[FM22] Statman and Robins

[Shrewnaldo]

Having just come off the back of an, ultimately underwhelming, club and country save in which I focused primarily on youth recruitment, I wanted to go with something very different for my next job. Nevertheless, I still wanted to introduce some additional challenge. As we were avoiding youth development, we would obviously be concentrating on recruitment and what better way to make that recruitment more difficult than by removing the numerical attributes?

Now I haven't gone completely 'attributeless'. I don't think it's really feasible, or realistic, to remove them in their entirety. Of course you're still going to have some sort of idea how good an individual player is at a particular activity, that's just logic - but in 5% increments? Probably not. So instead, I will be using a skin which grades the attributes in four levels, which I have set as:

• <8       - Poor         - Grey
• 8-11   - Average   - Silver
• 12-15 - Good       - Gold
• >15    - Elite         - Green

(I may change the colours in future)

The skin is a thing of beauty and is an amalgamation of the FM Rensie skin (found here) and the Star Attributes skin (found here). Huge thanks to FM Rensie who did all the legwork in combining the two and to Sebastien for the star attribute panels. I, of course, asked for their permission first.

The skin is amazing and removes all references to numerical attributes - from player profiles, scouting cards, squad views, etc etc. All gone. With luck, this will force me to utilise the statistics that are available in-game - both for squad decisions and, most relevantly, for recruitment. A more statistical focus is something which I've long wanted to do in FM but which I inevitably ditch in favour of the much simpler and more reliable tool of the attributes. When I tried this same idea back on FM20 with Académico de Viseu (Portugal), it really did work and I had a great save - so here's hoping that history will repeat itself.

I was hoping that the Data Hub would make this sort of save more viable but I've been a little disappointed by the new feature - primarily because you can't use it for recruitment purposes other than for players in your own league. Hopefully this is simply its first iteration and subsequent FMs will expand its functionality to adding players from your shortlists, or other selected leagues. This, and including a much wider range of statistics - plus making the statistics more reliable - would be the absolute ideal for me.

To bring this save even further from my standard game, there are two more key changes:

1. Youth recruitment at the club will be totally ignored. I have already released all the staff from the u-18s team and have, so far unsuccessfully, asked the board to cut back on youth recruitment et al. I can't quite 'do a Brentford' and kill off the newgen intake completely but the plan will be to run with purely a first-team squad of around 25 players by the end of season 2
2. I usually avoid saves in England like the plague. The easy money in the Prem drives me crazy but, given this will likely be only a 5-6 season save, I figured why not?

So which club have I gone with? Well the second half of the "Statman and Robins" title might give it away but I wanted to go no lower than the second tier but try to avoid former Premier League clubs. And my daughter was very much taken with the robin on the badge so it'll be...

... looking forward to it. Hopefully it'll be an interesting read.

[SteinkelssonFM]

A club close to home for me paired with the specific recruitment methodology means I for one will be following along. 

Hopefully you can bring City the success which they and the City of Bristol deserve. 

[Shrewnaldo]

47 minutes ago, MattyLewis11 said:

A club close to home for me paired with the specific recruitment methodology means I for one will be following along. 

Hopefully you can bring City the success which they and the City of Bristol deserve. 

Thanks Matty. I may or may not be taking a long look at your Monchi thread tomorrow *insert eyes emoji thing*

[rich ruzzian]

Interesting idea with the colours instead of the normal number. I will be following. 

[Bristol CityFC1981]

Shall follow! I can never do well with City, think it’s cos I’m too loyal to the current mediocre players haha

good luck 

[Shrewnaldo]

9 hours ago, rich ruzzian said:

Interesting idea with the colours instead of the normal number. I will be following. 

 

8 hours ago, Bristol CityFC1981 said:

Shall follow! I can never do well with City, think it’s cos I’m too loyal to the current mediocre players haha

good luck 

Thanks both. Hopefully I can do justice to the club, as well as provide an interesting thread.

[Shrewnaldo]

#1 - State of the club and early planning

As I said in my earlier Perú and Cantolao thread, planning is one of my favourite parts of Football Manager. This is likely, in no small part, because it's part of the fantasy element of the game and how you convince yourself you could run a real club, if only circumstances were different. It also appeals to my personality - preferring logic, reason and evidence to emotion or gut feel. Which is probably why I prefer strategy games over action and single player over online. It's only logical that this would extend to how I play a game that occupies vast swathes of my free time.

To plan, however, you need to know two things:

1. Where are you starting from?
2. Where do you want to end up?

Simple, right?

Question 1 is relatively straightforward. Bristol City are a Championship club from the South West of England with Bristol being 'famous' as the largest English city never to have had a representative in the Premier League. City have recently been the most likely of the city's two clubs, the others being Rovers, to make it to the big time, spending considerable time in the second tier and even making it to an unsuccessful playoff final in 2008 but have regressed recently.

They've recently completed construction of the Robins High Performance Centre, providing the club in-game with excellent training facilities; and play at the 27,000 capacity Ashton Gate.

And that's about it, really. There isn't much of a trophy cabinet to speak of - a second tier title in 1906, an Anglo-Scottish Cup when that was still a thing and, most recently, the 2015 Papa John's Trophy. But a decent starting point and certainly better than my usual saves start with. Not to mention some rather fetching kits. I'm a particular fan of the away shirt.

So then we come to Question 2 - where do we want to get to? I guess the initial answer is obvious - the Premier League, and from there Europe. Competitive success is simply football's raison d'être so I'd rather treat the question as the structural destination that we hope will bring that success. As mentioned in the opening post, that structure will hopefully look much, much different to my usual saves. Like most FMers, I'm typically a fan of developing youth players. In 99% of my saves, I sign only domestic players, mostly under 20, make a transfer profit every year, never spend more than £10m on an individual player and pump cash into the youth system. Most of those guidelines will not apply to this save.

Instead, we will jettison all youth development, empty the u-23 and u-18 squads and operate a system that utilises a senior team only. This senior squad will have 25-30 players, aiming for a core of players aged between 25 and 30 and will concentrate much more on compatible personalities than passports. As for the transfer profit and transfer fee ceiling... well I'm still Scottish so those might hang around.

The initial target, though, will be for the two youth sides to be empty of players by the end of season 2. Anyone good enough will be promoted to the senior squad and anyone not, well we know what happens with them. I'll keep asking the board to reduce the youth level, cut back the youth recruitment and spend less on junior coaching but I suspect it will take some time to get there as, whilst it is only listed as 'favoured' the board, like just about every other sodding team in FM, wants a culture of developing players using the youth system. At Cantolao, we were spending around £4m a year on the youth setup alone and we'll be looking to keep it well below that here. I've already laid off all the u-18 coaches and Head of Youth Development to save on staff wages, but will keep the u-23 coaches around as they also work with the senior squad and so provide much-needed additional staff capacity.

Conversely to the u-18 staff, I'll be looking to increase the numbers on the recruitment side as much as possible. Having, as always, blocked the first transfer window for the save, we have until the January window to get the initial scouting system up and running but, more likely, primarily aiming at summer 2022 for the first major moves. I'd also really like to make use of the Recruitment Analysts but, to be honest, I can't see how I can do anything with them other than the very bland Analyst Report that is produced - and I can't even see how to specifically request that specific analysts report on specific players... so any tips there would be welcomed.

I'll do a separate post on the scouting and recruitment strategy, as it's likely to be a key part of the save and I want to really get into a lot of detail on this area. For the moment, I'm having a hard enough time analysing my own squad and which roles each player is most suited to. Assessing this, for example, is much more challenging than running the rule over the numbers.

Challenging but, it has to be said, thus far a whole lot of fun.

[BadManager]

Interesting idea to abandon the youth. I wonder if that would cost if it was more than a 5-6 season save.

Probably not. As you say, you should have plenty of money if premier league success comes your way.

[tomblogsfm]

Very interesting idea, I hope it works out for you. I always find it surprising how much teams spend on their youth systems, with no or little success to come from it.

I'm looking forward to reading more!

[Shrewnaldo]

14 hours ago, BadManager said:

Interesting idea to abandon the youth. I wonder if that would cost if it was more than a 5-6 season save.

Probably not. As you say, you should have plenty of money if premier league success comes your way.

