Search the Community
Showing results for tags 'ai behaviour'.
This is a project I have been tinkering with on-and-off for quite a while. After playing FM (and CM before it) over many years, I've become more and more interested in the game’s AI – its sophistication and limitations. After several years of creating various edited databases, I decided to create an entirely new set of clubs and their own league to play in, almost entirely divorced from the parameters of the real footballing world. Part of the motivation was to create a new league to play in for a bit of change, but it was also a way of seeing how the game would react to such a situation: would it allow this new league to function in the way that it could (i.e., leverage its wealth and resources to become potentially the leading competition in the sport), or would it maintain the status quo, dominated by UEFA competitions and established European leagues? Moreover, it is an opportunity to see how a league unfolds from the very beginning. None of the clubs have history, staff (apart from some basic backroom staff and owners) or any form of legacy: they all need to establish themselves and either sink or swim in their environment using the resources at their disposal. As a follower of American sports, I took this project down the US ‘franchise’ route: closed leagues, divisional structure, playoffs, vibrant logos and ostentatious team names… I appreciate this is tantamount to heresy to some*, but if you’re interested in something a bit different, read on… *I’ll state at the outset, this isn’t one for the purists: many, many aspects of the game have been dismantled, altered, tweaked and tinkered with; the in-game editor is on and tools like FM genie scout and FMRTE will also be used to ascertain information and data: I’m not an active manager in the game and won’t be claiming any ‘achievement’ on route – this is solely about setting up a fictional scenario and seeing what happens. All the same, I appreciate that many like the game to be played ‘properly’ and will detest the very notion of this. I like to play the game in it’s traditional format too, but also enjoy having a mess about and being a bit creative. Likewise, I'm a bit of a data nerd, and reports will be, at least in part, focused on analysing data from the game - again, I appreciate that some will find this utterly pointless and an absurd way to 'play' FM... Premise: · How will FM’s AI cope under extreme conditions, namely the disruption of the underlying logic and structure of the game ‘world’? This is an experiment to find out, by creating a new league and new teams that will rival the old establishment of world football · Based nominally in Canada, a new ‘closed’ league has been created, populated by 32 franchises, spread across the world (although consolidated primarily in more economically realistic nations and cities as would be the case in a real-life situation) · Every team is extremely well financed via commercial deals and TV revenue (although relatively evenly financed, much like how US sports use collective revenue sharing), and has state-of-the-art facilities and stadiums. They are the football (soccer) equivalent of the NFL, where the ‘closed’ structure ensures their revenue and status are protected. Club reputations have been set high (between 9000 and 6500 on a ratio'd basis) so as to provide the new league with a surge of interest to kick-start it · Teams from the IFF (the Independent Football Federation) do not qualify for any existing continental or world club tournaments. Game Setup: · English, Spanish, Italian, German and French top divisions are running on full detail in addition to Canada, which is running the IFF World Championship and IFF Division II (for B Teams) · The game has 24,500 players active on a 'medium' sized database · All other nations are running as normal at game start-up New League Structure: · The IFF season is structured as follows: August – October: The Divisional Series: teams are regionally divided into 4 leagues of 8 teams – The Pacific Football Conference (PFC); The North American Football Conference (NFC); The European Football Conference (EFC); The Atlantic Football Conference (AFC). Teams play each-other twice, with winners crowned champions Additionally, the Pentland-Meisel Invitational Cup runs during August and September. This is a tournament in which the previous year’s IFF champions take on the winners of the continental cups (European Cup, Libertadores etc) November – May: The IFF World Championship starts with a group phase, in which all 32 teams are randomly drawn in pools of 8, playing each-other twice. The winners and runners-up in each group qualify for the championship playoffs (single-legged seeded knockout rounds played at the home-team stadium, with a final at a neutral venue) Running alongside the World Championship is the Intercontinental Challenge Cup; a 2-legged knockout cup featuring all 32 teams, drawn randomly To keep things interesting in these shorter format leagues, 2 points are awarded for a win, 1 for a draw and none for a loss. Initial Questions/thoughts: · How will this new league interact with the established game world? Will players and managers be attracted to the upstart league by the prospect of the money on offer? Will the AI recognise other factors such as the draw of the cities that teams are based in (Melbourne, New York, San Diego, Tokyo, Los Angeles, Miami….?) or even the relatively low number of matches they are required to play in per season? · Will there be an immediate influx of talent for big money as ambitious owners look to exploit their wealth and grow their empires? How effective will scouting and recruitment be? Will these new clubs, starting from scratch, be to ‘see’ the best players by building a scouting network? How much will this limit their ability to improve quickly? · In time, will the IFF grow its own superstars with the vast facilities and resources available to clubs? · What kind of competitive dynamic will emerge in the league? Will the ‘traditional’ football nations and regions (i.e., Europe) mimic their domination over this new league, or will North America become dominant? Will this affect the quantity and quality of youth prospects clubs can generate? Likewise, will dominance emerge from within specific regional conferences, or will competition be balanced? · Without the restrictions of FFP, will the IFF teams simply be able to outmuscle even the strongest of clubs in the European leagues, or will prestige and tradition ensure the likes of Real Madrid, Barcelona and Bayern always retain their status as the giants? · Over time, will the IFF offer a higher standard of football for the world? Will we see a consolidation of talent, with 32 highly tuned, high quality teams going head-to-head each year in the toughest club challenge on the planet, or will it simply become another retirement home for past-it stars of Europe, where lavish salaries are paid to has-beens to see out their twilight playing years in glamourous surrounds? Let's find out... Update Format: · Each Season I will do a report on the goings-on in this alternative football universe; · I will cover the winners of each tournament and provide some analysis of team strength and development using the in-game editor and Genie Scout to gather background data such as CA, PA, reputation as well as financial information · Initially, I will analysing changes to numbers of players with PA and CA bands in each of the major leagues, changes to club values, using a ‘control’ group of clubs from Europe and tracking the transfer spending of the IFF clubs and the ‘control group’ clubs. As the game evolves, there will be additional sections added and developed
