Would using genie scout ONLY ON YOUR OWN TEAM be cheating? I think not...

[RobinGoodey]

Right....

And your personal attack on certain people in this thread added what to the conversation exactly?

Hit a nerve did I?

And where did I claim that I was adding anything?

Unlike some......

And if you can point out where I made a 'personal attack' on you, please do - pointing out that you and 2 others have repeated the same posts 50 times isn't a personal attack - it's a valid opinion..... oh wait - if I don't agree with your point of view, clearly that's not allowed........

[x42bn6]

For those of you refusing to use the information at your fingertips to estimate whether a player has decent PA, a few tips to get you started:

I suppose the key point for me is that this information is definitely available, but not in the correct way.

A manager cannot make his own rating of a player's potential.

If a top manager moved to a club with rubbish coaches and the board refused to give him the wages to sign good coaches with his JPA/JPP, then why can the manager not use his own ability to judge a player instead?

The manager, who is supposed to be the head-honcho of the team, knows less about a player's potential than his coaches, even if he has been at the club longer than all his coaches.

So yes, this information exists - but where is the own manager's ability to judge?

1) Ask all your coaches what they think of him, and do this at semi-regular intervals, say 2 - 3 times in the first season. (Or scout the player with as many players as you can if he's not your player, although this thread was originally about your team it has gone very OT). Then take into account how good the coaches are at rating the player. Note that the longer a coach has worked with a player/more times a scout has scouted a player, the more accurate their reports become in general if they're any good at JPA. Take whatever they say with a hefty pinch of salt of course; that's why you ask several scouts to look at the same player over several matches/a season.

2) Look at the attributes. Despite what you think, they reveal a lot more about PA than you think. This is largely down to how SI appear to generate newgens, and may not be completely up to date...

2.1) Look at the key attributes for the player's position. In contrast to the AI, you as a human can rate players based on their attributes. The AI apparently does not. So you'll automatically know that a defender with 14 tackling might well be equipped to develop into a decent defender. Look at their key attributes, and see roughly how good they are. Add on a bit for development each year, and ask yourself if this player realistically could play for you in 1-4 seasons (depending on age). This is not potential as in PA, but potential as in "good attributes". Looking at the attributes in this way is equally important with fully developed players - CA isn't worth anything if they have the wrong attributes. But because players use templates for positions, the key attributes are the first thing to look at. If these are really low (below say 7 or 8, depending on your division etc. etc.), forget the player completely.

2.2) Look at attributes that hardly change. If many of these are high, the player may well have a high PA. Especially in conjunction with 2.1 and 2.3

2.3) Look at the levels of the lower attributes. Are these grouped, or spread? A player whose lesser attributes are all around the same level, but not that much (say 5 points) lower than his better attributes may well have decent potential. If the player has loads of attributes less than 5, and attributes are spread unevenly (particularly among the lower attributes with the odd stand out higher one), 9 times out of 10, he's crap. This is to do with how CA works and is spread between attributes, and an indicator that CA is already relatively close to PA, and PA is therefore low. Obviously not foolproof, but in general allows you discount a lot of players.

Is there a link where this is explained? Interested in knowing more...

3) Look at their stats; how well are they performing in the U-18s/reserves? Not as good an indicator in the U-18s (where flat track bullies can be quite effective), but a possible indicator of potential.

4) Look at their contract demands. Do they want "hot prospect"? How much money do they want? If they are asking for noticeably more than other young players, they probably have decent potential (remember, the AI "knows", so let it tip you off).

5) Check out the player profile, to see if the player is "promising" or a "wonderkid"

6) Check out their training development, if they've plateaued, and you're playing them regularly, they're probably crap and have reached their PA. Give them at least a year though.

7) Take all of the above, plus whatever else comes to mind and mix it all up. If all your coaches (at least with decent JPA) are raving about a player, he's got great primary attributes for his position, other attributes aren't too low, he wants a hot prospect contract etc. etc. then yes, he's probably got potential. Quite how much - only time will tell Which is how it should be, as you should never no for sure what the future will bring.