Aye, I'm not sure how many academy players end up coming through and playing first team football for Premier League clubs, not that many really. Let's say, conservatively, that a decent youth setup is going to cost us £5m a year, plus wages for the staff at, let's say, £2k a week for 5 staff - that's another £0.5m a year. Plus the player wages for the whole youth squad, not just the one or two who will make it into the first team - say 40 players at an average of £1.5k a week? That's another £3m a year. So £8.5m a year. conservatively, for a system that might bring through a player or two that'd be good enough for first team football. I reckon a smart recruitment system can absolutely beat that average and for a much lower cost.

4 hours ago, tomblogsfm said:

Very interesting idea, I hope it works out for you. I always find it surprising how much teams spend on their youth systems, with no or little success to come from it.

I'm looking forward to reading more!

Thanks. As above, I reckon it could be a much cheaper and more reliable system this way but we'll see.

[Shrewnaldo]

#2 - Recruitment Strategy, Part 1

The title might be a little pompous for a computer game but I do like the idea of having a more structured, logical and proceduralised method of bringing players into the club. Many of you will have read @MattyLewis11's blog post on the Monchi Masterclass, and as Monchi makes clear in the source material for Matty's blog, the aim of a recruitment strategy is to take as much luck as possible out of buying players. And that is most effectively done by applying what is, in truth, pretty straightforward logic in a structured manner. In a typical FM save, I rarely have to consider much of a strategy - the sole point of scouting is to reveal players' attributes and then I can make a decision based on some very specific and precise parameters. Having removed the numerical attributes from my game, that path is now closed to us and I need to build an alternative strategy for assessing potential signings in an effort to minimise that luck factor.

Both Matty's blog and the Monchi YouTube series are excellent and I would recommend you take some time to consume both. Neither are saying "this is the only, or even the best, way to do it", but they offer great insight from a hugely successful Director of Football on real life recruitment, and then how that could be implemented within FM. This post will outline how I intend to take those lessons into my save.

As in my previous post, planning is all about knowing where you are, where you want to be and then how you intend to get there. In what might seem like a vastly illogical step, I'd like to deal with the last of that triumvirate first here - the how. This is because both the where we are, in terms of playing personnel, and therefore what we need to get, i.e. areas we need to improve, require a bit of assessment over the course of our first season in charge. Whilst we do that, we can start the planning on the how. (note, I left the numbers visible for staff attributes as there is no other way of gauging staff's effectiveness)

The first thing to note is that our Recruitment Team is a bit limited for now. I've cleared out the deadwood and brought in, via the Job Centre, a decent team and am still on the lookout for a suitable Director of Football. Noting that we're still a Championship side and work permits for overseas players are likely to be difficult for the first couple of seasons, I've concentrated on British scouts with the exception of Vicente. Over time, expanding our scouting knowledge will be important, with two key paths to doing so - first through staff, both in terms of numbers and the individual knowledge they bring, and secondly through affiliate clubs. We've already started with the latter, setting up an agreement with Slavia Sofia that automatically brings full scouting knowledge of Bulgaria.

But for now, we'll be concentrating on the domestic market and we'll do that in two steps:

1. Gather as much data as possible
2. Filter that data through a selection process until top targets are identified

This will be a repeatable process where, in quarters 3 and 4 of any given year, all the scouts will be sent out on standard scouting assignments - to specific countries or nations - and left to identify as many players as possible. At Christmas, those assignments will end and, in Q1 and Q2 of the following year, scouting will instead consist of watching specific players at actual matches. 

In season 1, this means our scouts will be distributed as follows up to Hogmanay 2021:

• 1 scout assigned to Next Opposition reports (this scout requires Tactical Knowledge, as well as JPA)
• 3 scouts assigned to scouting English tournaments:
  • 1 to Premier League and Championship
  • 1 to League 1, League 2 and the National League
  • 1 to the Reserves / Under 23s and Under 18s - although we aren't looking to pick up young players, we are looking for discarded squad players from the big sides
• 1 scout assigned to the rest of the UK

Once more scouts are available, we'll initially send them to three specific European markets where we can try to find value - Austria, Bulgaria and Denmark.

This initial phase is deliberately crude. We're not looking for specific positions, we're just casting a wide net to gather as much information as possible. That data will then be fed into a funnel in Q1 and Q2 of the following year, preparing for the summer transfer window. I will manually ask individual scouts to go watch specific players for 3 matches at a time. Those reports will then dictate whether or not the player continues through the funnel. It's not ground-breaking or complicated, but I've set up 4 separate shortlists like this:

Simple really. The key step is from Tier 2 to Top Targets. Up until Tier 2, I won't filter out anyone by position, role or personality - as both the player and my requirements may change over time and these shortlists can be revisited in subsequent seasons so that we don't always have to start from Step 1. But to make it to the 'Top Targets', a player will need to meet 5 pre-requisites:

• Recommended B or above by multiple scouts
• Has been watched by me for at least 2 games
• Suitable personality
• Meet the statistical requirement for his role's profile
• Be a priority role for the next transfer window

The first three of these are self-explanatory, the last two will be subject to another post detailing the strategy further. 

And that's about it really. It's not complicated, it's just common sense and is about building up a databank that will facilitate the specific selection of players once we're in a position to identify specifically what we need.

 

 

[extudassex2]

Really interesting approach, one that'll be following close!

[toygungoktas]

If it is okay, can you share the skin? I like the Rensie skin but also want to try attributeless.

[Shrewnaldo]

10 hours ago, extudassex2 said:

Really interesting approach, one that'll be following close!

Thanks, hope I can keep it interesting 

3 hours ago, toygungoktas said:

If it is okay, can you share the skin? I like the Rensie skin but also want to try attributeless.

I can't, I'm afraid. When I asked permission of the skin developers, I said that I wouldn't then share it on further. If you're interested, probably best to ask them directly? 

[_Ben_]

A really interesting concept @Shrewnaldo. Not the kind of thing I'm used to reading from you but that is the thing that probably makes it even better! I remember you toyed with a skin like this in Portugal a couple of years ago? I also played an entire save with Guadalajara using it.

[Shrewnaldo]

2 hours ago, _Ben_ said:

A really interesting concept @Shrewnaldo. Not the kind of thing I'm used to reading from you but that is the thing that probably makes it even better! I remember you toyed with a skin like this in Portugal a couple of years ago? I also played an entire save with Guadalajara using it.

Cheers Ben. I did, yes, well remembered. With Academico de Viseu. It was pretty good fun but I never went as hard into the stats as I wanted then or am trying to now. I remember switching back to the regular skin after we won the Portuguese title and being shocked by how poor some of my favourite players' attributes were. Fun way to play though

How did you find it? Guadalajara only use players from their own area, don't they? Similar to Bilbao? Any thoughts or tips? 

[Shrewnaldo]

#3 Robins around the Christmas Tree

So the next step of the recruitment strategy will be to identify what we need to procure. And to do that we need to compare what we've got with what we want to end up with. And to know what we want to end up with, we need to know the system that we want to play in. And that leads me to the titular pun.

Again in an effort to make the save a bit different from other saves, I wanted to go with a formation that I haven't used in a while. And who doesn't love a #10, so why not use two? Hence, the 4-3-2-1 or, as he it has become known, the Christmas tree. For those who don't know, it's typically shown in this sort of shape.

Pre-ordaining a tactical system would once have been anathema to me, instead I'd have tried to build a system that utilised the strongest talents in our squad and then go from there. Luckily enough here, the starting squad is relatively well suited to the 4-2-3-1, not having much in the way of a winger and some strength in depth through the middle. Besides, with the board having relatively modest expectations (top half), we have a little bit of time for trial and error.

But 4-3-2-1 is just a bunch of numbers and we need to know a little more about each role in order to understand whether it's working with the personnel available and, if not, what we need in a new signing to become more successful. So, back to front, here's what I think each role should entail and, to provide a very crude profile by which to baseline the sort of player I'm aiming for, I'll give an example of top player that I think would be ideal for the role.

I would start from the back and work forward but I don't think there's anything specific about the 'keeper and centre halves - so we'll just skip over those and accept that standard attributes and metrics apply to those roles in terms of getting the best you can. Assume Oblak, van Dijk and Alaba.

Fullbacks

It doesn't take a genius to look at the team shape and understand that the fullbacks are going to be responsible for an entire flank each. They will need to cover an awful lot of ground getting up and down the park, so in FM terms we're looking at a Wingback on support or, more likely, attack duty. Stamina to maintain fitness through the match, and Natural Fitness to maximise recovery speed between games are obvious priorities. As they need to stay wide, any traits that include coming inside are undesirable and instead we should be looking for something like Hugs Line or Dribble down x flank.

