Jarrad Branthwaite

[Free Agent]

Was Badiashile injured this season or out on loan? (Just googled: injured)

[Shogun]

It's the player at the same age.

Badiashille currently probably undervalued despite being a similar profile to one of the top young defenders in the prem, partly because he's a big black guy.

I reckon Michael Keane could play 10 matches for Chelsea and get some decent stats as well.

[brap2]

I reckon Michael Keane could play 10 matches for Chelsea and get some decent stats as well.

Worth testing if they'd like to stump up the fee

[brap2]

Why is his passing stat showing inside the 25% ring when its 82%?

I would imagine most CBs have fairly high pass %s, so despite JBs 82% rate, he still ranks in the lower 25% of records. That said it must be fairly tightly grouped considering where Badiashille is.

The various scales looks inconsistent too, suggesting a greater visual disparity than the figures themselves.

I dunno what this bit means sorry. So the rings are percentiles, lowest 25%, mid 50%, mid upper 75%, top most 100%, wherever the player lands in the data set is where the marker extends to.

So I'm not suggesting badiashille is better than JB, and I literally love JB and would be delighted if he captained us for ten years and lead us to glory.

I'm just saying - you know what CBs are hard to judge, even more so these days than in the past. He's a superb player, but I am sure there are £30m players out there who could step in over a year or so and contribute to the day to day performance of the team in the same manner, and maybe leave us with £30-£40m to add to the bank for The Big Rebuild.

They're hard to judge! Badiashille is basically a meme among Chelsea supporters, but at one time he was one of the most highly rated young CBs in Europe. People are out there man.

[Cods]

I dunno what this bit means sorry. So the rings are percentiles, lowest 25%, mid 50%, mid upper 75%, top most 100%, wherever the player lands in the data set is where the marker extends to.

I think what Matt is saying is that it looks a little misleading with baseline set to 'lowest score'(percentile) rather than 'zero', accentuating the variance which might not be that large in absolute terms. I can see why it's done like that though. (Charts without zero as baseline in general can be viewed as misleading whether intended or not - ).

I know it's probably industry standard too, but I'd be a bit concerned around the data reliability on very small samples particularly with Possession Adjusted stats; adjusting between one of the lowest possession teams (Everton ~ 40%) and one of the highest (Chelsea ~ 60%) at a 1.5 factor the assumption can blow out the results a bit.

I do take your general point though, CBs may be relatively easily replaced without too much upheaval, but with Branthwaite being premium with further to grow we should not be budging until absolute top dollar is on the table.

For the buyer, our rival, 100m is, relatively speaking, often neither here nor there, and they're potentially receiving a vital piece to their puzzle.

[Evertonian in NC]

Waiting a year also alllows us to bring somebody in (presumably younger and from France, etc.) and let them get acclimated to the Prem, before throwing him into the fire next to a newly-baptized JOB and saying "don't fuck it up, lad."

JOB will be a full-time starter for the first time heading into this season (Tarks can hopefully still be #3 after that surgery, though it can be a career killer). Let's not change anything else immediately.

Part of the pre-season focus also needs to be finding out how much potential Myko has at LCB, for when we bring in other options at LB.

[Matt1878]

I would imagine most CBs have fairly high pass %s, so despite JBs 82% rate, he still ranks in the lower 25% of records. That said it must be fairly tightly grouped considering where Badiashille is.



I dunno what this bit means sorry. So the rings are percentiles, lowest 25%, mid 50%, mid upper 75%, top most 100%, wherever the player lands in the data set is where the marker extends to.

So I'm not suggesting badiashille is better than JB, and I literally love JB and would be delighted if he captained us for ten years and lead us to glory.

I'm just saying - you know what CBs are hard to judge, even more so these days than in the past. He's a superb player, but I am sure there are £30m players out there who could step in over a year or so and contribute to the day to day performance of the team in the same manner, and maybe leave us with £30-£40m to add to the bank for The Big Rebuild.

They're hard to judge! Badiashille is basically a meme among Chelsea supporters, but at one time he was one of the most highly rated young CBs in Europe. People are out there man.

Thanks Brap, if the rings are percentiles that makes sense. As you say the grouping for passing must be very dense, if Branthwaite's 82% passing puts him in the bottom 25%. Maybe Dyche bypassing the midfield more dragged him down, never seen him as being a liability on the ball tbh.