How does this match-up to real-life though? Talent is more innate. Take a midfielder, say Jock Walshore - he has good flair, creativity, passing and technique. But this is CA-determined, not PA-determined, and there exists a player with low PA and there exists a player with high PA such that his current set of attributes is equal to Jock Walshore right now - in other words, CA and PA are correlated but they are not necessily causated to any high degree. Another thing is that Walshore tries difficult things during games - but high flair doesn't mean anything.

All the key attributes unfortunately describe CA - it is not a discriminant. It is more like "if they are talented, then their nice attributes will become better" - talented implies better future. However, talent only partially corresponds to how good a player is now - talent is more innate and I am not convinced the match-engine shows all of this in great detail.

For the "promising" description, it is simply a description that is not a function of the manager - if the player has a high CA and a decent CA-PA gap, he gets that description, no matter how supposedly good the manager is at rating players. Ditto wonderkid. You don't find managers lookup up the player profile on some computer database and seeing the description "wonderkid" next to him to judge whether this player is a wonderkid - this description of "wonderkid" lacks a "why" that is a function of either aggregated scout, coach AND/OR manager reports.

So yeah, on day 1, you won't know that player X is great (and why should you?). But it should only take a year or 2 before you know which of your U-18s are worth a punt and which aren't. Many won't quite make it, even the better ones, especially if you're playing as a top team. You can of course ignore all the advice available to you, rant about it online, etc. etc. but in that case using Genie Scout really would be better. Or you can use the tools at your disposal. A bad workman, and all that

You should know because the player has likely "been at the club" longer than he was generated. It is not like every single one of your youngsters who comes through has joined your club on exactly July 1st (or whenever the regen date is). The generation date stops the game from modelling extremely young players (which is more realistic, but has all sorts of issues).

You do have tools, but I have mentioned that they are all not that realistic. The descriptions and contract all beg the question (they have high PA hence they get the moniker "wonderkid" or "promising" - but I want to know the reverse implication - what tools do I have to determine PA? If I use "promising" as a discriminant, I will be suggesting "promising" implies high PA implies "promising", which is a circular reference). There is something "more" that determines talent beyond things like this.

Appreciate the time taken to explain this, and if mentioned above, if you have a link to the attribute distribution then could you ping that across? Many thanks.

[Vangelis21]

i was sure that a topic like that would have so many replies! All i can say is play as you like. no one can dictate to you how you should play. just enjoy. my personal opinion is that only fmrte is cheating. enjoy the game

[sjm]

The way I form my own opinion is by doing what I said before. I don't see why we should have a button that does it for you...

As I said ages ago, we don't have a tactical attribute either as managers, as how good or bad we are tactically is determined by the tactics we use. Equally, our ability to judge how good a player is reflected in the players we sign. But from what I can tell, you seem to think their should be a "Judge this player" button that you press and then tells you what you think of a player. I don't see the point of that, much like I don't see the point of a button that magically provides you with a good tactic.

To take your chess analogy: you're playing chess against the computer, and always ask the computer to suggest your next move because "as a chess player, you should have a good understanding of playing chess, and it would be unrealistic to make mistakes". Which is quite obviously nonsense Either you're good at chess, or you suck; either you're good at judging players in FM, or you suck (or are somewhere in between). Some managers have it, some don't (RB signed loads of so-so players, SAF and AW have a much better track record, albeit with some high profile mistakes as well). There's no manager out there who knows what'll happen next week, let alone in 4 years.

Remember: it's a game, so it's an abstraction, so of course you can't do what AW and SAF do. That's why you get star ratings and scout reports. And SAF and AW don't have a list of numeric attributes that represent players either, but no one complains that that's unrealistic. Just because it's an abstraction, and you can't go about things in exactly the same ways a RL, doesn't mean you can't form your own opinion on a player. Really. Just do it. You might be wrong, but at least you'd discover that it is possible without someone doing it for you

As for the attributes, I'll go into that in more detail at some other point when I'm not taking a quick break at work.

[Sharps]

At one time I would have regarded the practice as having a distinct advantage over the game itself in that you would know which regens to try to snatch at the start of each season etc which could be equated to cheating.