I don't know what good or bad looks like yet, but the stats we'll have to consider will likely be:

• distance covered
• crossing success
• dribbles per game
• tackle success ratio
• interceptions

The classic example would be Cafú who provided the entire right flank for Carlo Ancelotti's Milan when the xmas tree was employed and any number of wingbacks exist these days who would be able to do a job - it's more of a Andy Robertson than a Trent Alexander-Arnold style of wingback, but then you'd probably want to go with the better Scottish left back and take Tierney over Robertson. But you get the idea. Defensive ability is important but probably not quite as much as sheer physicality and providing an attacking outlet through their crossing ability.

Midfield three

With the wingbacks pushing on very aggressively on the flanks, the obvious role for the midfield three is to provide some measure of cover and defensive stability. The two wider midfielders should be able to support the flanks - helping the wingbacks out in possession, forming triangles with the 10 on their side and also preventing counter attacks down their side. In the defensive phase, I'd expect these wide midfielders to push across to help their wingback out - particularly against teams who play two on each flank and look to overlap with their fullbacks.

So, similarly to the wingbacks, we're looking at the wider midfielders in particular facing a lot of physical strain, requiring high Stamina and Natural Fitness, not to mention Teamwork and Work Rate. In terms of roles, I reckon the 'meta' would be to use mezzalas here and overload the opposition's defence with runs from deep but that feels a little too much like "gaming the match engine" to me and I'd rather go with what I consider to be a realistic and logical set-up. As such, the Carrilero role looks ideal.

On the basis that we're aiming for this 2-3-5 shape in possession, I simply want that midfield three to control the game and dominate the space. Attacking attributes are less of a priority than physical and defensive attributes, but mentals are probably the key here. I've already mentioned Teamwork and Work Rate, but Concentration, Composure and Decisions at the very least. In terms of player templates, I'm thinking of the likes of N'Golo Kanté, John McGinn or James Ward-Prowse. Endless balls of energy, constant closing down but no small amount of technical ability.

Important stats for these players will likely be:

• pass completion
• tackle success ratio
• interceptions made
• possessions lost and possessions gained (sadly only available in match stats, I think?)
• distance covered per 90

With the carrileros providing the energy, that allows the central of the three midfielders to dictate play. In Ancelotti's xmas tree, this role was famously performed by Andrea Pirlo and his sumptuous beard. I'm not sure that I'd like this player to be the main creative outlet, but I do foresee him being the player with the most passes - perhaps performing more of a Busquets role in controlling the game from deep whilst the 10s further forward provide the final creativity. I do, though, want the MC to be able to play those raking passes to the flanks, so suitable player traits along with the technical capability certainly makes sense.

Initially I'd gone with a DLP(D) for this role but found that this provided a little too much positional freedom and I very much want, in the starting formation, the central midfielder to stay within 20 yards of the centre circle so have settled on a CM(D) which appears to be working well.

Important stats for the central midfielder:

• passes per 90
• pass completion
• key passes per 90
• headers won % (due to being on the end of goal kicks etc)
• possessions lost

One alternative set-up, and noting that we don't have a lot of midfield movement between the lines, is to switch the wide two into ball-winning midfielders on defend and then change the central player to a midfield runner, i.e. CM(A). This would be my back-up plan if we were chasing a game late on and needed more presence further forward.

Front three

And that leaves the two 10s and the 9. Initially my plan here was to use a traditional number 9, penalty box poacher scoring the majority of the goals whilst pairing a creator and a runner as the two 10s behind him. But then I also had a thought about a creative 9 who drops off the line and leaves space for the two 10s to burst through - quick one-twos and drawing defenders out of position. As usual, it's in the final third where I am most indecisive.

What I've decided on is to leave the specific roles fluid. It makes sense that this front three will be the most reactive part of the system. If the opposition plays a high line and looks to squeeze the midfield, then it makes sense for us to look for a 9 that plays on the shoulder, looking to exploit the space behind. If the opposition sits deep and compact, then it makes sense for us to utilise the wingbacks as much as possible and utilise some aerial strength in the middle. Or we could look to create space with a false 9 or similar, who looks to disrupt that compact defence with his movement and create space for runners at 10.

This produces a bit of a clash with the recruitment system, though, as we need specifics for the profile of players to target and metrics by which to judge their suitability. Instead, we can aim at 3 different profiles of player that we will need in the front 3 regardless of whether they're playing a specific game at 9 or 10:

• Creator (Bruno Fernandes, Ødegaard, Riquelme)
  • Passing, Vision, Technique, Decisions, Tries Killer Balls
  • pass completion, key passes per 90, chances created per 90, assists per 90
• Speedster finisher (Vardy, Owen, early Anelka)
  • Acceleration, Pace, Finishing, Composure, Anticipation, Looks to break offside trap
  • xG cf Goals, goals per 90, shots per 90, shots on target %
• Physical finisher with aerial presence (Lukaku, Zlatan, Ronaldo)
  • Finishing, Composure, Anticipation, Off the Ball, Strength, Jumping Reach, Heading
  • xG cf Goals, goals per 90, shots per 90, shots on target %, headers %

I'm not saying that our front three will always consist of one of each - it's entirely possible that we may go with two creators and one speedster for a particular game - but, as there will be times across the season when the option to use each of these players will be useful, it makes sense to prepare a well-balanced squad. Some of the possible combinations might be:

By assigning these 3 profiles, it will allow us to tailor the recruitment accordingly. It is entirely possible, even preferential, that one player could fulfil the requirements of two profiles but at least this helps us to identify gaps in the squad and what would be needed to fill those gaps.

Overall

So that gives me a general idea of the individual roles and how they'll fit within the system. For now. There are, of course, universal considerations to generate a club culture - typical preferences for FMers like Determination, Teamwork, Work Rate and Stamina. 

With the inherent weakness of the system on the flanks, I've gone with a couple of team instructions that I hope will provide some mitigation but I'm sure a lot of this will change over the course of the first season.

I also have some early thoughts on what gaps exist within the current squad - primarily right back and up front - but that can wait for some more detailed thoughts when applying the recruitment policy to the system.

Edited January 9, 2022 by Shrewnaldo

[Shrewnaldo]

As an addendum to the tactical stuff above, the selected roles are certainly not set in stone. I don't see the concept changing much but finding the best way to have this reflected in the match engine might.

For example, I've had some early issues where either of the 10s on a support duty tend to clash with the midfield 3, the gap between the lines being compressed by both our midfield and the defensive line. So I'm experimenting with this alternative approach which essentially uses the exact same concept but with the midfield line withdrawn slightly. I love working through these sorts of tactical conundrums.

 

[SteinkelssonFM]

A big fan of your breakdown in relation to your vision in terms of the tactic and role requirements.

I'm not sure if you have seen my Zemanlandia thread but it resonates with some of your roles and there are a few baseline metrics in relation to some of the key metrics to base performance on. 

Also a fan on the less is more TI's will you look to shape/mould these based upon the eye test within the ME or will you favour PI's and traits? 

[_Ben_]

3 hours ago, Shrewnaldo said:

Cheers Ben. I did, yes, well remembered. With Academico de Viseu. It was pretty good fun but I never went as hard into the stats as I wanted then or am trying to now. I remember switching back to the regular skin after we won the Portuguese title and being shocked by how poor some of my favourite players' attributes were. Fun way to play though

How did you find it? Guadalajara only use players from their own area, don't they? Similar to Bilbao? Any thoughts or tips? 

I really enjoyed it and it took my statistical analysis to another level, at the time. I kind of do want to go back to that style and I've been working on a skin that uses polygons (with an average score for each area - e.g. pressing or press resistance rather than one attribute) and then relying on statistics to drive that. That might come in a future save though.

No, there was no restrictions for them because this was the Spanish one rather than the Mexican side. They are just a lower league side that played in purple, so I chose them!

[Shrewnaldo]

2 hours ago, MattyLewis11 said:

A big fan of your breakdown in relation to your vision in terms of the tactic and role requirements.

I'm not sure if you have seen my Zemanlandia thread but it resonates with some of your roles and there are a few baseline metrics in relation to some of the key metrics to base performance on. 

Also a fan on the less is more TI's will you look to shape/mould these based upon the eye test within the ME or will you favour PI's and traits? 

Thanks!

I hadn't read it - the chaos of Zeman systems don't tend to appeal to me it has to be said!! :-) But I'll definitely go take a look. Keen to see how others baseline KPIs

I tend to go with very few TIs in my tactics these days. I like to use the player roles and traits, plus a few PIs, to mould the system more than TIs. I would think that a couple more TIs will be added over the course of this season as the tactic develops but we'll see.

