Jump to content
Sports Interactive Community


  • Content Count

  • Joined

  • Last visited

Community Reputation

1 "What we've got here is a failure to communicate"

About x42bn6

  • Rank
    Development Squad


  • Biography
    Manchester United armchair fan, glory hunter, geek, virgin

About Me

  • About Me
    Pro-consumer, anti-DRM. Never be satisfied with any answer. Dig until you drop.


  • Interests
    Shouting at the telly whenever Anderson loses the ball, programming, working on my blank website

Favourite Team

  • Favourite Team
    Manchester United

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. I want the game to treat the researchers' estimates as precisely that - estimates - that can go up and down in-game. It is across all FM versions. The very concept of a static PA is wrong. It would be wrong in any version of FM, or any version of Championship Manager, or any football game. I don't think PA changes should be based on a few games. Researchers don't base their changes on a few games. It would be based on longer-term statistics, i.e. how well a player does over a season, rather than a few games of good form. Nor do I think the researchers' opinions should be ignored. Their opinions should be well-considered at the start of the game, but become less and less meaningful as the game progresses, in the same way that a researcher's opinion of Cristiano Ronaldo at 17 is probably very different to that same researcher's opinion today. This is the bug! The game should not generate the future until it is actually simulated. The game has made a very strong assumption that the researcher's opinion is infalliable wasn't underestimated. The same researcher may well disagree with that rating himself/herself after a year. The game is in a position where it can make an objective reevaluation of a player, so it should do so. As stated above, it doesn't matter. This issue is across all FM versions and even across any football management game that has a similar concept. It was a simple example. I know CA weightings aren't that simple, but the fact remains that CA is a weighted average that constrains players' attributes. There is, however, a very clear difference in the way that an 80/80 player develops and an 80/200 player. An 80/80 player cannot have the same overall quality as an 80/200 player at their peak simply because a CA 200 player is overwhelmingly-likely to have a better set of attributes than a CA 80 player. If it isn't a huge limiting factor, then why not just give Messi 80 PA? You're conflating CA with PA. You can make Bale play in a variety of ways with a variety of performances in the present, but that is CA, not PA. PA refers to his overall future development. On average, it is clear that a young player with PA 140 is more likely to perform better in the future than one with PA 100, irrespective of roles or underlying attributes ("on average"). You can mitigate the latter's performance with good training and more appropriate roles, but I don't think you should ever need to mitigate anything. If these two players are equivalent except for PA, whatever you do to make the PA 100 player amazing can be done to the PA 140 player with even better results. The reevaluation of a player's estimated peak has little to do with the roles you give him today.
  2. I agree that attributes are the real story, but the fact is that PA is a huge limiting factor. A player who is 80/80 has extremely limited scope for development - he effectively has to lose attribute points somewhere to gain them elsewhere. For example, if an 80/80 striker is to develop, he may have to lose 1 Tackling point to gain 1 Finishing, while an 80/100 striker can just gain 1 Finishing. In my mind, this doesn't make sense because why does a player have to get worse in one area to get better? Because a researcher thought he wasn't up to it? Maybe it's true that each person has a set potential, but it still doesn't change the that this isn't the value that is put into the database. It's an opinion of a person that is put in. So I don't actually see your opinion in this respect and moving PAs as contradictory at all. If someone has a set potential, another person's opinion of that value doesn't actually affect it. Cristiano Ronaldo isn't a lesser player today because a random person on the Internet thought he was a showboating crybaby at 17 rather than an amazing talent. I simply don't think the game should take the database value as the value to use. It's a very strong assumption - one we know is wrong - and the game would be more correct if the assumption was relaxed or even removed.
