Jump to content
Sports Interactive Community

Svenc

Members
  • Content count

    3,724
  • Joined

  • Last visited

  • Days Won

    1

1 Follower

About Svenc

  • Rank
    Semi Pro

Biography

  • Biography
    "At the end of the day, all we can do as humans is create a tactic which dominates possession, creates clear cut chances and gets shots on target." -- perceived football wisdoms of an eternally to be frustrated Football Manager.

About Me

  • About Me
    Germoney

Currently Managing

  • Currently Managing
    Dishonored II

Recent Profile Visitors

7,121 profile views
  1. Any i3 on the market (even the first ones released 2010ish) is multiple times as fast as the 2001ish Pentium 4 being the minimum target in the meantime. Then again, Out Of The Park Developments, are still supporting P4s too. Plus, who knows how "good" the game actually runs on those, given those are chips you got alongside to Champ Man 01/02 back then when you got a new PC to play it!
  2. The more curious thing in terms of specs is that FM 2019 basically targets the same processing units as FM 2009 did -- ten years ago. This likely also carries far deeper implications. On the 3d side of things, you can scale plenty. Just add a couple sliders that let people adjust the graphical details. However the baseline complexity of transfer markets/match engine calculations/general AI activities/generally load on the CPU -- that's a bit more difficult.
  3. This is the kind of survey I'd love to see run amongst FM players -- no less as it would likely include just as curious contradictions (and contain a hobby psychologist's delight). If run by SI, they could set it up in such a way that it would give them hints where they could improve their feedback -- or outright scrap some of it. What kind of manager in football listens to the guy on telly yelling "How did he miss!?", for a start? It also seems surprising the amount of times players haven't grasped how to use the feedback that's there, e.g. the feedback in the team reports. I've personally lost count on how often somebody was arguing he had unbelievably finishing woes, when the data on that page doesn't merely fail to back that up -- it outright contradicted that perception with showing numbers that were well above the competition's averages. Needs to be much more "in your face!", apparently. The data analyst may require a bit of an FM 2019 upgrade, either way.
  4. Svenc

    Woodwork - just too many

    It contains a lot of flat out garbage too though. This needs to be pointed out no less as it distracts from the issue of your thread. "Unmissable chance" is one of such, and he cannot be to blame (sorry for the linking this so much). And tbh, if there was a match where the goal was empty 5 times, I'd upload that as a possible keeper bug instead, in a separate thread. There is no artifical difficulty, as proven by countless threads and just as many community sites. There may, however intended to be a bit of randomness given that this is intended to be a football simulation game -- the entire crux of the matter seems to be of how random one perceives football to be, and how the game presents such. I'd arge the latter is a seperate game issue on its very own, and I don't merely mean any woodworks or shots off target -- but the entire feedback, including match commentary, the in-game media, and to a degree naturally also the naturally repeat player motion capturing animations (which in a game, will naturally always exist to an extent).To me football is a lot more random simply due to the fact that Football Manager by definition is a fantasy that glorifies the influence a manager really has (long-term, but also short-term over the outcome of a match -- or else what we can achieve in-game wouldn't be at all possible, including scoring just as many fantasy goals per match he argues would be impossible to score). Unfortunately, due to such, the game is a bit more predictable because of it. Simply fixing the number of woodworks won't fix anything much for him anyway, except that there would be less woodworks.
  5. Svenc

    First Half Goal Bias

    I don't think it is, however, you can do it for your own team on any save (screenshot I posted above). It shows such goal times under team stats. edit: You can check the AI teams goal times too if you scout them in the same screen. Naturally, it may be a bit influenced by management -- the reason that my team (nor the opposition) scored much in the 2nd half was that every time I was up, I went into "protect the lead" mode. This involved playing keep-ball in a way that a) not much shots for my team went off (barely any players were pushed forward) and b) not much shots from the opposition went off either (they didn't much get the ball back anways). I'm not surprised that this piece comes from a betting site! Similar to football people and the analyst linked to above, those people are looking for information that is valuable. I still don't think you'll find much if you only look at big wins where most of the goals are scored in the first half (what about the other big wins?) Not merely due to AI at HT then oft going into "keep damage to a minimum" mode. But as mentioned before: Lots of goals scored require shot conversions that aren't sustainable, no matter when they happen. After all, if FM has it correct, even most of the biggest chances excluding penalties or tap-ins can be roughly compared to a coin toss. How many times does the coin land heads in a long sequence if you flip it? Half of the time, patterns seemingly spotted are the result of human brains fooled by randomness either way.
  6. In the past the game has also never modeled how top clubs such as Valencia are/were pressured to sell their best assets or anything. Likewise, even in the Bundesliga, even the small clubs make gazillions over the long run. Would be an interesting research how much this influences the transfer markets in the long run also, en detail. Thinking about AI managers and their selling decisions also. The difficulty naturally seems not merely getting the English home model "right" -- it's that economically leagues can be hugely different, in parts to national regulations which are some more and less strict, depending on the country and competition. But yeah, in general, the actual struggle that lots of clubs have to go through, that's never been modeled by the game -- arguably recent releases have made this even worse due to the ever growing TV money. Whilst that is a detail that's always considered, country by country, it's completely unbalanced the financial side of things. You basically can't not not make lotsa money. And if you can spend millions as a club just being promoted to the Bundesliga, it's pretty far off what's actually happening.
  7. Svenc