1 hour ago, _Ben_ said:

I really enjoyed it and it took my statistical analysis to another level, at the time. I kind of do want to go back to that style and I've been working on a skin that uses polygons (with an average score for each area - e.g. pressing or press resistance rather than one attribute) and then relying on statistics to drive that. That might come in a future save though.

No, there was no restrictions for them because this was the Spanish one rather than the Mexican side. They are just a lower league side that played in purple, so I chose them!

Ah, oops. I saw those polygons on some of your screenshots. I love that sort of thing, really good innovations.

 

In other news, I'm becoming increasingly irritated by the lazy and haphazard way statistics are used in this game. For example, statistics like "possession won" are available in only a couple of side screens - they aren't available in squad views and they aren't recorded per 90. When analysts score potential signings using this metric, they compare the total number against the rest of the league, regardless of the amount of game time that each player has had. For a game that's based entirely on numbers and is supposed to be a sort-of simulation of a sport that's increasingly stat-focused, it drives me insane.

[Shrewnaldo]

2021/22 - Halfway update

It's a long old season in the Championship. In Perú, the entire league season was 25 matches and we might have had 5 Copa Bicentenario games in the middle. Here I'm 23 matches in and just reached the halfway stage of the season so it's time for a bit of an appraisal of our situation.

So that's pretty good. The board want a top-half finish whilst the media predicted that we'd finish 19th so we've got to be pleased with the first half of the season. We've been really streaky - 3 without a win, 6 without defeat, 3 draws in a row, 3 losses in a row and the run that propelled us up the table, 6 back-to-back wins.

we also had a handful of Carabao Cup games, beating Northampton 1-0, Cambridge 4-1 and Forest 1-0 before disappointingly slipping up against League One Lincoln 2-1.

In terms of the playing personnel, here are the key squad statistics with the players listed in order of minutes played.

I've ringed a couple of standout statistics in red, should you be able to read it.

The first thing of note is that Nahki Wells has really disappointed. He's scored only 3 goals this season, all of them coming in the Carabao Cup. From an xG of 5.59 in the league, he has scored nothing. Zip. Zilch. Nada. Bugger all.

On the other side of the coin, I've highlighted some really surprising statistics from Joe Williams and Matty James. These two have become established as our first choice carrileros in the wide midfield positions - roles that I expected to be more defensive and about retaining possession than creating any attacking output. But Joe Williams has 2.15 key passes per game, creating 0.36 chances per 90; James has 2.53 and 0.62 respectively. Both are standout figures within the squad and will mean that I'll need to revise my requirements for players who fulfil that role in future.

In terms of overall team statistics, we're not deviating from the mean too much. What surprises me most is that we're below average for tackles per game by 2.3 and making 4.48 fewer fouls per game than the league average. Perhaps I need to instruct the team to be more aggressive because I'd like to see us putting the foot in more often, particularly with a combative midfield, and thereby dominate the ball a bit more than we currently are.

In terms of the Data Hub, a few statistics stand out:

Our aerial ability is shocking. Yes, we don't contest as many as other clubs but our success rate is woeful. Which probably ties into our worst-in-class cross completion ratio despite putting in a decent amount. And, matching the tackle and foul stats above, we don't win the ball back often enough.

Looking back at the squad statistics above, it's highlighted to me the really poor performance of one of our starting centre backs, Tomáš Kalas - winning just 69% of his 10.79 headers per game. That's absolutely not good enough. Kalas has the highest average ratings in our squad at 7.2 but losing that many headers is unacceptable for a centre back. That will mean Zak Vyner, how has been covering at right back whilst George Tanner has been injured and Danny Simpson has been ancient, will definitely move inside.

Further forward, our 3 strikers have the lowest heading success in the team - Semenyo on 33%, Janneh on 25% and Wells on 22%. This looks like a problem. The strikers tend to contest around 10 aerial challenges per game and whilst some of these will be from goalkeeper clearances, it only needs one look at that cross completion graph to know that many of them are chances to get a header on goal. We've already set the team to "low crosses" in an attempt to avoid this but it really does highlight that we have no-one who fits the profile of "physical finisher with aerial presence" that we'd identified in the recruitment strategy.

It might even be something we look at loaning in during the January window. This would give us the opportunity to experiment with the physical striker option and understand the statistical pre-requisites ahead of signing a permanent option in the summer.

The other areas I'll be looking at during the second half of the season will be right-back and another midfielder. But more on that in a later post.

Edited January 9, 2022 by Shrewnaldo

[keeper#1]

0 goals on a 5.59 xG?!?!?  That sounds like numbers I would put up but, at least, I had excuse of being a goalkeeper who occasionally played in the field when we were winning big. (Thanks to unlimited sub rules in the US high school game)

[Sonic Youth]

The title had me intrigued, and the thread interested 

That shot of stars instead of skills was definitely curious, and tricky to digest. Do you find yourself making notes about certain players?

[Shrewnaldo]

7 hours ago, keeper#1 said:

0 goals on a 5.59 xG?!?!?  That sounds like numbers I would put up but, at least, I had excuse of being a goalkeeper who occasionally played in the field when we were winning big. (Thanks to unlimited sub rules in the US high school game)

To be fair to Nahki, I think whilst my output would be the same, I wouldn't manage to get into positions necessary to accumulate 5.59xG :-) 

1 hour ago, Sonic Youth said:

The title had me intrigued, and the thread interested 

That shot of stars instead of skills was definitely curious, and tricky to digest. Do you find yourself making notes about certain players?

Thanks, I do like the title a lot!

Re notes, not yet but it's a good idea. I wish the game had a better notes system, something you could keep as one of the widgets on the player profile and just type in as bullet points 

[Shrewnaldo]

January 2022 business - Stat Attack

Having reached our first transfer window, a series of activated clauses in historical transfers had provided me with a reasonable kitty should I wish to invest in the squad. However, having already exceeded the wage budget through contract extensions for key players, we would need to sell before we could buy. 

In truth, we weren't in any position to buy anyone anyway. With the attribute restrictions and recruitment strategy in place, I need the second half of the season to solidify a list of targets through refined scouting. Comfortably top-half, there is no real pressure to make early signings to improve our form, so I was quite happy to take a little time and instead use January to get rid of the deadwood. As such, we had a handful of departures with Chris Martin joining Motherwell, Danny Simpson allowed to leave for Rotherham on a free and Nahki Wells offloaded to Derby for £650k. Between Simpson and Wells, that was a solid £45k a week off the wage budget - bringing us well under our budget and providing the opportunity to snap up Bosmans should our scouting through Q1 and Q2 identify a reasonable target.

Whilst we weren't making any permanent transfers in January, I did see an opportunity to use the loan market to our advantage. Without committing to a long-term deal, a 6-month loan allows us to bring in a player that I might be interested in signing permanently and get a much better look at him than scouting will ever give me. Indeed, for players who would otherwise sit on another team's bench, it provides just about the only opportunity for some statistical analysis. But the wage budget bust, at least until Simpson and Wells were sold, we couldn't afford any wage contributions for loans so I had to find a work around.

Take Millwall midfielder Ben Thompson, for example. He's out of contract in the summer and I'm very interested in him as a carrilero. He's got some great looking mental attributes - aggressive, brave, hard working, team player, runs all day. Perfect. He was getting a handful of games at Millwall but not enough to really give me a good spread of statistics. Trying to loan him in, Millwall wanted us to fund 60% of his wages - £3600 a week. We couldn't afford this, so instead I offered them £16k a month loan fee, covering his £3600 per week (£3600*4 = £14,400) and changing the wage contribution to zero. Millwall accept, we get our man without any hit to our wage budget and a total fee of £96k. Lovely.

We also secured similar loan deals for Marko Kvasina from Oostende, Jonathan Bolingi from Antwerp and Ousseynou Ba from Olympiacos. Only the latter two interest me in terms of potential long-term deals - Bolingi as a physical striker and Ba as a centre back. Kvasina, at 6'5", allows me to experiment with a player I know to be a unit in the air and very limited outside that. If this proves a success then it informs what I should be looking for over the summer. And early signs are, that it may work out quite well.

 Here, for example, is Kvasina acting as the classic target man - holding the ball up, pulling a centre back out of position, laying it back to Thompson at carrilero who has a better passing range to play the through ball to the third man run from Weimann at shadow striker.

