Trying to be too clever

[CanaryLegend 18]

Seems like our scouting based on stats has been the making of our undoing. Probably so obsessed with the likes of Buendia being a success we have lost sight of traditional scouting methods.

Signings like Placheta who is amazingly quick but has no end product and Josh Sargeant who covers lots of ground yet delivers very little else.

I am not sure what the thinking was behind these signings and where the assessment about ability to score goals would from. In Hugill, Idah, and Sargeant we have three strikers with very underwhelming goal statistics.

The soccerbot as well? I mean come on, players will not improve this way. It’s a waste of money.

Accept we are going down as without any signings let’s face it that’s the reality, blood players like Rowe and get us prepared for the Championship with squad hungry to take us forward.

[A Load of Squit 5,128]

Yes, you are.

[Michael Starr 519]

I feel like the soccer bot could have a place and the players seem to enjoy using it, especially good for the u23s squad but yes, it does feel like having a 100% stats based scouting method may have held us back a bit. Sure, a player can cover lots of ground etc, but you wonder what they are prioritising stats wise. We'll say, this is how a Norwich City striker looks, we aren't going to get a complete all rounder so let's get someone who works hard, covers ground, has a good rep, is young, available and within our budget... then you end up with a lower league standard player with the hope that he may develop into a Prem player.

[hertfordyellow 460]

32 minutes ago, CanaryLegend said:

The soccerbot as well? I mean come on, players will not improve this way. It’s a waste of money.

Based on what evidence? The soccerbot isn't replacing traditional training, it is there to enhance it. It simulates real game scenario's and is designed to help improve speed of thought and feet. 

If you are going to make such sweeping comments without a hint of understanding of the subject can you at least say 'I reckon' at the start so we know you are doing that pub bore thing and can instantly reject what you say before having to read the rest of the paragraph?

Also Brentford who are held in such high esteem on this board are heavily into the Moneyball method of player analysis.

[Ulfotto 623]

The club pitch is that it is a club that develops players and makes them better. Whatever level they are when they join.

That is a hard sell if you have tin pot facilities. That is why we have invested in soccerbot and all the mod cons.

Data has a role but it is not the be all and end all you just have to watch the Moneyball movie to know that. I’m sure the club is aware of that.

[Branston Pickle 3,635]

I’m undecided about the soccerbot thing - it doesn’t look like we are reaping huge benefits from it, but then it’s early doors.  
And let’s face it, the cost was large for most industries but it only cost about 3-4weeks of Trippier’s reported salary at Newcastle and even for our ‘poorly’ paid players is barely half a year’s salary of one player.

[Bannana Boy 30]

Not going to pretend to be an expert on this, so this comment is an assumption.

Could the thinking be that by focussing on some key stats the club belives, rightly or wrongly, that it can find players with important attributes all sorted (strength, speed, work-rate etc) and then improve the other attributes (eg being able to hit a cow's harris with a banjo) through training?

Not saying this is the right approach, or the wrong approach, just wondering if this is the thinking.

[Thumbbass 318]

Peter Crouch said on his podcast that at the end of his career there were players that were obsessed with their stats and would deliberately do things to make them look better - on paper. Pass completion being an example is meaningless if you're playing sideways or have played a ball back to a defender who is under pressure. I read someone on here digging at Normann for his pass completion rate but he clearly plays more positive football and is overall a far superior player than our other midfielders.

Stats have a place but can only take you so far down the route of making a decision.

[hogesar 9,618]

15 minutes ago, Branston Pickle said:

I’m undecided about the soccerbot thing - it doesn’t look like we are reaping huge benefits from it, but then it’s early doors.  
And let’s face it, the cost was large for most industries but it only cost about 3-4weeks of Trippier’s reported salary at Newcastle and even for our ‘poorly’ paid players is barely half a year’s salary of one player.

The reality is for the money spent all the soccorbot has to do is make any young potential signings think "Ah, that's cool" - and it'll probably pay for itself.

[hepphep 175]

There are increasing amount of teams that are doing well using metrics heavily in their scouting process. I think in every case (including us) there is still level of manual scouting as part of that process.

Therefore, I don't think it is wrong way to try to do the things. However, as others are doing it actively as well, you need to be good on defining what you are looking for and what kind of data you analyse. Apparently we have not been very good with metrics required for EPL relegation dogfight. Might be that it is position we need to fill even more than the notorious CDM position..

[hertfordyellow 460]

1 minute ago, hogesar said:

The reality is for the money spent all the soccorbot has to do is make any young potential signings think "Ah, that's cool" - and it'll probably pay for itself.