Apparently FM 2020 is just not going to be unable to get this one right. I have not played since the end of December until tonight, though I had been playing and actively giving feedback since beta day 1 till then. I'm 46 and have been playing this game since the very first release of CM and this one has been the most frustrating ME I can remember, maybe because you can see the next level of immersion just over the hill and out of reach. Here is an example of what I have been talking about. (I had to crop it down and record it playing in the media player to get it down from 4K to get under the 9MB limit for the post) It crashed not long after this, first crash I've had in FM 2020 (posted .dmp in bug forums) First off, thank you for continuing to try to improve the ME. Various Match Engine (version 2039) changes including (but not limited to): - Improved one-on-one finishing - Recalculations regarding clear-cut chances - Tweaks to penalty kick conversion rates - Lowered frequency of players overrunning ball and losing possession - Improved realism of player recoveries from slide tackles - Improved crossing variation and instruction behaviour - You're having a laugh on this one mate! - Improved defensive tracking from long balls - Improved heading accuracy - Improvements/tweaks to referee behaviour - Throw-in positional adjustments - Various other balancing tweaks/improvements You know what is crazy? The best I have ever seen FM behavior from these positions was in the very first beta release of 2020! Yet the very first patch changed it and they went back to smashing into the nearest defender. Why???? Look at the acres of space Mo and Trent have to go past and then cross, or drop the shoulder and drive in past them. Never will they do it, never. And then the Inter defender slide tackling his teammate for a corner kick. No words on that. Especially Trent here when he picks the ball up, he's already at pace and the defender just gifts him the space beyond him with his poor initial movement, but oooh nooo Trent is already winding up for a blast, perfectly timed to HIT THE DEFENDER. My 11 year old nephew would've skipped right past that defender having read the space, the speed of everything happening and the options, seriously. This isn't like an bizarre heliacal event that is hard to track down, fire the game up, play with two IF/RMD/IW and tell them to cut inside, or not, to sit narrower, or not, shoot more often, or not, cross more or less, or not, cross low, pre-match talk- cross low, individual personal instructions- cross low, and they will never cross low, or none of the former, and any other combination of instructions (channels, get forward, have a bevvy) and they just continue to blast the ball into the first defender that gets near them. They will even wait until they get close and then smash it into them. I am using Liverpool not only because they are my team IRL (since '96!) but that they are the simply the best team on the planet at the moment. If I was using Rangers or Stuttgart (my other saves) one could use the 'the players just aren't that good' argument. FM 2020 Wide Attack Play.mp4 You can use the slider and go frame by frame since I rerecorded it at 30fps and you can see the defender begin his slide tackle before Salah begins his cross attempt, so Mo should have easily skipped past him. Or, has the game already calculated that Salah will cross and the defender gets the jump on him by being able to 'see the future'?
So myself and 4 other friends run a fantasy draft online save but we use the pre-game editor to do it since it's just cleaner in our opinion. We realised that as a result of our draft some teams get absolutely demolished, to the point they no longer have a senior starting 11. As a solution to this, I proposed when we run our next save (which will be when FM19 comes out) we'll compensate the affected teams as fairly as we can. To make sure this worked I tested the theory on FM18. In our draft, Napoli is the worst affected team. So by our calculations, it would be fair to give them an extra 400mil transfer budget and extra 720kpw wage budget. First time I ran the simulation, I increased their budget to 300mil since it was the max it would allow and gave them 1mpw remaining wage budget - they hardly spent any money, less than 10m, highly confusing. Then I thought maybe it's because I didn't increase their balance to reflect the new transfer budget, so I increased the balance, there was hardly any change in transfer activity. So I thought maybe they didn't have the reputation to buy anyone they deemed worthy to be bought, so I upped their reputation to 9000. They then went onto spend 25m, still ridiculous. I added manager and realised it seems the AI wasn't able to manipulate the transfer and wage budget sliders and had stupid amounts of money as transfer budget while having no wage budget left. So I did another simulation with a remaining wage budget of 2mpw. They've now spent ~65m, which is an improvement but still not really the desired outcome. I continued the game to January and they did make a lot more signings in January, totalling to 190m (minimum value). I added manager and saw they had 200m budget left with 250kpw budget left. So clearly still a lot of money left. So I've been left baffled on how to get them to spend this money, it's also not just a Napoli problem, every team I've tried to compensate has reacted like this. I did just think now that it may be to do with FFP - that's my only idea. So I've come here to see if anyone knows if I'm right and also, if I am right, how do I fix it? Attached are screenshots The screenshot shown says 1mpw remaining wage budget but that's an old version, the latest version has the 2mpw.