In the same game, Kvasina would also score a close range header from an open play left wing cross. In the next game, Bolingi - a more rounded player but still a physical specimen - would score and assist. This has really given me pause for thought on the priorities for the summer. Something which I've tried to summarise in a handy 'cheat sheet'.

Listing out all the roles and profiles that were identified in the posts above, I've assigned a quick traffic-light score to the current squad status for both the first choice option and a back-up. To ensure that we have appropriate squad depth, no player can be counted twice and each player is scored on 3 key criteria: ability, age and departure potential.

So our back-up central midfielder scores all reds because we don't have one; the back-up centre back option scores red for ability because he's awful, the back-up right carrilero scores red for age as he's 33 and will need replacing imminently. Departure potential covers a number of possibilities - either the player's contract is out soon, or they might be the subject of strong transfer speculation or, as is the case for a number of players here, they are on loan. Either way, I really need to be looking at replacements with some urgency.

The traffic-lighting gives me a very quick, graphical overlay of where my priorities should be for summer recruitment. Clearly centre back, central midfield and big lumps up front are what is most needed.

Taking centre back as an example, I then tried to build a statistical profile to build on the 5 criteria each player will need to satisfy in order to become a 'top target'. I find the way that statistics are collated in Football Manager to be incredibly haphazard, to the point of frustration with many of them. Nevertheless, the Data Hub has helped in one key way - it is easier to identify what 'good' means for your league.

Using the scattergraphs, and the handily produced indicators for mean level, we can quickly know that ~75% win ratio for headers is good for our league. Combining this with some painstaking reviews of our players stats pages - some via their profile, some via the squad views and some via the league detailed statistics (because SI seem to want to make this difficult for their users) - I came up with a series of pre-requisites:

• 14 contested headers per game
• 76% aerial win rate
• 7 clearances per game
• 1.2 blocks per game
• 80% tackles win ratio
• 92% pass completion
• 10.5 possessions won/90
• <7 possessions lost /90

Obviously there has to be some flexibility because some of them will be dependent on the level of team or the style of their play, but that provided a framework. With a few days left of the window, I decided to go looking at our longlist to see whether this was a reasonable profile to target. One player stood-out immediately as meeting all of my criteria, along with the statistical pre-requisites:

Coming from the league below, I also allowed for a "quality distortion" of about 5% and he still met the criteria. I decided to go back and watch some of his matches as if I had scouted them all along, liked what I saw and put in a bid. Sadly, the only way we could agree terms was if Ipswich could agree to sign a replacement first and we all know how that goes. Nevertheless, come the summer, I've no doubt that we'll be back in for young master Woolfenden.

Win some, lose some, but that's a pretty handy example of what I hope to get from the recruitment strategy. Now we just need to spend 6 months finding our key targets - centre back, central midfielder and a big lump or two.

Edited January 12, 2022 by Shrewnaldo

[\'Appy \'Ammer]

On 06/01/2022 at 21:30, Shrewnaldo said:

I will be using a skin which grades the attributes in four levels, which I have set as:

• <8       - Poor         - Grey
• 8-11   - Average   - Silver
• 12-15 - Good       - Gold
• >15    - Elite         - Green

I am interested as to why you changed the parameters for the attributes? Good read by the way

[extudassex2]

Now the fun part begins! Looking forward for more updates full of stats.

[_Ben_]

23 hours ago, Shrewnaldo said:

Using the scattergraphs, and the handily produced indicators for mean level, we can quickly know that ~75% win ratio for headers is good for our league. Combining this with some painstaking reviews of our players stats pages - some via their profile, some via the squad views and some via the league detailed statistics (because SI seem to want to make this difficult for their users) - I came up with a series of pre-requisites:

23 hours ago, Shrewnaldo said:

Coming from the league below, I also allowed for a "quality distortion" of about 5% and he still met the criteria.

I love this Shrew! I can't believe I didn't think to look for mean values there. I've thought for so long about setting up KPIs for my players but now it feels so much more simple: "above average in the league we're in" 

Just makes it really annoying that they provide me with an average xG per shot statistic but, obviously, I can't search for anyone with an xG/shot of over 0.12, for example. Same with the unicorn-like Possession won and Possession lost statistics that appear it literally zero useful places! Don't even get me started that they've lumped full backs and centre backs in the defensive category! What I'd give for my centre backs to be nailing more than 1.1 key passes/90.

[Shrewnaldo]

10 hours ago, \&#x27;Appy \&#x27;Ammer said:

I am interested as to why you changed the parameters for the attributes? Good read by the way

I think I lowered one or two of the bandings slightly? It was just to align with what I would think was a good and excellent attribute for this level. I may bump it a little if/when we go up - although that may give me an idea of specific players'  attributes so I'll have to think about it.

9 hours ago, extudassex2 said:

Now the fun part begins! Looking forward for more updates full of stats.

I have been looking through FAR too many stats today. Exhausting.

1 hour ago, _Ben_ said:

I love this Shrew! I can't believe I didn't think to look for mean values there. I've thought for so long about setting up KPIs for my players but now it feels so much more simple: "above average in the league we're in" 

Just makes it really annoying that they provide me with an average xG per shot statistic but, obviously, I can't search for anyone with an xG/shot of over 0.12, for example. Same with the unicorn-like Possession won and Possession lost statistics that appear it literally zero useful places! Don't even get me started that they've lumped full backs and centre backs in the defensive category! What I'd give for my centre backs to be nailing more than 1.1 key passes/90.

Cheers Ben, appreciated.

I have realised one downside to using the mean values on the graphs - they average out values from ALL defenders, so including centre backs and full backs. This obviously skews the numbers down somewhat if you'd prefer to look at just centre backs. It's really annoying - it's like all the stats just aren't thought through. There's no consistency on what is available and, as you say, some of the stuff would be REALLY useful but is just hidden away and only really available on a game-by-game basis. In my view, ALL the stats should be available, on both a nominal and a per 90 metric, on ALL screens. So on Player Detailed screens within the League view, on the player profile, on the recruitment analyst report, on squad views and as search parameters in scouting / shortlists. Otherwise it just makes it harder for us.

The recruitment analyst reports are particularly annoying. A player might be marked down because he has, for example, only made 100 interceptions that season. But it takes no account for the number of games in which he has accumulated those 100. He might have been injured, he might have just not played. It should be done as a per90 metric. It's really just sloppy.

[BadManager]

This is really different. I have to admit I don't often look at the stats like that, and there is so much there (even on older versions). Interesting way of going about things. You're starting to see the gaps in those stats, which will be an issue. I suspect most people don't see it because they don't drill down far enough.

Upping the bands for different levels and leagues makes sense to me overall, though given you are already at a high standard its probably not necessary. If you had started in the conferences, setting 10-13 as "good" might have made sense and needed adjusting long before the championship. So long as you don't bump things too much so only 19-20 become "excellent" you should be able to do some adjustment without giving too much away.

 

[Shrewnaldo]

1 hour ago, BadManager said:

This is really different. I have to admit I don't often look at the stats like that, and there is so much there (even on older versions). Interesting way of going about things. You're starting to see the gaps in those stats, which will be an issue. I suspect most people don't see it because they don't drill down far enough.

Upping the bands for different levels and leagues makes sense to me overall, though given you are already at a high standard its probably not necessary. If you had started in the conferences, setting 10-13 as "good" might have made sense and needed adjusting long before the championship. So long as you don't bump things too much so only 19-20 become "excellent" you should be able to do some adjustment without giving too much away.

 

Cheers. Aye, it is a frustrating way to play as, like you say, you really pick up on the gaps in the way the data is available and presented. At least it's better than it was last year when the statistics and match analysis were just completely and utterly broken.

[deltablue]

Great thread title

[Shrewnaldo]

9 hours ago, deltablue said:

Great thread title

Thanks, I am a little proud of it, I have to admit 

[Shrewnaldo]

So erm... yeah

Bit of an unexpected turn. We got promoted.

I don't think we were even that good. We didn't win a game in March, for example. And we went through February scoring only 4 goals, albeit not conceding in the 6 games. In truth, the promotion feels really meh. Like unearned and I was seriously considering save scumming so that we didn't win in the playoffs but I just can't do it. Not least because trying to stay up with this team *should* be an almighty challenge and essentially cheating to avoid that challenge wouldn't feel right.

One of the biggest problems we're going to have is that all this careful scouting I've done, in order to inform our super-duper recruitment strategy, was all aimed at good Championship level players. I reckon there aren't more than two or three players on the longlist of Premier League quality. So that's a bit of an issue...