Considering we made 20 million on both Maddison and Godfrey, if it helps one player develop, it more than pays for itself.

[lake district canary 4,531]

Making decisions just on stats is always going to be problematic. Combined with human intuition and knowledge, they are of use, but to just use stats to judge everything is bonkers. Quite often stats are used as justification for decisions made, as in - "why did you buy that player"?..............."well his stats said he would be good".

I would place much more trust in  - "why did you buy that player"......"well I went to see him play twenty times and saw that he had many of the attributes we are looking for and the stats backed that up".  Just to use stats is just a poor substitute for effort, hard work and using your mind.  The brain is capable of so much more - judgement, intuition, creativity - stats are cold, limited and only give a small part of the picture - and that applies to nearly everywhere they are used.

Edited January 14, 2022 by lake district canary

[hertfordyellow 460]

4 minutes ago, lake district canary said:

Making decisions just on stats is always going to be problematic. Combined with human intuition and knowledge, they are of use, but to just use stats to judge everything is bonkers. Quite often stats are used as justification for decisions made, as in - "why did you buy that player"?..............."well his stats said he would be good".

I would place much more trust in  - "why did you buy that player"......"well I went to see him play twenty times and saw that he had many of the attributes we are looking for and the stats backed that up".  Just to use stats is just a poor substitute for effort, hard work and using your mind.  The brain is capable of so much more - judgement, intuition, creativity - stats are cold, limited and only give a small part of the picture - and that applies to nearly everywhere they are used.

I think the statistical data mining point the team to WHO they should scout rather than replacing the scouting side of the process.

[wcorkcanary 4,329]

37 minutes ago, Michael Starr said:

I feel like the soccer bot could have a place and the players seem to enjoy using it, especially good for the u23s squad but yes, it does feel like having a 100% stats based scouting method may have held us back a bit. Sure, a player can cover lots of ground etc, but you wonder what they are prioritising stats wise. We'll say, this is how a Norwich City striker looks, we aren't going to get a complete all rounder so let's get someone who works hard, covers ground, has a good rep, is young, available and within our budget... then you end up with a lower league standard player with the hope that he may develop into a Prem player.

Is our scouting really100% stats based.. no one watches players  ?

Or is this one of those say it enough and it becomes fact pinkun fact. 

[HertsCanary93 222]

It's probably among a host of reasons why I don't run development and coaching at top flight football clubs... But in my opinion, the soccerbot looks like a massive gimmick that wouldn't be out of place at Alton Towers. Especiallyat the price it was rumoured to cost.

Unfortunately with regards to stats driven scouting - under the current model, it is the only way we are able to sign players that might be able to play at this level. Any proven player will either be out of budget or get snapped up by another club.

[hertfordyellow 460]

1 minute ago, wcorkcanary said:

Is our scouting really100% stats based.. no one watches players  ?

Or is this one of those say it enough and it becomes fact pinkun fact. 

No, it's an assumption based on people wanting to bash the club again. We had an interview on this very site with the data analyst lead and it conclusively outlines that the data is used to find WHO to scout not replace scouting

[wcorkcanary 4,329]

3 minutes ago, hertfordyellow said:

No, it's an assumption based on people wanting to bash the club again. We had an interview on this very site with the data analyst lead and it conclusively outlines that the data is used to find WHO to scout not replace scouting

Thought so, though I doubt it'll stop the carp that gets posted on the subject. 

[hogesar 9,618]

9 minutes ago, hertfordyellow said:

No, it's an assumption based on people wanting to bash the club again. We had an interview on this very site with the data analyst lead and it conclusively outlines that the data is used to find WHO to scout not replace scouting

Correct, and it's also not done Brentford too much harm (and a long list of other clubs)

[Mr Angry 1,545]

No need to do any stats based scouting-just use the latest Football Manager database 😉

[Son Ova Gunn 165]

There was a game I think last year when Todd and one of centre backs, 10yrs from any opposition player just started playing tiny one twos. They must of passed it 15 times and I remembered thinking ok we are winning and their time wasting to kill off the game. Occurred to me after seeing the match stats exactly what they were doing and I did wonder at the time, does that really work. Any team looking at a player would surely look through each game played, even if it took 2-3 days to watch the games you would wouldn’t you? So then the scout see’s that a player primarily plays the safe backwards ball or 1yrd sideways option and any false inflation of stats is pointless? 

[hepphep 175]

19 minutes ago, Mr Angry said:

No need to do any stats based scouting-just use the latest Football Manager database 😉

Actually I did read somewhere year or two back that even some top teams actually use FM as one of their scouting tools. As it has so massive database collected from all around the world, it can be one of the things that help finding leads about missed potential targets. It was seen useful especially for find local talent from smaller regions that are not usually worth to scout extensively.