However, having done a few tests with it during several careers I have seen many cases were for example, John Smith, 17 year old with a current and potential abilities of 95/170 has 5 seasons later only gotten his current ability upto 120 meaning he has missed the boat to get upto 170 and into the world class bracket.

It is fairly relative to the clubs their development takes place at, how many games they get etc. While it is far easier to judge by genie scout, I wouldn't call it cheating as it's still not a sure thing.

[Maidel]

It is fairly relative to the clubs their development takes place at, how many games they get etc. While it is far easier to judge by genie scout, I wouldn't call it cheating as it's still not a sure thing.

But if you hadnt used genie scout in the first place would you have kept that player? If he didnt seem to be improving and possibly the scout/coaches reports were unfavourable then you would have sold him. But you didnt you held on to him because the hidden factors told you he could be great.

[scooter28]

Its cheating yourself but play it as you want only my opinion

[keeper tom]

Who cares !! ?? Do What you want. Its your game and you can play how you want to !

[Sharps]

But if you hadnt used genie scout in the first place would you have kept that player? If he didnt seem to be improving and possibly the scout/coaches reports were unfavourable then you would have sold him. But you didnt you held on to him because the hidden factors told you he could be great.

If i went out deliberately and put the players into an academy of my own with a load of other high PA players, they'd probably perform and develop better therefore not offering an accurate sample. Also, if i was managing a team myself, i'd also affect the development of a player, whether it be positive or negative.

Don't get confused. I never said i actively use the tool. I stated that i'd used several careers to look at the merits of the system and decided that the PA wasn't a sure thing.

[x42bn6]

As I said ages ago, we don't have a tactical attribute either as managers, as how good or bad we are tactically is determined by the tactics we use. Equally, our ability to judge how good a player is reflected in the players we sign. But from what I can tell, you seem to think their should be a "Judge this player" button that you press and then tells you what you think of a player. I don't see the point of that, much like I don't see the point of a button that magically provides you with a good tactic.

To take your chess analogy: you're playing chess against the computer, and always ask the computer to suggest your next move because "as a chess player, you should have a good understanding of playing chess, and it would be unrealistic to make mistakes". Which is quite obviously nonsense Either you're good at chess, or you suck; either you're good at judging players in FM, or you suck (or are somewhere in between). Some managers have it, some don't (RB signed loads of so-so players, SAF and AW have a much better track record, albeit with some high profile mistakes as well). There's no manager out there who knows what'll happen next week, let alone in 4 years.

It's not a gameplay decision made, though - by hitting this button, you don't gain any suggestions on how to train the player, or how to play him - you just get latent information. It just exposes more attributes.

Hitting the button is essentially a "give me more information" button to help you make better decisions, rather than a button that says "make a good decision for me".

After all, it is going to be similar to asking your scouts to "give me more information" about a player by asking for a scout report - this isn't pointless nor controversial.

Remember: it's a game, so it's an abstraction, so of course you can't do what AW and SAF do. That's why you get star ratings and scout reports. And SAF and AW don't have a list of numeric attributes that represent players either, but no one complains that that's unrealistic. Just because it's an abstraction, and you can't go about things in exactly the same ways a RL, doesn't mean you can't form your own opinion on a player. Really. Just do it. You might be wrong, but at least you'd discover that it is possible without someone doing it for you

Of course you can form an opinion but like I said, you have very little information compared with scouts so any opinion is really a ballpark estimate rather than an educated guess. In addition, what we have now is not a realistic abstraction.

For example, when judging a player surely someone who has lots of odd spikes in perhaps weaker attributes should be rated better in real-life because it shows he has other talents outside his current position - i.e. a player with 1, 1, 1, 8, 8, 1, 1, 1 as his weak attributes would be considered marginally better than someone with 3, 4, 3, 4, 5, 3, 4, 5, say. Therefore a low variation in weak attributes may not be considered a realistic barometer of potential.

I'm not a fan of letting the imagination fill in what is a flawed model - we should be wanting a more accurate model and letting our imagination dictate how we play. In other words, this "Judge This Player" button will make the game easier but will allow for us to use our imagination and concentration elsewhere. Similar to the new tactics wizard - it makes the game easier but people have made lots of fun and simple tactics, investing their energies into making the tactics rather than tinkering with sliders.