  3. It doesn't get ignored. The input value is taken as the most accurate estimate as of that date, and the estimate is reevaluated over time as the game sees fit. Put it this way - if a researcher increases a player's PA by, say, 10 points, the circumstances behind that PA rise, should they happen in-game, will also result in a similar rise. A crude example - 30 goals + 30 assists + an average rating of 9.00 = 25 point rise - whether in-game or by a researcher. Why are you trying to derail the thread with my thoughts on the series? We're talking about PA here. If you are desperate to know, maybe one day I will setup a blog with my thoughts and PM you the link. If you think my opinion on Valve is going to sway my thoughts why a fixed PA is unfalsifiable and therefore incorrect in the game, you are wrong. They are two separate concepts and you know it. And I'm disappointed you're accusing me of piracy. Let's not sink that low, thanks.
  4. In that case, the root cause - that is, the match engine exploit - should be fixed. Players developing too well is a sympton, not a cause. I don't think any one of us can answer that question. Not every bug is fixed and there are many reasons why this happens. But the important thing for this forum is to describe bugs and enhancements - and let SI triage it. Triage is not our concern. Why, just because of dynamic PA? Talk about throwing the baby out with the bathwater. I do write code in my free time but even there I have better things to do. It is the researcher's perceived potential that goes into the game - hence, it should go up.
  5. I sort of covered the last sentence with my goalkeeper example. However, let's take a player who is like Messi (except for ability)... Let us say that Manchester City in reality decide that thanks to an unlimited money supply, they have the facilities and coaches to turn anyone like Joe Awful, a winger in non-League, into Messi. Joe Awful is signed for a bargain fee of £100m from a club on the billionth tier of the pyramid. City then start to work on him. We know that first-team football is important for player development. Awful's first hurdle is therefore breaking into the City first-team. Needless to say, Awful is likely miles off the likes of David Silva in terms of actual ability, so it is going to be very hard to break in to the the first-team. Let us assume that for whatever reason, he does get his chance, such as every single attacking midfielder at Manchester City breaks their legs (extremely unlikely). The next hurdle is performing well. How many times have we seen young player have absolute stinkers while on loan, and therefore struggle to continue developing? Macheda is one for us - even when he was playing for the first-team, he was awful. Playing well opens up new possibilities because they now know what works and opponents need to try different things to combat them - which is the learning process. Needless to say, Awful is no David Silva so the chances of him playing well in the Premier League are close to zero. Think Ali Dia. Awful, being from non-League, is also naturally at a present ability disadvantage compared with his peers - i.e. his CA is going to be a lot lower than David Silva. As a result, Awful needs to have more luck on his side when it comes to injuries because he's already given his peers a headstart. Awful needs the stars to align perfectly to hit Messi levels, while players like Messi could possibly tolerate the odd serious injury or so. Awful also needs to have a fantastic attitude in training, especially since, as stated above, he is at a present ability disadvantage. Players who have such attitudes are rare. If they don't, he needs to change his attitude - the game has tutoring to help that - but tutoring may not always work, and it takes time - time Awful really doesn't have. Without a good attitude off the pitch, he risks turning into a Cassano rather than a Messi. Therefore in-game, even if you gave every single player an effective PA of 200, a well-built game should still demonstrate that Joe Awful's chances of making it to Messi's level are extremely rare - without actually holding them back. Joe Awful simply needs too much luck and then some to get there. A manager needs to deliberately try to get Awful up to those levels - when in reality, every manager would simply use the same amount of effort for a likely better reward by signing players above the billionth tier of the pyramid. Well, the OP has mentioned Bale in a previous iteration of FM, which demonstrates it is broken. Both of you (you and Loki679) are correct when it comes to the 50/190 player. The player is 28, so in order to hit CA 190, he needs to start developing and performing - yesterday. The odds against the player hitting CA 190 are so unlikely it really doesn't matter what his PA is (as long as it is greater than or equal to 50). However, PA in this case is less useful because players naturally start declining at that age. A better example would be a player who is CA 1, PA 200 at 18. Because his CA is so pathetically low, there is a good chance he'll never get anywhere near CA 200 at his peak (namely: he'd probably be the worst youth player in the team, lucky to even get on the bench for the team, and he'd play awfully if he did play). In this case, a movable PA would help: The researcher would give him a PA of 50 or something - but under the assumption that this player does get a chance (i.e. every rival in his position gets injured), and under the assumption that he defies all the odds, he performs and develops well (i.e. Bale), he "deserves" a PA rise. Maybe to 70 or something. Maybe one day he will continue performing to the extent that he could hit 80... 