    First Half Goal Bias

    Re: possible perception bias, something that may be worth considering here: Scoring lots of goals in one half, let alone two, no matter which, is something rather unusual and oft requires oustanding shot conversions that aren't sustainable over a season, let alone a match.Exclusively looking at big wins where the majority of goals were scored in the 1st half may only auto-confirm your observations due to the fact that scoring loads is against the odds, no matter when. Think of it like you rolling the 6 on a six sided dice thrice in a row -- you wouldn't expect that to continue likewise -- and it more oftenly won't. I'm not sure about your experience. But from mine as well as looking through various story sections, it seems obvious that not all big wins are caused by a landslide of goals in the 1st half. About shutting up shop: Luckily the AI isn't able to do such the way we are anyway, or else the perceived difference between halves wouldn't be a possible perception bias, but straight in your face. Aside of the tactical choices, which obviously influence goal spreads and goals scored in a huge way -- reckon SI must also look at what is actually going on in football. Whilst teams may "switch off" some in the 2nd half -- in tendency more goals are scored in the 2nd as in the 1st half. Something oft attributed to fatigue kicking in, and additionally errors. Likewise, teams are naturally more inclined to go "gung-ho" when the clock is running out too, increasing the likelyhood of scoring and conceding alike. Additionally, first halves tend to kick off mostly at a scoreline of 0-0 -- mostly. :P 2nd halves rarely do, oft forcing one team to do something or risk of going bust with a similar effect: Space becomes available on both ends of the pitch, and thus increases the likelyhood of witnessing goals. Statistical analysis for observing the effects of such is talking about changing "game states". SI likely balance and simulate this likely on how AI manages matches either way. For every AI that shuts up shop at half time, there is another AI that is visibly coded to go more aggressive, as it would risk losing that AI vs AI game -- a bit like the stuff that tends to happen in football.
  8. Not at all. Posting such has lead some of this board's most helpful threads by far. In particular when the OP kept an open mind, such as the in the fantastic thread just linked to. If I'd personally lecture about cognitive dissonance at schools or uni, I'd use FM as firstly young people tend to have at least a casual interest in football and games -- and it's a fantastic tool to demonstrate several cognitive dissonances such as perception and confirmation and hindsight bias, which affect everybody in some way in all ways of life. That said, it's not strictly a tactical debate -- though as any such dissonance rubs off on all kinds of choices, it may naturally affect tactical choices and interpretations too. And not necessarily in a way that may be of benefits to the own enjoyment of the game and, also, results. But now carry on with that tactics stuff. I would still strongly advice to always go through all the matches (against defensive teams first), and not exclusively focus on the ones where "stuff happened". I would also strongly advice to take a look how the AI managers of Juventus, Napoli, Roma, Lazio et all are performing in such matches likewise. They face a lot of rigidly defensive AI too. Who knows, you may be actually performing better than some of them. And given what the Lazio manager regularly did (pre-patch) anyway, you actually should.
  9. This looks similar to Chritoph Biermann's recently releases "Matchplan" (only available in German at the time of writing). Whilst I personally wouldn't predict a season outcome based off FM and place bets on it or anything -- the reason why following more advanced media has always been so beneficial is because mainstream Television still perpetuates cliche's, myths and sensationalist nonsense any time of the day. In fairness, a big part of the job of those guys is entertaining and exciting people -- no less as Television football is becoming an ever more expensive piece of primetime entertainment effectively competing against other pieces of primetime entertainment. As this piece from a few years back puts it: Industry people may be looking for something that is valuable; whilst the general TV audience may desire something that appears interesting. And some of it translates well over to FM overall -- such that contrary to expectations, if you experiment in the editor, finishing attributes, etc. don't make that hugely much of a difference to conversions, despite TV match commentary making players out as bonafide "killers" who supposedly finish everything and "bottlers" who screw it all every week -- or vice versa for keepers. tldr; as soon as you see somebody arguing how knowledgeable he is about sports as he's watched "Match Of The Day" religiously ever since he was a toddler, you know he'll have at least one big issue with the game.