Which is why I believe you've misinterpreted my chess analogy. The true realism of chess is the ability to see the entire board and all the pieces on the board. Genie Scout is being able to read your opponent's mind and seeing what he will do next in the next 20 plays (too much information), while the current game is you playing chess blind (too little information). The question is whether the first scenario is more realistic than the other.

This is opposed to asking the computer what is the best decision to make. If you like, I'm asking for more information to allow me to make better decisions, rather than a better decision - because the information is what I need.

I suppose what I'm asking for is something like this:

======================================================================================
Player                 "CA"      "PA"      Morale      Condition      Value    ...
======================================================================================
Lionel Messi            [color=gray]*******[/color]   [color=gray]*******[/color]   Very Good   80%            £40m
Emile Messi             [color=#000]***[/color]       [color=#000]*******[/color]   Very Poor   100%           £1m
Darron Gibson           [color=white]**[/color]        [color=white]**[/color]        Excellent   30%            £500k

Here, the "star" ratings are coloured and are "transparent" to "opaque". "Transparent" in this case refers to Darron Gibson - this reflects the fact I haven't scouted him but I know from various bits of information (i.e. young player for Manchester United, agent talk, having watched him during scouting a different Manchester United player (and it so happened Gibson came on), having listened to the gnashing of teeth for various Manchester United fans after seeing Gibson disappear against Southampton...) - i.e. I don't have lots of information on Gibson, so my decision should be taken with a pinch of salt. If you like, a white set of 7 stars is probably about as meaningful as a white set of 2 stars - very rough.

The "grey" stars are for Lionel Messi. I haven't scouted Messi but everyone knows he's a magnificent player - and I know this because he has a high reputation, everyone drools over him, editorials canonise him and agents are always looking to get him on their books. If I scout Messi, I might see my ratings go up or down a star or two because I get more information and my opinion may become more accurate, depending on how good I am at judging talent and ability.

The "black" stars is the opaque one and refers to a player whom I know well for various reasons. I've watched Emile Messi (Lionel Messi's evil twin) for several matches and know his agent well; he was on our books as a 14-year-old and we have seen him train. My opinion of Fat Messi is considered rather "accurate".

No matter how bad the scout is, eventually the scout will predict a player's PA with a high degree of accuracy - but it will take the poor scout a longer time to figure it out. The best scouts can do it at a click of their fingers - the worst scouts need to look at him for several seasons. In other words, the speed at which the stars go from white to black depends on how good the person rating the player is at judging ability and talent.

This is a bit like attribute masking.

The ideal scenario is that I can look at a list of 1000 players and quickly identify which ones are rubbish and which ones are worth looking at. If I look for all midfielders in Spain I may only have smatterings of information about each player but I can rule out quite a few of them immediately - worst midfielder at the worst club in the division? Probably not (two white stars). A young Argentine who has broken scoring record after scoring record worldwide? Probably (six white stars).

And the star usage is only an abstraction - it could well be a number that is like PA instead of stars - more granular - more realistic? That is a game design problem...

[Maidel]

This is opposed to asking the computer what is the best decision to make. If you like, I'm asking for more information to allow me to make better decisions, rather than a better decision - because the information is what I need.

I suppose what I'm asking for is something like this:.

So, basically, what you are asking for is an 'easy' mode for the game.

[x42bn6]

So, basically, what you are asking for is an 'easy' mode for the game.

No, I'm asking for a new dimension to the game.

If we had the ability to judge talent in-game and the ability to develop scouting abilities over time, then the game's complexity would move from an unrealistic and mundane forcing of scout report generations and suspended imagination in place of information we don't have, to a manager being able to use his own "virtual" judgement to judge a player's potential.

It makes the game more "accessible" (sometimes misconstrued as "easy"), and moves the difficulty on to something else more interesting.

Look at the tactics wizard - it's made tactics easier and arguably more realistic because it abstracts away from unrealistic sliders - but the difficulty has moved on to things like morale, team talks and "tactics-talk" rather than "sliders-talk" and various complex rules. The game is now easier to get into - but the game is not necessarily easier in itself.