Edited January 14, 2022 by hepphep

[Yellow Wal 314]

But all database compilations are based on opinion and interpretation of the data collector.

If a left winger goes down the line, beats five players and crosses to the striker who is two yards from goal who, in turn, puts it over the bar, that's not an assist.

If a ball is hammered across the box, hits the striker on the **** and it drops to an inrushing midfielder who scores, that's an assist for the striker.

Wonderful things, statistics.

[hertfordyellow 460]

29 minutes ago, Yellow Wal said:

But all database compilations are based on opinion and interpretation of the data collector.

If a left winger goes down the line, beats five players and crosses to the striker who is two yards from goal who, in turn, puts it over the bar, that's not an assist.

If a ball is hammered across the box, hits the striker on the **** and it drops to an inrushing midfielder who scores, that's an assist for the striker.

Wonderful things, statistics.

If you hit the bar with the keeper well beaten it's a shot off target, if you feebly pea role a shot from 30 yards into the Keeper's arms, its a shot on target.

[Ulfotto 623]

You need a database though right. If you want to know all the players under 23 with full international caps. It is a quicker to pull that out of a database than cobble it together some other way. 

[lake district canary 4,531]

14 minutes ago, hertfordyellow said:

45 minutes ago, Yellow Wal said:

But all database compilations are based on opinion and interpretation of the data collector.

If a left winger goes down the line, beats five players and crosses to the striker who is two yards from goal who, in turn, puts it over the bar, that's not an assist.

If a ball is hammered across the box, hits the striker on the **** and it drops to an inrushing midfielder who scores, that's an assist for the striker.

Wonderful things, statistics.

If you hit the bar with the keeper well beaten it's a shot off target, if you feebly pea role a shot from 30 yards into the Keeper's arms, its a shot on target.

These two posts sum it up.  if you want accuracy and context to stats you have to programme more and more parameters in - and then it becomes almost ludicous because there are so many permutations.  So stats should never be taken as the whole picture, but what does anyone ever do when commenting on anything, from coronavirus to football to anything else you can think of?.........."well the stats say" - as if the stats can't be wrong.  Says it all.

[Michael Starr 519]

1 hour ago, wcorkcanary said:

Is our scouting really100% stats based.. no one watches players  ?

Or is this one of those say it enough and it becomes fact pinkun fact. 

It's been confirmed in 2 cases that we signed players based on stats. We couldn't get to see them in person due to the restrictions

[TheGunnShow 5,932]

Stats are like miniskirts; what they show is revealing, but what they hide is f*cking vital!

[wcorkcanary 4,329]

36 minutes ago, Michael Starr said:

It's been confirmed in 2 cases that we signed players based on stats. We couldn't get to see them in person due to the restrictions

Roight, that sounds far more , sensible  and reasonable  than your previous claim that our scouting methods were 100% stats based. ..... can you name the two?  

A man could waste an awful lot of time questioning  the statements  of ' facts ' on this forum. Fortunately  for me I'm too lazy to chase up every single  line of Tosh that  is written  on here. .. but it is the duty of every balanced poster to pull up anything that is, how we say ,' poisonous  seed.'  ..... it is only damaging. 

Would be a great help if anything that was not true could be marked  ' pure conjecture ' ...  to help the gullible, like.

[A Load of Squit 5,128]

28 minutes ago, wcorkcanary said:

Roight, that sounds far more , sensible  and reasonable  than your previous claim that our scouting methods were 100% stats based. ..... can you name the two?  

A man could waste an awful lot of time questioning  the statements  of ' facts ' on this forum. Fortunately  for me I'm too lazy to chase up every single  line of Tosh that  is written  on here. .. but it is the duty of every balanced poster to pull up anything that is, how we say ,' poisonous  seed.'  ..... it is only damaging. 

Would be a great help if anything that was not true could be marked  ' pure conjecture ' ...  to help the gullible, like.

If you look at the Pinkun XT (Expected Truth) stats you'll realise how acurate posts are. 😀

[Jim Smith 2,314]

3 hours ago, Bannana Boy said:

Not going to pretend to be an expert on this, so this comment is an assumption.

Could the thinking be that by focussing on some key stats the club belives, rightly or wrongly, that it can find players with important attributes all sorted (strength, speed, work-rate etc) and then improve the other attributes (eg being able to hit a cow's harris with a banjo) through training?

Not saying this is the right approach, or the wrong approach, just wondering if this is the thinking.

Suspect there is something in that with the thinking being that if they still have development to do in certain areas then they are also cheaper.