  10. My argument wasn't so much that there was no struggle at all. However, cognitive dissonance is a big part of this, or else he wouldn't perform that well overall. It's not as if AI wouldn't drop points in such matches in sequence itself (same as happens in football...) -- in parts due to their weird picks which can be much worse than this. However, it is obvious it's getting exaggerated, including the rate at which opponents score in the last post. It's a paradox, but in a sense, part of these forums meant to help can have the opposite effect. As previously discussed in another thread, the line between an "exploit" and "tactical nous" on FM can be a fine one indeed, as both the engine as well as the AI has several weakness which don't exist in the game of football. As a direct consequence, it is possible in-game to have a shot conversion advantage over rivals that is non-existen in the history of sports. The majority of matches won't be even much of a contest, and player quality not much matter; whereas in football, matches are settled in finer margins, and even the big dogs **** it up. In either case, achievements from others shift the standards by which players judge themselves by. Players who "struggle" the most also tend to have a cognitive issue in here, namely that they tend to be "perfectionists". Unless they can't stop things happening altogether, it's tough. Naturally, this comes in many nuances. At the most extreme, players cannot accept that there are better and worse players on each video game out there. That's a rage quit garanteed, as eventually, somebody will always be a tad better than you. The key still is: Your rivals to beat aren't players. Your rivals to beat are AI -- and as of the AI, see above (curiously, they still don't have the **** that happens in football, despite their weirdo tactical picks). Mind, this is the tactical forums. But I thought this was worth pointing out -- in particular as the core perception always seems to be "doesn't happen in football". Overall, the game is meant to be a simulation of a football world. A bit of seemingly random crap may be supposed to happen (and to the AI, anway, does). The observations about the aforementioned "shenanigans" in his last post will be 99% perception bias either way. Unless one can get rid of that, one will always keep fuming.
  11. Considering the league position and the club picked, I'm smelling perception bias all over this thread/OP's posts. The first key to get rid of it is considering all matches against teams playing a tad more cautious, rather than solely focusing on the ones that don't go according to "expectations." As an example: Guardiola IRL last term conceded but 6 shots per match average, which is half of what the EPL average is. If he focused solely on the matches where he conceded (close to 30 goals in total despite it all), he would find the opposition scored from few shots in all of them. In one of them they even scored 3 from 4 (United). In parts, that's his (aggressive) system, naturally. Easier to score in yards of space rather than in space the size of a tunafish can. In general, every shot tends to be some kind of chance, and sometimes the worst are taken, that's simply random chance. Understanding this is absolutely key. You will find plenty players who rage across the shop due to not getting it. Unless SI were to develop an engine where it would be possible to not concede shots at all, players will concede goals over the season. And if they'd develop such a game, it would be something else entirelly than football. Additionally, struggle can be a bit subjective. A "struggle" to me, in-game anyway, is when several AI rivals do better than I do, for instance. Take a look what and how they do against such teams, and you may get a much better feel of it all. Oh, and football is much crazier than in-game across the shop -- you just have to follow it the way you do follow your game. All kinds of fascinating cognitive bias as well as weird expectations pretty much some up about 90% of the truly struggles players face on this, either way.
  12. Svenc

