Clement

880 posts

Clement in or out

Poll ended at 02 Oct 2018 15:29
Out
35
25%
In
68
49%
Out but in a funny accent
3
2%
Oxf*rd off Maff
17
12%
Ian Royal
16
12%
 
Total votes: 139
Brain Traysers
Member
Posts: 663
Joined: 06 Jul 2004 12:02
Location: Cascadia

Re: Clement

by Brain Traysers » 14 Sep 2018 20:33

A few different bits and pieces to respond to there. Sorry - its another long one.

Denver Royal made some very fair points - unfortunately we don't know how many clubs are at the more advanced end, simply because them making that public knowledge would reduce its edge over the rest of the competition as others would likely follow suit. I think we (and most clubs) will use stats at some level, but from a different angle, probably referred to as 'performance analysis'. From what I understand, this would be more of the sports science angle - looking at which players physical outputs have dropped (e.g. in-game fatigue, impact of injury on short sprint output etc). The fact we have an active position for Women's First Team Performance Analyst suggests we are not completely in the stone age, but the fact is pays £18k suggests they are probably looking for a graduate, rather than an expert (albeit less money in the Women's game). I have seen similar adverts for RFC Academy Performance Analysts, but nothing for first team (I suspect those roles are more frequently head hunted). Fulham advertising for a Data Scientist (more qualified for handing big data, no salary details given), along with their transfer business (excellent signings this summer, plus selling Aluko to us for a laughable fee) suggests they are one of the leading users. Other clubs will also have specific set piece analysts and training analysts to suggest tweak in game tactics, but to be completely honest its a bit of a black box with everyone rightly holding their cards to their chests. This guy posts excellent stuff and is employed by Huddersfield (another team punching above their weight in recent years). Arsenal were an early adopter and paid $250,000 for one year of services from Chicago firm StatDNA in 2010, and when others were sniffing around ended up buying it outright for $4m in 2012. There was a period a few years ago when lots of good data analysts/scientists got hoovered up by clubs and the amount of free/publicly available analysis really suffered. I think we are currently on the cusp of wave 2.

On set pieces - last season we were 16th best with 12 goals from set pieces, against league avg of 13.5 (WhoScored dont break down into different types, so this unfortunately almost certainly includes pens) with Cardiff and Wolves tied first for 21 goals. That said, we only scored 48 goals in total so 25% from set pieces is more mid table (11th best) - Ipswich and QPR stand out with 18 set piece goals each (32% and 31% of their respective totals) which suggests they targeted/exploited it well.

A few people mentioned Aluko and Kelly's deliveries not beating the first man - while a very real frustration I think it indicates we are not at the forefront. From what I have previously read on set piece analytics (not much beyond things already linked...), this is really common as teams are conscious of the counter attacking threat - at any professional level nearly all players are good enough to consistency deliver a ball beyond the first man into the right areas, but to avoid making it easy to clear/ be gathered by the GK who starts a counter, they have to try to whip it in to a much smaller area which is often then cleared by the first man. The more advanced coaches are seemingly very aware of this, and instead target areas easier to reach, but harder for the GK to get to - e.g. Trippers excellent World Cup deliveries were often towards the penalty spot rather than 6 yard box, with 2-3 targets there and Kane sniffing around the 6 yard box for any follow ups - that's a very very basic summary, there was a lot more too it e.g.. But I accept that doesn't necessarily mean Aluko/Kelly they are good set piece takers.

By coincidence, I used to work with the girl who was in charge of Prozone analysis at RFC right from when that was the new big thing (mid 2000s?), although I worked with her in a restaurant while in sixth form rather than at the club! I hadn't seen anything about her for a while, but a quick google tells me she left her role with us as head of Technical Scouting in 2015 to join Atlanta United in the US (currently top of their league and favorites to win the MLS cup) as the Head of Video and Technical Analysis. Good on her. We hired Tony Spearing in Feb 2015 (possibly her replacement?), but Wiki tells me Norwich appointed him in April 2016, and I cant see a similar direct replacement press release - we currently have Ed Franklin as 'Head of Elite Performance' but that seems a much broader role as described on LinkedIn, and his background is very Sports Science focused. We still have a few scouting analysts with data skills such as Michal Opilka who seem to be active (very active in Scotland), but that looks a much more junior level with an obvious recruitment focus.