We qualified for the final playoff spot with 77 points and played Fulham in the semi-final. Having lost both regular season games to the Cottagers (2-1 and 3-2), I was fairly confident that this would be the end of the road. However, with Michael Hector seeing red on 20 minutes, by which time we were already 1-0 up, we took the home leg 3-0. It was definitely tighter in the second leg, with Fulham coming at us particularly in the second half. But a 3-1 defeat was enough to put us through 4-3 on aggregate to face, somewhat bizarrely, Barnsley in the final.

We'd lost 3-1 at Barnsley in March, with their wide men Frieser and Gordon absolutely destroying us. But a lovely goal from loanee Jonathan Bolingi was enough to give the Robins victory in front of 90k at Wembley and victory in football's richest match. Which was a bit of a bummer.

So where does that leave us? The board have given me £25m to strengthen a squad that, frankly, has no business being anywhere near the Premier League. So before we get onto what we'll need to do with that budget, some analysis of how we got here.

^^^ My "Effectiveness View" with the players ordered by minutes played ^^^

Starting with the good, our 'keeper Daniel Bentley has been outstanding. Seriously, seriously good. No really, this good.

And, rather obviously, I've been really pleased with the other players who have got lots of game time. Centre back Nathan Baker, in particular, has been excellent with great statistics aerially (87% header win ratio) and on the deck (82% tackle success ratio). Indeed the defence, in general, has been solid throughout. Rotating between the starting options of Baker, Zak Vyner and Ousseynou Ba has worked out really well, with the loanee having the lowest "team conceded per 90" in our squad at just 0.63. Prior to the promotion, he was an absolute certainty as a summer target.

Even at right back we've done ok, rotating two players that I don't particularly rate - Kalas as generally unsuited for the role and Tanner as lacking the required talent. It was an area that I'd already identified as needing strengthening and it's now an absolute top priority.

On the other flank is the curious case of Jay Dasilva. He's supposed to be one of our best players and is clearly very good but there are two problems. The first is that his natural fitness is absolutely shocking. This means that it takes him longer to recover fitness in between games and he's practically incapable of playing two matches a week - which might be less of a problem in the Premier League. The second is his creative output. In 2913 minutes, he got 3 assists - 0.09 per 90. With the 4-3-2-1 formation, I'd really expect our wingbacks as the only wide players to be creating a lot more than this. However, he has been creating chances at a rate of 0.65 per 90 which is the best in our team, joint with Bolingi, and makes him one of the most creative players in the league.

So I'm willing to accept that his low assist record isn't his fault - he's creating the chances, it's just that the player on the end of his cross is missing.

I'm similarly happy with our options in centre midfield. The likes of Joe Williams, Tyreeq, Matty James and Andy King have just been solid all season. Unspectacular for the most part but completely reliable. A few statistics standout - such as Tyreeq's particularly poor tackle success ratio - of specific importance to the carrilero role he's been fulfilling.   And Matty James, who has played the middle of the three more often than most, has faced the fewest headers per game at a little under 4 - and despite being "only" 5'11" with average aerial ability has won around three quarters of them. This tells me that some form of aerial prowess is not as important as I had thought it might be.

And that takes me to the front three, where I've had most indecision (as usual) throughout the season.

The 'big man up top' experiment didn't entirely work out. Per the highlight in the previous post, I really liked the effect of the targetman dropping off the line to hold the ball up, lay it off and allow the midfield runners to get beyond him for through balls. But the other side of it - winning headers in the middle - hasn't really happened. To be fair, Kvasina (at 6'5") got injured pretty early on and didn't get many games but Bolingi, albeit less of an aerial threat at 6'2", just wasn't getting on the end of crosses.  So I tried to take what I liked - dropping deep to create space - and switched up the striker to a false 9. This moved the creator role to the 9 spot and let both 10s become runners. Weimann as a shadow striker was excellent, even if I'd prefer a little more pace, whilst Bolingi became an attacking midfielder and able to use his physicality from deep.

The biggest problem I have now is that, except for the shadow striker, my front three has been changed so often that I can't really baseline the statistics against my players' performances this season. Semenyo finished the season as a false 9, but played 9 games at AM and started the season as an AF. Bolingi was similarly varied.

Without the option of separating out stats from their time in particular positions or roles, I basically need to go with gut feel. That being said, it's easy to say that anyone in that front three should have good shooting statistics, cover plenty of ground and, to varying degrees, creative statistics like assists per 90 and chances created per 90. So let's just aim for the best that we can across those statistics.

And it's here that I think the majority of our summer's recruitment will focus. We need quality. Antoine Semenyo and Andi Weimann are decent players for the Championship but Premier League? Previous experience for the latter would suggest not and, whilst he will likely one day be Premier League standard, the former isn't quite there yet. Further, Matty James has done his achilles which means he'll be out until around November and, even then, is likely to return a poorer player. He'll be 31 when he returns and, at that age, there's a decent chance he'll struggle to recover from such a serious injury. That means dropping the excellent Han-Noah Massengo into the middle three and leaving another gap in the front three.

But, you know, I'm sure it's really easy for newly promoted sides to find Premier League level attacking quality for £25m.

One other item of note - FM doesn't record statistics season to season. Once we tick over into the next season, even the haphazardly presented statistics that are currently visible will be removed. To inform our statistically-based recruitment, I'll therefore be keeping a save from the end of each season. This will let me reload the game to review statistics for particular players over the history of the save. A bit labour intensive but that's about all we can do.

[Shrewnaldo]

A note on recruiting for the Premier League

I reckon I could go with one of two strategies:

1. Stick with the longlist that we've already developed
2. Ditch that and trawl bigger teams for discarded big names that we can bring in

I am very much going to go with option 1. I'm hoping the board will not expect us to stay up. We are almost certainly going to come back down. So we need to plan long-term and therefore the players that we've already looked at are what we should stick with. Ok an option or two might come up that we can't turn down, but I'm loathe to step away from the preparation that's been going on for the last ten months - particularly when I don't have the precision of numeric attributes to rely on.

What we can do is try to structure deals - preferably with some sort of survival bonus. £25m isn't much by Premier League standards but it's way, way more than I would usually spend in a transfer window so I'm relatively confident about finding the right sort of talent on a budget. But there's no need to throw money away for nothing.

[SteinkelssonFM]

On 13/01/2022 at 23:04, Shrewnaldo said:

I think I lowered one or two of the bandings slightly? It was just to align with what I would think was a good and excellent attribute for this level. I may bump it a little if/when we go up - although that may give me an idea of specific players'  attributes so I'll have to think about it.

I have been looking through FAR too many stats today. Exhausting.

Cheers Ben, appreciated.

I have realised one downside to using the mean values on the graphs - they average out values from ALL defenders, so including centre backs and full backs. This obviously skews the numbers down somewhat if you'd prefer to look at just centre backs. It's really annoying - it's like all the stats just aren't thought through. There's no consistency on what is available and, as you say, some of the stuff would be REALLY useful but is just hidden away and only really available on a game-by-game basis. In my view, ALL the stats should be available, on both a nominal and a per 90 metric, on ALL screens. So on Player Detailed screens within the League view, on the player profile, on the recruitment analyst report, on squad views and as search parameters in scouting / shortlists. Otherwise it just makes it harder for us.

The recruitment analyst reports are particularly annoying. A player might be marked down because he has, for example, only made 100 interceptions that season. But it takes no account for the number of games in which he has accumulated those 100. He might have been injured, he might have just not played. It should be done as a per90 metric. It's really just sloppy.

Completely agree with you on this one, all the stats on the player search tab also comes from every game that player has played that season rather than being able to focus on a specific competition.

So for a club which is playing in multiple competitions it will capture all of his minutes which then impact P90 metrics.

You can manually find minutes per comp on the player page under stats but as you said not all metrics are detailed which then means having to look under the league detailed invididual drop-down but again that doesn't list every player. 

😡

[Shrewnaldo]

1 hour ago, MattyLewis11 said:

Completely agree with you on this one, all the stats on the player search tab also comes from every game that player has played that season rather than being able to focus on a specific competition.

So for a club which is playing in multiple competitions it will capture all of his minutes which then impact P90 metrics.

You can manually find minutes per comp on the player page under stats but as you said not all metrics are detailed which then means having to look under the league detailed invididual drop-down but again that doesn't list every player. 

😡

Indeed. Or being able to switch it to only games where a player has played in a certain position. Say you have a player who covers both 9 and 10, you'd want to know their stats for these positions separately.