The same thing applies here - you will be faced with the dilemma whether to trust your own judgement or consult your scouts for more opinions, whilst opening up the opportunity for yourself to act as if you were a scout yourself - after all, you can perform your own "scouting" operations. The difficulty moves on to something else.

[Maidel]

No, I'm asking for a new dimension to the game.

I dont see it that way.

1) Managers may well be able to spot talent - but they arent the ones that actually find the players. Sure, if you play in the same league as the player you are considering then you will have a fairly good idea of his talent, but what you are describing is effectivel having a huge database in your head of players 'abilities' which is utterly divorced from the real way it works.

2) Scouts are sent out to find new players. Chances are the manager has HEARD of the players they find, but unless they have played them in the champions league (assuming they are foreign) or in a cup game (if they are different domestic league) then he most likely hasnt seen them play.

3) The scout gives his opinion on the player and if the scout is jumping up and down about how good this player is then the manager will watch some dvds of him or even go and see him personally. But it would appear to be very rare for a manager to suddenly stand up one day and say - 'wow there is this wonderful new player in outer mongolia who I think will be a talent'. Thats entirely the wrong way around.

4) the tactics wizard WAS an improvement purely because it moved things away from the made up world of sliders into the world of 'words' eg, rather than having a player set to defensive 4 and roaming 1 in position DMC - we now have 'anchor man' which makes infinate more sense.

5) What you are describing is a step back from realtity into 'game mode'.

[x42bn6]

I dont see it that way.

1) Managers may well be able to spot talent - but they arent the ones that actually find the players. Sure, if you play in the same league as the player you are considering then you will have a fairly good idea of his talent, but what you are describing is effectivel having a huge database in your head of players 'abilities' which is utterly divorced from the real way it works.

Is it? Without having seen a, say, League One player, I can make a very rough estimate of a youngster's talent based upon how much "agent-talk" there is about a player - look at Oxlade-Chamberlain, for example. It's extremely rough but it does exist.

The human mind does not store an entire database - I don't know every single Premier League player, let alone League One player. It does know things like players who get a lot of hype and the best players who pop up frequently in news and matches. But regardless of how this works, the manager can infer - if not know - things. How accurate they are is dependent on ability and duration of scouting - which is why first-impressions should be very rough.

2) Scouts are sent out to find new players. Chances are the manager has HEARD of the players they find, but unless they have played them in the champions league (assuming they are foreign) or in a cup game (if they are different domestic league) then he most likely hasnt seen them play.

Managers do scouting all the time. Fergie was at some match in France today.

The fact that it's unlikely suggests that it's possible.

3) The scout gives his opinion on the player and if the scout is jumping up and down about how good this player is then the manager will watch some dvds of him or even go and see him personally. But it would appear to be very rare for a manager to suddenly stand up one day and say - 'wow there is this wonderful new player in outer mongolia who I think will be a talent'. Thats entirely the wrong way around.

It generally doesn't happen because the manager needs to be with his team - his job is more working with the players, less going to Mongolia. It does not, however, stop Fergie from flying to Mongolia to watch a game - it's possible, although it tends not to happen.

'Arry was in Spain the other day, when he got mugged.

As for this Mongolian, the game can model this accurately - it is difficult to obtain scouting knowledge of Mongolian players, scouts would avoid Mongolia as it is unlikely to produce talent for a high-level team and the travel costs would be silly. Therefore the chances of a manager popping up with an idea that a Mongolian will become the next Messi is exceedingly unlikely - no Premier League staff member will be watching Mongolian football, and anyone who is silly enough to do so will only find one Lionel Messi every 100 years anyway - what's the point of sitting in some hotel in Ulan Bator waiting for the next Messi to pop up? There's no business justification for scouting in Mongolia for a Premier League club.

4) the tactics wizard WAS an improvement purely because it moved things away from the made up world of sliders into the world of 'words' eg, rather than having a player set to defensive 4 and roaming 1 in position DMC - we now have 'anchor man' which makes infinate more sense.