Separately I might start a separate thread for periodic updates on RFC related stats as the season goes on (that way the h8rz can choose to continue to ignore). I'm relatively new to devoting time into this - I'm trying to improve my data science skill set and football data seems a nice way to do it, with an active community of observers and a tonne of rich datasets.

User avatar
Zip
Hob Nob Legend
Posts: 22408
Joined: 30 Dec 2017 16:39

Re: Clement

by Zip » 14 Sep 2018 20:42

No reason why you shouldn’t start up a separate thread regarding Reading FC related data. Ignore the h8ers though.

Hound
Hob Nob Legend
Posts: 24934
Joined: 27 Sep 2016 22:16
Location: Simpleton

Re: Clement

by Hound » 14 Sep 2018 20:58

Sadly I really enjoy the stats posts

Did do it for A level so maybe related

Not too sure about the first man bit though - the number of times teams have countered us from just that suggests our coaches aren’t either

Snowball
Hob Nob Legend
Posts: 20679
Joined: 02 Jan 2009 18:35

Re: Clement

by Snowball » 14 Sep 2018 21:20

I notice with England that teams never seem to cover the DEEP corner
and Harry Maguire got a lot of chances or room to head back across
goal. Didn't we used to do similar with Pearce heading the ball back?

User avatar
Zip
Hob Nob Legend
Posts: 22408
Joined: 30 Dec 2017 16:39

Re: Clement

by Zip » 14 Sep 2018 21:43

Snowball I notice with England that teams never seem to cover the DEEP corner
and Harry Maguire got a lot of chances or room to head back across
goal. Didn't we used to do similar with Pearce heading the ball back?


We don’t cover the deep corners leaving opposition players umarked


User avatar
BR0B0T
Hob Nob Super-Addict
Posts: 15325
Joined: 08 Nov 2016 23:25

Re: Clement

by BR0B0T » 15 Sep 2018 01:28

Brain Traysers I'm relatively new to devoting time into this - I'm trying to improve my data science skill set and football data seems a nice way to do it, with an active community of observers and a tonne of rich datasets.


What Data Science stuff do you do? Assume you can program etc

SCIAG
Hob Nob Addict
Posts: 6362
Joined: 17 Jun 2008 17:43
Location: Liburd for England

Re: Clement

by SCIAG » 15 Sep 2018 08:44

A few people mentioned Aluko and Kelly's deliveries not beating the first man - while a very real frustration I think it indicates we are not at the forefront. From what I have previously read on set piece analytics (not much beyond things already linked...), this is really common as teams are conscious of the counter attacking threat - at any professional level nearly all players are good enough to consistency deliver a ball beyond the first man into the right areas, but to avoid making it easy to clear/ be gathered by the GK who starts a counter, they have to try to whip it in to a much smaller area which is often then cleared by the first man. The more advanced coaches are seemingly very aware of this, and instead target areas easier to reach, but harder for the GK to get to - e.g. Trippers excellent World Cup deliveries were often towards the penalty spot rather than 6 yard box, with 2-3 targets there and Kane sniffing around the 6 yard box for any follow ups - that's a very very basic summary, there was a lot more too it e.g.. But I accept that doesn't necessarily mean Aluko/Kelly they are good set piece takers.


About ten years ago the trend was for low, flat deliveries into the near post. These would often get cleared by the first man, but the pace on the ball meant that contact would often result in a goal - at least, that was the theory. Stephen Hunt used to get an awful lot of stick for these deliveries. He wasn't nearly as gifted as the set piece takers we'd had before him (Shorey, Convey, Little) or for that matter Kelly and Aluko. But eventually this approach paid dividends and we had one of our most successful spells at scoring set piece goals, particularly through Bikey.

Kelly and Aluko are both good takers. We know that because of how they've done at other teams. I think the main reason things aren't working too well for them is the lack of targets - our centre backs aren't very good at attacking corners but Moore and McShane have improved since joining us, and our current crop of forwards aren't much good at attacking crosses either. We can work on that.

I'm very pro-stats but I think most armchair statisticians are terrible at using them, which I think contributes to the collective groaning when someone uses "expected goals" to illustrate a point.

User avatar
Sutekh
Hob Nob Super-Addict
Posts: 18386
Joined: 12 Feb 2014 14:05
Location: Undiscovered pyramid somewhere in Egypt

Re: Clement

by Sutekh » 15 Sep 2018 10:36

You would still expect them to beat the first man even if the rest of the team was failing in their part. This is something they both regularly struggle to do. Perhaps it’s because they’re low on confidence and/or trying to hard but either way there’s room for considerable improvement.

