minutes ago
I don't understand when people say this because surely it's equally true for goals
like yes they can be total flukes that don't tell you much. but the idea is that the volume with which they happen tells you something. nobody has gotten 20 assists in a season from knocking it sideways to players scoring screamers
----------------------------------------------------------------------
Gerrard pretty much that season with Suarez did this and punted it long for him to score out of nothing.
posted on 25/2/25
If it's equally true of goals it's equally true of results. What's the point of looking at the score if we have a predicted score?
posted on 26/2/25
comment by Edinspur (U1109)
posted 13 hours ago
But it’s not strictly relevant. It’s not adjusting for player quality as that’s not the point of the metric and to get quantifiable data would be impossible. It’s just creating the value across the average of the whole dataset.
It’s the point in post-processing to determine what the data is telling.
----------------------------------------------------------------------
Footy chat!