As Cods says there will be context to the stats, based on the style of play around a player. Interesting comparison regardless.

[Cods]

Just a bit more context...(general analysis, not suggesting all is new information)

Badiashile was trusted against 11 teams in that 22/23 season; 8 teams from the bottom half, (including every one of the bottom 7 teams) and only 3 teams from the top half, 5th, 6th and 10th. 70% were home games. He played against teams who averaged 42pts, well below the league average of 52.6 that season.

In comparison, Branthwaite effectively missed 9 of our first 10 games in 24/25; our relatively easy (but dreadful) start where he could have padded out his stats against all of the bottom four. The teams he missed averaged 44pts, well below the league average of 52.4pts last season.

I'm guessing the possession adjustments (PAdj) stats would have been taken from the team's average possession across the season, as opposed to the games actually featured in, but if not they potentially could be weighted closer to a factor of 2.0, unfairly skewed further towards Badiashile; Chelsea had 65% v Bournemouth, 66% v Fulham, 68% v Everton, 71% v Hammers, 75% v Forest.

We know the stats are also a function of team style, For passing: whilst Chelsea knocked it around easy at the back and into the middle, we long balled it often from CB,... and out of possession Chelsea were arguably exposed relatively more often at the back as they threw more players forward, allowing for interception opportunities, whilst Everton: largely two solid banks to break through...CB tackles and interceptions therefore understated.

Our opposition don't tend to test our CBs heading ability as they know it doesn't bear fruit.

Blocks and clearances were more our style in the first half of the season, yet Branthwaite played more games under Moyes than Dyche this season.


Further criticism of the radar chart…they also show independent stats as a shaded 'area' which gives the impression of a second dimension to data that simply doesn't have one (the area can also be distorted depending upon which stat sits next to which, a totally arbitrary choice) and a scale range that is inconsistent across each statistic. To the untrained eye they're quite misleading.

When it comes to reliability, these sort of analyses wouldn't pass muster even within other sports (due to inherent complexity), let alone in other industries.

They've even been criticised from their creators, specifically in relation to how effective they are for analysing centre backs, below comments from the fella that designed them for Davie Weir:

These were developed later, and to be perfectly honest, they are less valid overall than the other positional templates....They give you a sense of how a centerback plays, but become tricky beyond that.

All added up, the 'visual' represented by the chart when compared with the underlying numbers and context can paint a very different picture.

[Shogun]

Yes Cods

[Evertonian418]

Just a bit more context...(general analysis, not suggesting all is new information)

Badiashile was trusted against 11 teams in that 22/23 season; 8 teams from the bottom half, (including every one of the bottom 7 teams) and only 3 teams from the top half, 5th, 6th and 10th. 70% were home games. He played against teams who averaged 42pts, well below the league average of 52.6 that season.

In comparison, Branthwaite effectively missed 9 of our first 10 games in 24/25; our relatively easy (but dreadful) start where he could have padded out his stats against all of the bottom four. The teams he missed averaged 44pts, well below the league average of 52.4pts last season.

I'm guessing the possession adjustments (PAdj) stats would have been taken from the team's average possession across the season, as opposed to the games actually featured in, but if not they potentially could be weighted closer to a factor of 2.0, unfairly skewed further towards Badiashile; Chelsea had 65% v Bournemouth, 66% v Fulham, 68% v Everton, 71% v Hammers, 75% v Forest.

We know the stats are also a function of team style, For passing: whilst Chelsea knocked it around easy at the back and into the middle, we long balled it often from CB,... and out of possession Chelsea were arguably exposed relatively more often at the back as they threw more players forward, allowing for interception opportunities, whilst Everton: largely two solid banks to break through...CB tackles and interceptions therefore understated.

Our opposition don't tend to test our CBs heading ability as they know it doesn't bear fruit.

Blocks and clearances were more our style in the first half of the season, yet Branthwaite played more games under Moyes than Dyche this season.


Further criticism of the radar chart…they also show independent stats as a shaded 'area' which gives the impression of a second dimension to data that simply doesn't have one (the area can also be distorted depending upon which stat sits next to which, a totally arbitrary choice) and a scale range that is inconsistent across each statistic. To the untrained eye they're quite misleading.

When it comes to reliability, these sort of analyses wouldn't pass muster even within other sports (due to inherent complexity), let alone in other industries.