The per90 thing is what bugs me most, though. Because you'd think that'd be simplest thing in the world to present

[Shrewnaldo]

The hunt for a creative 10

Apologies in advance but this post will be a bit of a stream of consciousness as I try to work through identifying a primary target for that creative 10 spot.

The way I've been trying to build the front three is that we have certain profiles of player who can play both at 9 and 10. So far, we'll be carrying the following into our debut Premier League season:

• speedy runner:
  • Antoine Semenyo - decent, but more promising than immediately effective
  • Andreas Weimann - decent, not as quick as I'd like but runs all day
• physical unit:
  • Saša Kalajdzic - newly signed from Stuttgart for £4.2m, 6'7" and capable on the ball
  • Jonathan Bolingi, playoff hero, smaller than Kalajdziv at 6'2" but quicker and puts himself about
• creators:
  • Alex Scott - promising but just 18 and really isn't ready for Premier League football
  • Callum O'Dowda - just not Premier League quality

Therefore it's clear that the priority is that creative role and, should we have any money left, try and pick up a runner - perhaps in the loan market.

OK, so we want a creator but what does that mean? Assists? Certainly. Key passes? Probably. So the first thing I did was set up a custom view on the shortlists, looking at the key statistics for the role in order to assess the players in Tier 1:

And Tier 2:

Two of the players who immediately stood out were Forest's Joao Carvalho and Domingos Quina, who had been on loan at Fulham from Watford. Carvalho was soon ruled out, however, as he signed a new contract and, as was to become a frustrating habit, it quickly became apparent that his statistics were heavily distorted by set pieces. Taking the majority of his club's freekicks and corners, his key passes stats were artificially inflated against a player like Quina who could only obtain such metrics from open play.

Quina, though, has an optional fee agreed with Fulham and besides, having watched Fulham a number of times primarily with an eye on Harry Wilson, he has a tendency to go missing in games - including in both legs of our playoff semi-final. Other options were required.

A couple of other options came to the fore. Tom Lawrence at Derby played the majority of the season as a 10 in narrow 4-2-3-1 that wasn't too dissimilar to our christmas tree. And whilst his key passes / 90 metric was skewed by his set piece taking, the 0.67 chances created per game is pretty impressive for a player that got relegated. Also going down, Peterborough's Sammie Szmodics is a late entrant to the shortlist having impressed against us despite losing 6-1. Meanwhile, QPR's Chris Willock and Ilhas Chair had impressed as dual 10s in an attacking system not too dissimilar to our own and Keane Lewis-Potter had some potential but had played the majority of his season from the left for Hull and statistically looks more like a runner than a creator.

One outside option is Tom Cairney - a player I like in real life and whose numbers were relatively impressive, albeit only across a small sample size of 457 first team minutes at Fulham. He's also one-footed which I find can be a limiting factor with 10s who I expect to receive the ball under some amount of pressure.

The last option was one which had no basis on statistics whatsoever. With our unexpected promotion, I went looking for a few options that I liked across the footballing world just based on my own knowledge. Rapid Wien's Yusuf Demir was an obvious one and, with Barca not taking up their option, I set up a £6.5m deal (rising to £11m with clauses). However, I've since cancelled this. I have no doubt he'd be excellent but it's really going against the premise of the save and wouldn't feel right just to abandon that to go pick up a known wonderkid.

In the end, I've narrowed it down to these 5 options.

As an aside, the distance / 90 metrics in Football Manager are a tad excessive.

In terms of how they compare to my own players, these are the equivalent stats for all the players in the Bristol City squad who can play at 10 - although only Massengo, Weimann, Scott and O'Dowda have done so with any regularity.

Taking that as a benchmark, anything north of 1.25 key passes per game and 0.5 chances created per 90 is going to be useful. Taking each of the 5 short-listed players in turn then:

• Tom Cairney
  • Pros: cheap, good chances created per 90, fits character of the squad
  • Cons: scouts reckon Championship level only, small sample size, one-footed, 31
• Domingos Quina
  • Pros: top-rated by scouts, excellent stats without set piece distortion, exceeds xG, relatively cheap, stats available whilst playing as 10, excellent dribbling, two-footed
  • Cons: potential inconsistency, low work rate
• Harry Wilson
  • Pros: set piece strength, positive traits (Tries Killer Balls), relatively two-footed, can also play 9
  • Cons: expensive, underperforms xG, potential set piece distortion on creativity stats, stats from playing wide right not as a 10, low assists stats
• Tom Lawrence:
  • Pros: cheap, dribbling, set piece strength, consistent, enjoys big matches, stats from playing as a 10, two-footed, can also play as 9
  • Cons: might be injury prone, shoots from distance trait, potential set piece distortion on stats, mental attributes look low
• Sammie Szmodics
  • Pros: cheap, work rate and mental attributes in general, two-footed, quick, professional, no set piece distortion, good player traits, can also play 9
  • Cons: scouts aren't keen, technically not the best, stats come from playing as 9 and 10 (50/50)

And so, on the basis of all of that I'm leaning towards Domingos Quina as the primary target, potentially bringing in either Cairney or Szmodics as a back-up. What are your thoughts? Anything I'm over-looking here?

[Goughy7]

48 minutes ago, Shrewnaldo said:

The hunt for a creative 10

Apologies in advance but this post will be a bit of a stream of consciousness as I try to work through identifying a primary target for that creative 10 spot.

The way I've been trying to build the front three is that we have certain profiles of player who can play both at 9 and 10. So far, we'll be carrying the following into our debut Premier League season:

• speedy runner:
  • Antoine Semenyo - decent, but more promising than immediately effective
  • Andreas Weimann - decent, not as quick as I'd like but runs all day
• physical unit:
  • Saša Kalajdzic - newly signed from Stuttgart for £4.2m, 6'7" and capable on the ball
  • Jonathan Bolingi, playoff hero, smaller than Kalajdziv at 6'2" but quicker and puts himself about
• creators:
  • Alex Scott - promising but just 18 and really isn't ready for Premier League football
  • Callum O'Dowda - just not Premier League quality

Therefore it's clear that the priority is that creative role and, should we have any money left, try and pick up a runner - perhaps in the loan market.

OK, so we want a creator but what does that mean? Assists? Certainly. Key passes? Probably. So the first thing I did was set up a custom view on the shortlists, looking at the key statistics for the role in order to assess the players in Tier 1:

And Tier 2:

Two of the players who immediately stood out were Forest's Joao Carvalho and Domingos Quina, who had been on loan at Fulham from Watford. Carvalho was soon ruled out, however, as he signed a new contract and, as was to become a frustrating habit, it quickly became apparent that his statistics were heavily distorted by set pieces. Taking the majority of his club's freekicks and corners, his key passes stats were artificially inflated against a player like Quina who could only obtain such metrics from open play.

Quina, though, has an optional fee agreed with Fulham and besides, having watched Fulham a number of times primarily with an eye on Harry Wilson, he has a tendency to go missing in games - including in both legs of our playoff semi-final. Other options were required.

A couple of other options came to the fore. Tom Lawrence at Derby played the majority of the season as a 10 in narrow 4-2-3-1 that wasn't too dissimilar to our christmas tree. And whilst his key passes / 90 metric was skewed by his set piece taking, the 0.67 chances created per game is pretty impressive for a player that got relegated. Also going down, Peterborough's Sammie Szmodics is a late entrant to the shortlist having impressed against us despite losing 6-1. Meanwhile, QPR's Chris Willock and Ilhas Chair had impressed as dual 10s in an attacking system not too dissimilar to our own and Keane Lewis-Potter had some potential but had played the majority of his season from the left for Hull and statistically looks more like a runner than a creator.

One outside option is Tom Cairney - a player I like in real life and whose numbers were relatively impressive, albeit only across a small sample size of 457 first team minutes at Fulham. He's also one-footed which I find can be a limiting factor with 10s who I expect to receive the ball under some amount of pressure.

The last option was one which had no basis on statistics whatsoever. With our unexpected promotion, I went looking for a few options that I liked across the footballing world just based on my own knowledge. Rapid Wien's Yusuf Demir was an obvious one and, with Barca not taking up their option, I set up a £6.5m deal (rising to £11m with clauses). However, I've since cancelled this. I have no doubt he'd be excellent but it's really going against the premise of the save and wouldn't feel right just to abandon that to go pick up a known wonderkid.

In the end, I've narrowed it down to these 5 options.

As an aside, the distance / 90 metrics in Football Manager are a tad excessive.