Best corners I’ve seen are those where the ball is dropped to the edge of the D for someone to hit a clean volley that no one can stop. Very difficult to pull off but you’d think it’s worth a try once a game esp. when the standard ball over is continually not getting anywhere.

And that’s another point, Reading players don’t shoot enough. Too much time is spent diddling about looking for the space that never opens before losing the ball.

User avatar
Snowflake Royal
Hob Nob Legend
Posts: 39398
Joined: 20 Jun 2017 17:51

Re: Clement

by Snowflake Royal » 15 Sep 2018 13:00

Sutekh You would still expect them to beat the first man even if the rest of the team was failing in their part. This is something they both regularly struggle to do. Perhaps it’s because they’re low on confidence and/or trying to hard but either way there’s room for considerable improvement.

Best corners I’ve seen are those where the ball is dropped to the edge of the D for someone to hit a clean volley that no one can stop. Very difficult to pull off but you’d think it’s worth a try once a game esp. when the standard ball over is continually not getting anywhere.

And that’s another point, Reading players don’t shoot enough. Too much time is spent diddling about looking for the space that never opens before losing the ball.


Failing to beat the first man is a consequence of the small margins involved in the type of set piece delivery they're going for though.

You can deliver a floated ball to the back post to be contested pretty much every time. But your chances of an attacker winning it and getting enough on it to score are slim. Or you can try the harder flat whipped ball where any touch can see it fly in the net, but you risk hitting the first man and it not even being contested.

People get overly frustrated with the failed first man corner. Give me an attempt at that sort of corner over a floated one to the back post most tries. We do need to get better, but we need to get better at absolutely everything, so singling corners out seems pointless.

+1 on the stats interest by the way... obviously shouldn't be taken too seriously or as proof on their own. But useful. BT's posts are better than most on here.


User avatar
Zip
Hob Nob Legend
Posts: 22408
Joined: 30 Dec 2017 16:39

Re: Clement

by Zip » 15 Sep 2018 14:14

If we lose today Clement can expect a lot of criticism. I can’t believe Aluko starts and that Kelly and Swift both start. Our only player who has scored more than one goal is benched. :|

Hound
Hob Nob Legend
Posts: 24934
Joined: 27 Sep 2016 22:16
Location: Simpleton

Re: Clement

by Hound » 15 Sep 2018 14:33

Do we know Aluko and Kelly are good set piece takers? Don’t remember Aluko ever taking them for previous teams particularly (not saying he didn’t but certainly wasn’t famed for them). Kelly has only played for us at any sort of level

User avatar
John Smith
Hob Nob Regular
Posts: 4717
Joined: 20 Jan 2010 23:47
Location: Astronauts The New Conquistadors

Re: Clement

by John Smith » 17 Sep 2018 09:24

Cape Town Royal I feel we really needed to have won the Blackburn game, 2 up away from home and largely in control, that was a game where had we seen it out would have given the team a huge lift. The result and the nature of how it happened was a large kick in the teeth...

Boy do we need an ugly win or two....

The only reason we didn't win that game was Paul McShane. There were no statistics to save us that night.

Hound
Hob Nob Legend
Posts: 24934
Joined: 27 Sep 2016 22:16
Location: Simpleton

Re: Clement

by Hound » 17 Sep 2018 09:57

John Smith
Cape Town Royal I feel we really needed to have won the Blackburn game, 2 up away from home and largely in control, that was a game where had we seen it out would have given the team a huge lift. The result and the nature of how it happened was a large kick in the teeth...

Boy do we need an ugly win or two....

The only reason we didn't win that game was Paul McShane. There were no statistics to save us that night.


think the most interesting stat from that game was their only 2 shots on target in the second half were the penalties.

4 points from the next 2 games and I'd be absolutely delighted


Snowball
Hob Nob Legend
Posts: 20679
Joined: 02 Jan 2009 18:35

Re: Clement

by Snowball » 17 Sep 2018 10:03

If flat corners keep finding a defender at the near post
why don't we have an attacker standing in front of him?

And if the attacker moves in front of the RDG player
he's now dragged out of position.

User avatar
John Smith
Hob Nob Regular
Posts: 4717
Joined: 20 Jan 2010 23:47
Location: Astronauts The New Conquistadors

Re: Clement

by John Smith » 17 Sep 2018 10:08

Snowball If flat corners keep finding a defender at the near post
why don't we have an attacker standing in front of him?