    First Half Goal Bias

    AI tactical picks in Europe, in particular two legged ties, used to be pretty interesting in general on older releases. Back then patch notes used to be super detailed, giving a few hints of how things work here and there, as every tiny little thing adressed saw a line, including AI reworks (<- link as an example under the "Match engine version history" header in particular). Following such, it became apparent that back then anyways AI prior to their picks work out "expected scorelines" they target. IIRC there was an issue with such in particular in two legged ties encouraging comebacks. Let's imagine AI A (Barcelona) had a heavy win in the first match. The AI calculation for the "expected scoreline" for the 2nd leg thus was a 0-0, as that was all for AI A needed to progress. Thus Barcelona would shut up the shop completely inviting the OP team in its half, whilst the other AI B went all attacking (its "expected score" calculated was a big win, as else it couldn't possibly progress). It's something that was noted by human players likewise, how extreme things could get e.g. http://www.fm-base.co.uk/forum/football-manager-2016-discussion/358243-why-barcelona-playing-defensively-against-me.html (as a side note, that's 22 Liverpool corners... Basically a corner for every 3rd minute the ball is "in play". In the abscence of added statistical feeedback, I'd watch those shots and whether they aren't primarily headers under pressure off of those corners, which aren't tap-ins and likely never be converted anywhere close no matter the player, but I disgress). So it's very well possible that the internal logics for how AI work out what they need in a group stage in the last game could be some faulty too -- or at least, produce curious picks. It's even possible that the Goal Difference isn't even much a factor.
  13. This goes straight back to the injury balancing -- the actual balancing is done primarily on AI simulations. Evidently, some players are consistently below the average, and some are well above the AI's injuries, including how their forwards score. There's nothing that keeps you from scoring (within somewhat "realistic" limits of course). Or having fewer injuries overall. Except for the input. A SOT conversion of 1 in 6 seasonal conversion is either way an easy fix for the time being, technically. It will primarily relate to the movement of players (or lack of it) a human player is expected to put into the UI by picking player roles and duties. Arguably it should be highlighted in the UI by showing movement arrows when picking such, visual cues that many football video games have had. It is debatable whether things work similar on an actual football pitch -- but without a semi-decent mix, movement on the pitch will be up to non-existent, and shots may still go off, but mostly not in that much space. Movement is pretty basic either way in any sports. There seems an ongoing debate naturally whether the UI is up to the task. If you think there is something wrong, perhaps triggered by how you play, I'd simply report the save. That it is possible to have SOT conversions this low can be seen by teams managed by AI -- and the root cause is similar. It's always one of the bottom teams. They play this ridiculously defensive (even by real football standards) over much of the season that they barely push enough players forward to ever have a decent move from open play. This relates to anything in terms of balancing. Is it really because of the game that Barcelona in-game don't have as much "possession" as in real football, considering that human players can do it, and yes, to which extent? Does Ronaldo score fewer goals in-game because the game can't replicate him, is it how the AI utilizes him given how human players can get him scoring fine, or is it maybe a bit of a mix of both (as arguably often the case given the game's many flaws)?
  14. Svenc

    First Half Goal Bias

    As a warning, the simple stats FM provides are deeply flawed, in particular if you cannot interpret them in a meaningful way. The in-game media itself doesn't either, nor do the assistants. This is actually also very relevant as of the topic at hand, considering the aforementioned tactical switches made by AI that try to prevent further goal leakage when they are about to lose big. Some AI tactical switches can be very short-term -- like an AI going gung-ho for but 10 minutes of a match and getting and equalizer from that --and then switching back. That simply won't show in the stats. As in football, it is the play making those stats, not the other way around. On the topic, if the AI in the 2nd half sits 10 players behind the ball, that will be tougher to break down, visually. Sometimes they may pack the centre, sometimes they may pack the flanks, and each approach may be considered in a different way. That is second by second what's happening, and the stats are derived from that. You can be hugely successful in lots of ways -- some do so by not even paying much attention to anything come match day! However, if you simply manage by FM's stats, that's a route to additional frustration -- at the least in isolated matches. At the very least, you will never fully understand what is going on. But then there's no manager in football that manages by stats alone too.
  15. Looks fun, but in a sense, not that massively far off*. https://www.whoscored.com/Regions/252/Tournaments/7/Seasons/6365/Stages/13832/Show/England-Championship-2016-2017 I reckon players are now gonna muse about clear causes of this -- like, the top teams being in a league of their own. Perhaps even musing about flaws they perceive -- e.g. "Football Manager = Morale Manager -- once yougo on a streak, you don't get out of it!" but realistically, it will simply have happened by nothing but random chance. And think about it -- *the only step from the above table is: That for each team at the top, 1-3 close matches that may have gone either way (football with its pitifully few points a match!), went for the teams at the top. Which are oft slightly more likely to win, anyway. That's the plus +6-9 points required, nothing more. I think chance is to be considered whilst playing FM also, at any one time, but that is a different matter. That said, still a mental season! An arguably more curious thing would be the goal differences, which in football too, are oft usually a better, or at least additionally hint of what's going on -- the league table lies! Better teams oft tend to win bigger, worse teams oft tend to lose bigger and those are big goal differences for each of the top 8.
×