In terms of how they compare to my own players, these are the equivalent stats for all the players in the Bristol City squad who can play at 10 - although only Massengo, Weimann, Scott and O'Dowda have done so with any regularity.

Taking that as a benchmark, anything north of 1.25 key passes per game and 0.5 chances created per 90 is going to be useful. Taking each of the 5 short-listed players in turn then:

• Tom Cairney
  • Pros: cheap, good chances created per 90, fits character of the squad
  • Cons: scouts reckon Championship level only, small sample size, one-footed, 31
• Domingos Quina
  • Pros: top-rated by scouts, excellent stats without set piece distortion, exceeds xG, relatively cheap, stats available whilst playing as 10, excellent dribbling, two-footed
  • Cons: potential inconsistency, low work rate
• Harry Wilson
  • Pros: set piece strength, positive traits (Tries Killer Balls), relatively two-footed, can also play 9
  • Cons: expensive, underperforms xG, potential set piece distortion on creativity stats, stats from playing wide right not as a 10, low assists stats
• Tom Lawrence:
  • Pros: cheap, dribbling, set piece strength, consistent, enjoys big matches, stats from playing as a 10, two-footed, can also play as 9
  • Cons: might be injury prone, shoots from distance trait, potential set piece distortion on stats, mental attributes look low
• Sammie Szmodics
  • Pros: cheap, work rate and mental attributes in general, two-footed, quick, professional, no set piece distortion, good player traits, can also play 9
  • Cons: scouts aren't keen, technically not the best, stats come from playing as 9 and 10 (50/50)

And so, on the basis of all of that I'm leaning towards Domingos Quina as the primary target, potentially bringing in either Cairney or Szmodics as a back-up. What are your thoughts? Anything I'm over-looking here?

So weird to see you managing my IRL team, Shrew. Surreal. I did the same as you this year and got City up via the play-offs unexpectedly (especially considering us IRL, oh god) and managed to finish nice and mid table.

As far as a 10 goes, Szmodics is a weird choice to read given we sold him to Peterborough after not getting a fair shake here, no way he’s good enough for the Prem even as a back up. I went for Nicolae Stanciu from Slavia Prague for about £4m I think it was, magnificent 10. Somehow managed to get Reiner in on loan from R. Madrid too, what a coup. 
 

Loving that Sasa K. signing, I went with Lorenzo Lucca and Thomas Chory upfront, units. That **** worked amazingly. 

Edited January 16, 2022 by Goughy7

[Shrewnaldo]

2 hours ago, Goughy7 said:

So weird to see you managing my IRL team, Shrew. Surreal. I did the same as you this year and got City up via the play-offs unexpectedly (especially considering us IRL, oh god) and managed to finish nice and mid table.

As far as a 10 goes, Szmodics is a weird choice to read given we sold him to Peterborough after not getting a fair shake here, no way he’s good enough for the Prem even as a back up. I went for Nicolae Stanciu from Slavia Prague for about £4m I think it was, magnificent 10. Somehow managed to get Reiner in on loan from R. Madrid too, what a coup. 
 

Loving that Sasa K. signing, I went with Lorenzo Lucca and Thomas Chory upfront, units. That **** worked amazingly. 

I noticed that about Szmodics. Funnily enough there have been a couple of players that I've looked at bringing back into the club.

But it's nice playing without any preconceptions about anyone - I know absolutely nothing about Championship football and Szmodics played really, really well against us, quite possibly due to the former club thing. I checked his stats and he'd been one of the most creative players in the league all season. I'm still thinking about bringing him in as back-up but the scouts absolutely agree with your view of him.

Sounds like you're smashing it. How far through are you now?

[_Ben_]

8 hours ago, Shrewnaldo said:

The per90 thing is what bugs me most, though. Because you'd think that'd be simplest thing in the world to present

Shall we start with the fact that some aren't even called per90!! I don't understand the oversight that sees some '/per game' and some '/90'

3 hours ago, Shrewnaldo said:

What are your thoughts? Anything I'm over-looking here?

This is about as thorough as you're going to get! Really inspirational stuff here Shrew - and can be adapted (as I think you've seen in my own thread) whether you can see the attributes or not!

[ifinnem]

Will be reading with interest to see how this plays out. Great concept and especially together with the star attributes. I just worry that stats are so dependent on systems and roles that using them for recruitment will not be that translatable but maybe I’m proved wrong.

[extudassex2]

Certainly not easy to overcome the quite serious limitations of data presented, but you're certainly doing an great job of it!

[weecoOk]

Great thread Shrew, just wondering how you go about deciding on roation? is it purely a fitness thing?

[Shrewnaldo]

On 16/01/2022 at 18:47, _Ben_ said:

This is about as thorough as you're going to get! Really inspirational stuff here Shrew - and can be adapted (as I think you've seen in my own thread) whether you can see the attributes or not!

Thanks. I ended up going with Wilson, purely down to the set piece strength funnily enough as we need some sort of edge. Might follow-up with another bid for Quina though - in no small part because...

20 hours ago, ifinnem said:

Will be reading with interest to see how this plays out. Great concept and especially together with the star attributes. I just worry that stats are so dependent on systems and roles that using them for recruitment will not be that translatable but maybe I’m proved wrong.

Thanks. Absolutely chime with that concern. It's something that really bothers me about stats in real life but mostly in FM - that there must be an element of post hoc ergo propter hoc. I do try to check the systems that the clubs are using to find out where the player is being used and try to take that into account but it is definitely a concern. Take defenders, for example - players in a poor side that do a lot of defending are always going to have more blocks and clearances than better defenders from bigger sides.

19 hours ago, extudassex2 said:

Certainly not easy to overcome the quite serious limitations of data presented, but you're certainly doing an great job of it!

Thanks. We'll see how far I get before the frustrations overtake me.

18 hours ago, weecoOk said:

Great thread Shrew, just wondering how you go about deciding on roation? is it purely a fitness thing?

Thanks. On rotation, it is nearly always just based on fitness, but I also have a default squad view and tactics view which includes a column for playing time happiness. So this gives me an early heads up if anyone is getting concerned that they aren't starting enough games. And then there's always loss of form that drives me into a change.

Lastly, I am trying to be more reactive - particularly with the front three. So I might play a pacy AF against a really slow defence, for example.

Edited January 18, 2022 by Shrewnaldo

[Shrewnaldo]

Will provide a proper update summarising the summer's recruitment but this Premier League malarkey is a proper bloody challenge.

Been really tough so far, requiring a lot of micro-managing. We've been absolutely battered in some games (Leicester and Man Utd), rode our luck in a couple of others (Spurs and Villa) and been genuinely unlucky in others (Norwich and Leeds). 

Goalscoring has been... difficult.

[Sonic Youth]

Tough start to the Prem. 6/14/-8 for 8pts from 9 games is tight at the back but not enough goals going forward. Hopefully there are three or more teams struggling and in a worse position.

[Sonic Youth]

I was going to say: Is there a way to make skins that put these stats in a more organised and logical arrangement? But then re-read your mentionings of the dilemmas from where to pull the stats from.

I noticed that it could for some, put game-by-game singularity stats might not be so possible (though that could mean they change after ever game and or require a bit more knowledge of writing a code that collates the stats wanted when putting into a skin if possible).

Would collecting and keeping stats in-game of all leagues loaded in view/full league modes add much to processing times and game sizes?

It would be good to have more passing stats. They already record and show pass maps. Surely it wouldn’t be a stretch to add pass distance, vertical/diagonal/horizontal/forwards/backwards/lofted passes into the game along with pass percentage. If they’ve got this, breaking down were the passes happen (ie into final third, into the box, line breaking, into space) could just as easily be worked out. From that we could get Packing Rate, xT and other stats included. Though just having xA would be a start. And separating all the stats from set piece/penalties (non-penalty goals for example) would be hot on xA’s heels!

Edited January 19, 2022 by Sonic Youth
Speeling

[SixPointer]

16 hours ago, Shrewnaldo said:

Will provide a proper update summarising the summer's recruitment but this Premier League malarkey is a proper bloody challenge.

Been really tough so far, requiring a lot of micro-managing. We've been absolutely battered in some games (Leicester and Man Utd), rode our luck in a couple of others (Spurs and Villa) and been genuinely unlucky in others (Norwich and Leeds). 

Goalscoring has been... difficult.

Tough step up in class. But the fact you felt hard done by in a couple of games, drew with spurs and beat West Brom and Bournemouth I think you can mix it up they're and give a good fight.