Isn't that the point? Different interface, different accessors, different gameplay.

5) What you are describing is a step back from realtity into 'game mode'.

If anything, I think only scouts and coaches as being able to judge PA is less realistic than reality.

----

This does have a basis in realism - Bayesian inference. You can think of it like a normal distribution - the more information you have, the thinner the normal distribution becomes - when you have perfect information, the normal distribution is infinitely thin. The starting point will have a very wide distribution suggesting just about anything could happen - but you have a very rough idea based on things like current position in the league and "agent-talk".

[Maidel]

Sigh.

Going round in circles again and you bring in 'bayesian inference' which actually has nothing to do with the conversation other than to make your posts look intelligent.

Im not going to reply point by point - there is no point...

I dont think what you are suggesting is realistic, nor does it improve the game play significantly.

You do.

Impass.

[x42bn6]

Did you actually read the article? Bayesian inference is exactly what I am talking about when I mean "more information".

http://www.vosesoftware.com/ModelRiskHelp/Analysing_and_using_data/Bayesian/Introduction_to_Bayesian_inference_concepts.htm

Here, the parameter is PA.

The more informed we are, the more focused the prior distribution will be:

I'm not dressing things up here. Please don't assume bad faith and say I'm doing so. I'm a mathematician who has studied this stuff at a fairly high level (not PhD, though).

[Maidel]

Did you actually read the article? Bayesian inference is exactly what I am talking about when I mean "more information".

I'm not dressing things up here. Please don't assume bad faith and say I'm doing so. I'm a mathematician who has studied this stuff at a fairly high level (not PhD, though).

Ive also got maths and further maths A-levels and a degree (and part of a masters) in quanity surveying - which is LOTS of maths, so lets not play 'whose is bigger'

What you are saying is 'true' but utterly irrelavent in a real world context, its entirely theoretical and has no direct comparason to 'realism' or 'fun' both of which are important in this game.

[x42bn6]

I wouldn't say it's entirely theoretical. Scout reports work just like Bayesian inference - the more scout reports, the more accurate they become. You aggregate the scout reports - the more aggregation, the more accurate your scouts' opinions collectively become. So is this realistic?

[Maidel]

I wouldn't say it's entirely theoretical. Scout reports work just like Bayesian inference - the more scout reports, the more accurate they become. You aggregate the scout reports - the more aggregation, the more accurate your scouts' opinions collectively become. So is this realistic?

Yes.

And that has relavence to your point how? You keep going on about the managers, not the scouts.

You just think that you should know everything about every player and that somehow based on that theoretical model somehow that would make the game better.

[x42bn6]

Yes.

And that has relavence to your point how? You keep going on about the managers, not the scouts.

You just think that you should know everything about every player and that somehow based on that theoretical model somehow that would make the game better.

Because I don't see why we should distinguish between a scout and a manager in what they can do (and not do).

Just because they have different roles doesn't mean a scout can't act as a manager and vice-versa.

They may not be as good in the other role as their most comfortable one, but this shouldn't stop them from trying.

In other words, whatever a scout can do a manager can also do - although perhaps not better (and perhaps not worse).

[Maidel]

Look - Im going to bed.

Sometimes I wonder if you actually play the game, or simply think about it all the time.

Scouts can act as managers - if they get a job as a manager.

Managers (including us) act as scouts all the time - you dont need to be able to 'see' his PA in order to make a judgement about him - as many other people have pointed out.

But none of this actually has any relavence to the discussion, as I have said many times before you are far happier arguing semantics, than you are actually discussing the point.

Genie scout, according to everyone in this thread appart from you and the thread starter, is NOT REALISTIC. (and in many other peoples words is cheating/cheating yourself).

Not once have you managed to put forward a constructive arguement as to why it is realistic, despite wheeling out random mathematical models, none of which disprove the fact that genie scout allows you to be omnicient and clairvoyant, which cannot under any circumstances (unless you actually are God) be called realsitic.

[x42bn6]

I'd dispute that but I'm fairly bored of this thread.

Truce?

[Maidel]

I'd dispute that but I'm fairly bored of this thread.

Truce?

*Waves huge white flag*