90... 150... 200? The amount of luck required increases dramatically, as you raise the PA, though. This would fix the Bale scenario. Say Bale had -8 PA at 16, which generated a PA of 140. For a number of years, his development is ho-hum, maybe more high-profile than your average youngster, but ho-hum nevertheless. Suddenly, a manager realises that he might be a better winger, and starts giving him a try there - and he starts to deliver. The game sees that his performances are better than his PA warrants, and nudges it up a tad to 145. He then starts scoring and assisting for fun, burning world-class full-backs and ruining goalkeepers. The game also sees this, and bumps his PA up to 160. He then scores nearly 30 goals in a season and single-handedly wins too many games to count, and draws the attention of a certain Madrid-based club. The game bumps his PA up to 175.
  6. I'm not confusing the two. There are actually two forms of perceived potential ability - one that is the process used by a researcher or generated via a PRNG for regens; and an in-game mechanic that lets a scout, say, judge a player's possible peak by a variety of factors like current ability and form. Barside brought up the latter, while I'm actually interested in the former. The first thing to note is that the actual value isn't what's in-game. We know this because it often changes between FM iterations, and even within an iteration between data patches. Indeed, the editor even lets you enter negative PAs to give you a range. It's not a fixed value. This limit isn't a moving target or range in reality, is it? If it were the true value, you could never change that value ever between FM data releases. Mathematically, you could describe what gets entered into a database as the upper bound of the n% confidence interval (and n can never be 100 without a crystal ball). What actually gets entered is the estimated potential by a researcher, who uses various strategies such as watching games and gathering statistics to put a value into a database. In other words, it's a guess. An estimate. It's not the real value - it can't be, without a crystal ball. It's an estimate and should be treated as such - including the possibility that the researcher underestimated the value. There is a mechanic in reality where a researcher will adjust a player's PA in response to real-world events, such as a ridiculous scoring burst like Bale, a serious injury, or just even plain hype (or otherwise). These even can happen in-game too - so why not replicate the mechanics in-game as well? Because estimated potential goes up and down in reality, it should go up and down in-game too. Just like Passing or Finishing can go up and down as a researcher deems fit. This is not as controversial as it sounds! True, the game does have perceived potential (the former definition in my first paragraph), but this has no impact on the player's real development. It's like a scout report - but scout reports can conflict, and if I think a player will turn out to be rubbish, it doesn't actually affect his peak. My mind isn't that powerful. Does it mean that all players have the potential to be the best in the world? No. However, if you were asked to model this statement in a piece of software, I would say that the most correct thing do is to say "yes". If you answer "no", you would be able to pick out with certainty which players won't be, but you don't have a crystal ball. This isn't a contradiction - like I said above, you don't put the real value in the game. You put an estimate. This doesn't mean that a moving PA will result in every player becoming Messi. Consider goalkeepers. There are only so many top-class teams in the world, and each can only play one, maybe two (i.e. Real Madrid) goalkeepers in a significant number of games. We know that first-team football is important for development (i.e. players going out on loan) and we know that world-class training facilities like those at Real Madrid are also key to development. However, because of the limitation on the number of goalkeeper positions on the pitch, it is impossible for every goalkeeper in the world to turn out to be world-class. Unless every single club becomes world-class, in which case the worst world-class club isn't actually world-class in relative terms any more. The same applies to all positions - a club can only have so many players, and only so many players can be considered part of the "first team". So even if you gave everyone 200 PA, the game should still result in a good spread of players - some world-class, but most not-so-much. Are there genetic factors that influence development? Of course! However, until the