Jump to content
Sports Interactive Community

How long would you endure to wait for a "Transfer AI Update" twice a season?

BiggusD

By BiggusD,
in Football Manager General Discussion

How long would you endure to wait for a "transfer AI update" twice a season?  

69 members have voted

  1. 1. How long would you endure to wait for a "transfer AI update" twice a season?

    • <1 minute
      9
    • 1-5 minutes
      17
    • 6-10 minutes
      19
    • 11-20 minutes
      14
    • 20 minutes - 1 hour
      4
    • >1 hour
      6

Recommended Posts

BiggusD

BiggusD

Posted
Posted

I have suggested many times that SI implements an actual deadline/update where each AI club in the world decide who to sell and who to buy, as a possible improvement to the current opportunistic system. Twice a season, just before each transfer window, short and long-term plans would be laid and I believe this would go a long way of solving the AI squad building issues as well as the lack of bids for players at human-controlled clubs.

Suppose that SI implemented this feature and that it fully fixed any current AI squad-building and transfer AI issue, how long would you care to wait for that progression bar to go to 100%?

Link to post
Share on other sites
Smurf

Smurf

Posted
Posted

It's as easy as just saying "implement it now!"

What criteria would the AI use, would every club use the same or different criteria?

Would the AI just sell players with Potential under 3 stars or 4 stars or 4.5 stars?

What age group would it allow?

Would it be if 15 - 18 year old CA was all silver stars but PA of 3 gold stars - would these go on development list or transfer list?

What about 18-24 year old with 1 gold star and 3.5 potential star - would these go on development list or transfer list?

25-30 year old out injured for 6 months returns to squad with 3 CA and 3 PA stars - do they get transfered even though previously they were 4.5 star potential?

Would the AI have to analyse each player position and decide it has too many DL and sell the weakest player? Or would they offload the player most prone to injury?

What if the AI had only 1 striker but 5 AMC that could play as striker but not officially down as a striker... would the AI retrain these - or would it sell off 3 AMC because there's too many - then would it attempt to buy 2x ST to offset this?

Would the AI offload players without having players lined up - or would the AI analyse the weak spots in the team, buy in players and then sell players? Or would it sell players to fund the buying of another player?

There's a tonne of different scenarios and not all clubs would operate or sell these players or work completely different ways.

Good luck coding it - sounds like a nightmare...

Link to post
Share on other sites
Al_Pacino

Al_Pacino

Posted
Posted

I'm very impatient, but chose 6 - 10 minutes. I reckon any longer than that would see me hit the reset button in belief my game had crashed :D

For what it's worth, I finally had an unexpected bid for one of my players last night. Chelsea bid £7.5M for my young DR/C (valued at £3.6M).

I didn't want to sell him but was just so relieved to get a bid out of the blue for decent money that I accepted happily...

Link to post
Share on other sites
Kenco

Kenco

Posted
Posted

I'd happily wait an hour or more if it worked but will it? How would it work?

It can't simply be a list of players to buy and sell, there has to be a complex structure of if statements and contingency plans in case all the targets are bought by other teams or they sell a player they didn't originally intend to sell.

Link to post
Share on other sites
Smurf

Smurf

Posted
Posted
@Eugene Tyson and Kenco

Well there's already a process of "finding transfer targets" at the beginning of the game and if it's working properly then SI won't have to change much.

You don't know that - that's an assumption on your part.

Link to post
Share on other sites
Fancy Gaffot

Fancy Gaffot

Posted
Posted
If it's working properly then why is there a discussion to give it an overhaul?

It runs once before the game. The discussion is about making it run twice a year.

You don't know that - that's an assumption on your part.

I don't know what? Take a look at your screen when you click start new game :)

Link to post
Share on other sites
Arnoldzhu

Arnoldzhu

Posted
Posted

Oh god. I don't like potential rated with stars at all. I don't like potential either. Let alone the AI manage their squad based on valuing their players' potential rating is worst. I could buy and pile up bunch of players with high potential rating, but their current attributes could be poor and their criteria attribute may also be poor, which is not sensible, how on earth people would think they are promising to become good player in future. Contrary, some young players with good attribute distribution may only be appointed with low potential ability, that's pretty sad and disappointing and unrealistic. Back to the topic, the AI should use different criterias to make various decisions based on their managers style and as for young players, age and current performance should be considered as to their future is with the club or to be sold.

Link to post
Share on other sites
BiggusD

BiggusD

Posted
  • Author
Posted
It's as easy as just saying "implement it now!"

What criteria would the AI use, would every club use the same or different criteria?

Would the AI just sell players with Potential under 3 stars or 4 stars or 4.5 stars?

What age group would it allow?

Would it be if 15 - 18 year old CA was all silver stars but PA of 3 gold stars - would these go on development list or transfer list?

What about 18-24 year old with 1 gold star and 3.5 potential star - would these go on development list or transfer list?

25-30 year old out injured for 6 months returns to squad with 3 CA and 3 PA stars - do they get transfered even though previously they were 4.5 star potential?

Would the AI have to analyse each player position and decide it has too many DL and sell the weakest player? Or would they offload the player most prone to injury?

What if the AI had only 1 striker but 5 AMC that could play as striker but not officially down as a striker... would the AI retrain these - or would it sell off 3 AMC because there's too many - then would it attempt to buy 2x ST to offset this?

Would the AI offload players without having players lined up - or would the AI analyse the weak spots in the team, buy in players and then sell players? Or would it sell players to fund the buying of another player?

There's a tonne of different scenarios and not all clubs would operate or sell these players or work completely different ways.

Good luck coding it - sounds like a nightmare...

Well, how is this different from what is a nightmare to code right now? All those questions must be answered also for the current system...