They've even been criticised from their creators, specifically in relation to how effective they are for analysing centre backs, below comments from the fella that designed them for Davie Weir:

These were developed later, and to be perfectly honest, they are less valid overall than the other positional templates....They give you a sense of how a centerback plays, but become tricky beyond that.

All added up, the 'visual' represented by the chart when compared with the underlying numbers and context can paint a very different picture.

And breathe.

[brap2]

One thing I would say is the fella who posted it does absolutely know his stuff like, especially when it comes to Chelsea and French football. He's very critical of Chelsea's owners and their business practices, and his point is probably - why would we buy takeaway when we have pizza in the freezer.

Chelsea tend to buy hot new things and abandon the previous one based on the current managers preferences or agent pressures or whatever.

[AndSeel]

Think if he is to go this summer. We should rinse Chelsea. We dont need or want to sell, so go and get your 80 million plus for him

[brap2]

Just a bit more context...(general analysis, not suggesting all is new information)

Badiashile was trusted against 11 teams in that 22/23 season; 8 teams from the bottom half, (including every one of the bottom 7 teams) and only 3 teams from the top half, 5th, 6th and 10th. 70% were home games. He played against teams who averaged 42pts, well below the league average of 52.6 that season.

In comparison, Branthwaite effectively missed 9 of our first 10 games in 24/25; our relatively easy (but dreadful) start where he could have padded out his stats against all of the bottom four. The teams he missed averaged 44pts, well below the league average of 52.4pts last season.

I'm guessing the possession adjustments (PAdj) stats would have been taken from the team's average possession across the season, as opposed to the games actually featured in, but if not they potentially could be weighted closer to a factor of 2.0, unfairly skewed further towards Badiashile; Chelsea had 65% v Bournemouth, 66% v Fulham, 68% v Everton, 71% v Hammers, 75% v Forest.

We know the stats are also a function of team style, For passing: whilst Chelsea knocked it around easy at the back and into the middle, we long balled it often from CB,... and out of possession Chelsea were arguably exposed relatively more often at the back as they threw more players forward, allowing for interception opportunities, whilst Everton: largely two solid banks to break through...CB tackles and interceptions therefore understated.

Our opposition don't tend to test our CBs heading ability as they know it doesn't bear fruit.

Blocks and clearances were more our style in the first half of the season, yet Branthwaite played more games under Moyes than Dyche this season.


Further criticism of the radar chart…they also show independent stats as a shaded 'area' which gives the impression of a second dimension to data that simply doesn't have one (the area can also be distorted depending upon which stat sits next to which, a totally arbitrary choice) and a scale range that is inconsistent across each statistic. To the untrained eye they're quite misleading.

When it comes to reliability, these sort of analyses wouldn't pass muster even within other sports (due to inherent complexity), let alone in other industries.

They've even been criticised from their creators, specifically in relation to how effective they are for analysing centre backs, below comments from the fella that designed them for Davie Weir:

These were developed later, and to be perfectly honest, they are less valid overall than the other positional templates....They give you a sense of how a centerback plays, but become tricky beyond that.

All added up, the 'visual' represented by the chart when compared with the underlying numbers and context can paint a very different picture.

Yeah, they are very hard to judge. I don't think the graph does or intends to, show JB as any less of a player. Just similar players in different systems.

The point being similar template of players are out there and worth the minutes and development rather than jumping to the next one who has had the fortune of playing.

Or for us, similar players are out there if we can find one and bank the change.

[superpull]

Hearing more and more rumours that we're looking at offering him a new deal

[Cods]

Yeah, they are very hard to judge. I don't think the graph does or intends to, show JB as any less of a player. Just similar players in different systems.

The point being similar template of players are out there and worth the minutes and development rather than jumping to the next one who has had the fortune of playing.

Or for us, similar players are out there if we can find one and bank the change.

Yes, agree to an extent.

Just think we're better with him than without him at the moment.

We've been down the route of big sales for big players, Rom, Stones and Rooney come to mind, but have not really improved the squad to any great extent with that cash. We've not really rebuilt, and have paid well above the odds for largely inferior replacements (even excepting the squandered 'Moshiri years').

I think a star player like Branthwaite actually brings out the best of others around him, granted its tougher since we've been depleted of almost all our last remaining quality, I just think if we keep him, and sign another marquee player, then we start to look like we're positioning ourselves differently.