game is going to model stuff like VO2 max and muscle fibre density, I think these low-level biological factors are more of a question for scientific research than a game. And these low-level biological factors don't necessarily imply a limit from a footballing perspective - football isn't a true endurance sport along the lines of marathons or professional cycling, so VO2 max may never be fully-utilised in football. The fastest sprinter in the world isn't the best footballer in the world, and arguably wouldn't be. The most intelligent person in the world isn't the best footballer in the world, and arguably wouldn't be either. They are just attributes of a player - just like Strength or Passing are - the higher, the better, both today and tomorrow at your peak.
  7. Wouldn't it therefore be better if he could, in-game, beat all the odds and become a world-class player - regardless of how people perceived him at the time?
  8. If a player has a low PA, then it doesn't matter what his perceived PA is - no matter how many 10.0 ratings he gets, or how professional he is in training, or whatever. He will always be stuck.In reality, if he started outscoring Messi and kept training well, he would get a healthy PA boost. There is a mismatch between reality and in-game - this is the bug.
  9. Headbutts I'm less sure... Accidental "bunts" of the head maybe, but headbutts should be in a similar category as biting, personally.Either way, I think a better example would be from contact sports like boxing - boxing is about as close as you can get to battery (not really assault - battery is closer to the actual offence). Even Tyson's ear bite on Holyfield wasn't a criminal offence. However, the moment the fight is over, consent is over - as James Butler's sucker punch demonstrated (he was later charged with battery).
  10. Why is varying the PA over time not an improvement? It allows the game to model the uncertainty more accurately.Nobody cared 5 years ago, but wouldn't be even better if you could go back go FM09 or whatever and mimic Bale's meteoric rise? It is just that in this case, you would fix it for FM15 and see the benefits in 2019 or 2020 or something.
  11. Even if it's his third bite, are FIFA permitted to use that as evidence? The Bakkal incident wasn't cited in the FA's report when he bit Ivanović as far as I can see, for example. To the English FA, it was his first time. If they were to use his previous bites as evidence, they would need to call up the Dutch and English FAs. I'm not sure this has happened. And it could be a slippery slope where if FIFA wants to nail someone with a case, they can always ask for more incriminating evidence elsewhere. If FIFA can consider it as his third time, and are allowed to ban him from club competitions too, what is fair? Personally, I'm veering towards somewhere between 3-12 months, although I'm not sure where exactly is fair. 3 months is about 15 matches and I can't help but think that that's a bit lenient for a third offence, so maybe 6 months at a minimum. If they are allowed to enforce it, a mandatory psychological evaluation too. I think life bans are maybe a bit harsh although for a third offence, not as stupid as it sounds. In reality, I think he'll probably get 6-8 games in FIFA competitions and a giant fine. Liverpool will sanction him separately, probably dock his wages for a few weeks and make him go for a psychological evaluation if FIFA don't tell him to. People will cry foul, but FIFA won't listen. The FA will look into the case to see if they can enforce something further, but find nothing.
  12. It's the perceived potential that goes into the game (we know this because it goes up and down between FM versions - see the aforementioned Bale), so it should go up and down in-game too, depending on circumstances. So if Bale suddenly goes on a 20-goal scoring spree in-game, it should go up just as it did between versions. The value that goes in has a degree of uncertainty and the uncertainty is due to the fact that a player can become better or worse than expected. So if the player does turn out, in-game, to be better or worse than expected, it should shift correspondingly.
  13. His grandfather has a shoulder. He'd never bite anyone's shoulder.
  14. I don't see why this needs more processing power. It just looks like a disconnect between how good the grey players are versus the real players, where loading up a country means more non-greys are generated. So maybe the grey player strength needs to be strengthened, for example. Or maybe the game too-easily calls up non-grey players (including weak youngsters) if the team is not loaded (so is predominantly grey), so the national team is actually weaker than reality as a result.
  15. I had a shocking 5-1 defeat to a team in the relegation zone. Awful performance really, and well-deserved. I also lost 24-0 in FM08 where a match managed to start without any of my players on the pitch. Very weird. Suffice to say, I reloaded this match.
  • Create New...