The only difference is that there would be a time-span for their interest in a player. Now they can have a player on their wishlist for 20 seasons without ever making a bid unless you make the decision for them and offer him out. With the system I propose, all the variables and squad building logic currently in place would still be there, but there is a framework for actual decision-making on top. A list of priorities could be for instance one new player in each tier (forward, midfielder, defender, goalkeeper), with numbered transfer targets, so that they will go for no.1 of each tier first, and if they get that player all other players in that tier are removed from the list. A better manager could be allowed to pinpoint left back, right back and centre back weaknesses in the first team and therefore make priorities within each tier. When a first team signing is done, a youth prospect or backup signing can fill that tier slot but have a lesser priority than the other first team tiers until as such time all four tiers have been strengthened. This way they won't use all their £40m on a 15m first team striker and a 25m 15-year old striker.

As for which player to offload, it should depend on club and manager philosophy as well as existing interest in players - among hundreds of other variables. The point is that decisions must actually be made. There should also be a distinction between "offloading" and "looking for a buyer". The latter is when a small club has developed a star and he's interested in moving on. There should be an option in the game (for all of us and the AI) to ask a player's agent to market their client to bigger clubs, to subtly indicate that the club won't stand in the way of him fulfilling his dreams. That way, there is a "trigger" to activate decision-making of the agent feature as well.

Link to post
Share on other sites
Overmars

Overmars

Posted
Posted
I'd happily wait 10-20 mins if it meant that the game was still a challenge in the future. Can just go and make a sandwich.

That's a big sandwich! I'd wait even longer though. If I know I have to wait an hour or so, that isn't a problem. The overall speed and playability of the game is much more important than one or two predictable long pauses.

Link to post
Share on other sites
slipknot67

slipknot67

Posted
Posted
But after the first season, there are not enough transfers!

I just checked the transfer histories of the clubs at the top of the top divisions in England, Italy, France, Germany and Spain in my current game, they all brought in between nine and fourteen players in the space of the two transfer windows in a season. The problem is quality not quality and as Neil Brock already mentioned the game already has the ai managers select short and long term transfer targets, it's just not as good as the human player at doing this and is never likely to be due to the nature of coding ai into games. I don't see how your idea would change anything at all with regards to ai squad building.

For me what needs to change for long term games to improve is for the ai to be able to recognise that winning a meaningless end of season league game when the title is already won or a cup match where their board consider that trophy unimportant 10-0 by fielding their highest ca players every single time is not necessarily in the best long term interest of the squad and club as a whole and for the ai manager to better be able to select a starting 11 taking more into account than just who are my 'best' players, but as I mentioned before coding this time of reasoning into FM would be very difficult.

Link to post
Share on other sites
jayahr

jayahr

Posted
Posted
Errr...the game already pretty much does this? They have internal shortlists and plan out their long (prospects) and short-term signings.

I assumed nothing else. Can't see what such update would or could do on top of that.

That said, the AI can certainly do with some improvement in that area.

Link to post
Share on other sites
Herbert Fandel

Herbert Fandel

Posted
Posted
Errr...the game already pretty much does this? They have internal shortlists and plan out their long (prospects) and short-term signings.
I assumed nothing else. Can't see what such update would or could do on top of that.

That said, the AI can certainly do with some improvement in that area.

For me you could reword it, the AI is massively failing in that area. Big clubs sign players for big fees and never play them. Playing in the SPL I have sold 2 players to fellow SPL clubs, both are straight into reserves and transfer listed soon after, never played for the club I sold them to. Although clubs make some good signings too, the instances of players never playing are far too often. For every Bebe IRL there are a handful in the game. That, and squad rotation need massively looked at.
Link to post
Share on other sites
BiggusD

BiggusD

Posted
  • Author
Posted

Yeah, my suggestion is really to implement a more hands-on, time-specific decision-making process for the AI. Hence the question; would FM-players really want to wait longer for a more thorough transfer AI - one that didn't make those absolute suspension-of-disbelief-killing transfers at the very least. It seems that the majority of players wouldn't mind having a progress bar "update" going for a few minutes twice a season if it meant a more competitive AI.

So if SI are looking for solutions to what is actually a small problem (I wouldn't say big, that would imply that it is not fun to play after a few seasons. It is but I would like not having to take over Real Madrid every savegame and buy them a new team because they have the original 2013 team still in 2020), but are afraid of implementing a too resource-demanding AI coding, then this little poll would indicate that it is an avenue they could investigate.

Link to post
Share on other sites
Herbert Fandel

Herbert Fandel

Posted
Posted
Yeah, my suggestion is really to implement a more hands-on, time-specific decision-making process for the AI. Hence the question; would FM-players really want to wait longer for a more thorough transfer AI - one that didn't make those absolute suspension-of-disbelief-killing transfers at the very least. It seems that the majority of players wouldn't mind having a progress bar "update" going for a few minutes twice a season if it meant a more competitive AI.

So if SI are looking for solutions to what is actually a small problem (I wouldn't say big, that would imply that it is not fun to play after a few seasons. It is but I would like not having to take over Real Madrid every savegame and buy them a new team because they have the original 2013 team still in 2020), but are afraid of implementing a too resource-demanding AI coding, then this little poll would indicate that it is an avenue they could investigate.

I agree that, given that the current system is working to a level one could describe as ideal, and assuming there exists a coding possibility by which more regular lengthy calculations could increase the AI, it was a very worthwhile poll
Link to post
Share on other sites
RobertPage

RobertPage

Posted
Posted

Personally i don't think the transfers are as much of a problem as people make out. I just think that some work needs to be done just to allow the AI to recognize which areas need to be stronger rather than just buying specific targets. I don't think that these 'internal shortlists' should exist. They seem to be the main problem with the un-realistic and un-necessary signings if anything

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Go to topic listing

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...