And if the attacker moves in front of the RDG player
he's now dragged out of position.

Brian Howard?

User avatar
NewCorkSeth
Hob Nob Addict
Posts: 9519
Joined: 05 Jul 2013 00:17
Location: Wherever Nameless may be.

Re: Clement

by NewCorkSeth » 17 Sep 2018 10:17

I am unable to back this up in any way but I believe I read that short corners (as in not into the box) result in more goals that corners crossed deep into the box.

Corners and goal kicks both wind me up a little as in both cases pumping it long often lead to either a counter attack or sustained possession for the other team a large amount of the time.

Its one of the reasons I prefer a team to play it out from the back. It also lessens the amount of head tennis I have to watch which I really hate..

Stranded
Hob Nob Subscriber
Hob Nob Subscriber
Posts: 19584
Joined: 14 Apr 2004 12:42
Location: Propping up the bar in the Nags

Re: Clement

by Stranded » 17 Sep 2018 10:34

NewCorkSeth I am unable to back this up in any way but I believe I read that short corners (as in not into the box) result in more goals that corners crossed deep into the box.

Corners and goal kicks both wind me up a little as in both cases pumping it long often lead to either a counter attack or sustained possession for the other team a large amount of the time.

Its one of the reasons I prefer a team to play it out from the back. It also lessens the amount of head tennis I have to watch which I really hate..


Corners are statistically awful when it comes to goal scoring chances, which makes sense when you think about it (ball has to go away from goal to give the team a chance and the defending team are virtually all back). A good short corner, should increase your chances. Take our 2nd on Saturday for example, short ball pulls 2 defenders out, quick ball releases Aluko and suddenly you are pulling back from a better position with more fluid movement in the box.

If used properly, short corners can be very dangerous.

User avatar
NewCorkSeth
Hob Nob Addict
Posts: 9519
Joined: 05 Jul 2013 00:17
Location: Wherever Nameless may be.

Re: Clement

by NewCorkSeth » 17 Sep 2018 10:41

Stranded
NewCorkSeth I am unable to back this up in any way but I believe I read that short corners (as in not into the box) result in more goals that corners crossed deep into the box.

Corners and goal kicks both wind me up a little as in both cases pumping it long often lead to either a counter attack or sustained possession for the other team a large amount of the time.

Its one of the reasons I prefer a team to play it out from the back. It also lessens the amount of head tennis I have to watch which I really hate..


Corners are statistically awful when it comes to goal scoring chances, which makes sense when you think about it (ball has to go away from goal to give the team a chance and the defending team are virtually all back). A good short corner, should increase your chances. Take our 2nd on Saturday for example, short ball pulls 2 defenders out, quick ball releases Aluko and suddenly you are pulling back from a better position with more fluid movement in the box.

If used properly, short corners can be very dangerous.

I think (I may be well off mark here) that more goals are scored from corners than direct free kicks (when averaged out obviously)

Our corner on Saturday for the 2nd goal was lovely but won't work every time which is why Bacuna, who has already shown to be better at corners than Aluko, should be in charge of corner duty. If swift comes close every time then hell draw a man every time and Bacuna can choose to go short or long.

The fact that we don't have a left footed player taking the right hand side corners also baffles me. In swinging corners are waaay more effective than out swinging ones.

User avatar
John Smith
Hob Nob Regular
Posts: 4717
Joined: 20 Jan 2010 23:47
Location: Astronauts The New Conquistadors

Re: Clement

by John Smith » 17 Sep 2018 11:14

SCIAG ...eventually this approach paid dividends and we had one of our most successful spells at scoring set piece goals, particularly through Bikey.

Who scored 6 goals in 3 seasons for us. What are you on about? Stop guessing to try and sound clever.

BraisingsteakRoyal
Hob Nob Regular
Posts: 1955
Joined: 24 Mar 2011 18:33

Re: Clement

by BraisingsteakRoyal » 17 Sep 2018 11:49

Regarding the flat corner / hitting the first man - i used to remember that Ian Harte used to do that 2/3 times a game (to groans from half the ground) before the next one allowed Pearce/Morrison to score.

I don't have the stats to hand, but I feel like it resulted in thousands of goals.

880 posts

Who is online

Users browsing this forum: WestYorksRoyal and 440 guests

It is currently 28 Mar 2024 16:07