Posts by Salix

    Thus, each player can choose not to repeat the game with other players.

    So cheaters can choose to continue to cheat? Doesn't that completely negate the whole point of this system?
    It's not just about being able to avoid encountering cheaters, it's also about preventing cheating being possible in the first place. If I can opt out of this system, I can still cheat.

    Adding NPC to clash after a timeout.

    We already have that, actually. That's the second way to cheat.

    Or: doing a promotion.

    Yeah, currently my ideas go in a similar direction. Not exactly the same. Promotion can do only so much and it's effect wears off if you repeat it. But I think Clash could be limited to certain timeframes. For example: "You can play Clash in the first 15 minutes of each hour between 5 AM and 10 PM". Something like that. Not sure if that would be "focused" enough, but playing around with the time brackets could do the trick.

    Create tv ads

    Nope. TV ads are f*ing expensive (both production and airing them) and do not really have an immediate effect of player numbers. They have other purposes (like brand awareness), but for this particular purpose it would be as close as "literally throwing money out of the window" as you can get ;)

    And also add RN players for best season (autumn and winter) for online games.

    In general we already consider that, whenever we can. Lots of factors go in the decision when exactly we do something, the season being one of them.

    Best ideas are simple

    Quantum physics disagrees! Probably a special case though ;)

    But seriously: I think Naikes idea is definitely a good basis. It has some flaws and so far I am not sure if it's possible to work around them...but maybe. A good path to explore!

    So my next question about this idea would be:
    Assuming we use daily resets for this system, we still have very big queue times for everyone outside of prime-time-players. This is even a problem right now with Clash! as it is. This system would multiply it.
    Any thoughts about possible solutions for that?

    why not change the random selection so that you can only play once against the same player(s).

    That is a possible solution, but with a downside: You might run out of players to be matched with. If you are a prime time player and play around 6 PM, you play at times where lots of other players are online, you might not encounter that problem. If you are an early bird who plays at....idk...5 AM, you only have a very tiny selection of players to play with and they will very quickly be excluded from being matched with you.

    In other words: Clash! is a fast game with short matches, so people only play the game for a very small fraction of the day. This divides a "Clash day" in dozens of tiny time brackets, resulting in a relatively low amount of possible matches, even if you play during prime time and especially if you don't. If you exclude matches in addition to that, waiting queues will quickly explode.
    Assuming 10.000 players play Clash every day and they are divided in equal time brackets of 30 minutes, we have 48 brackets with 208 players each. With your system, you exclude 20 players from your bracket per day. So just after 10 days, you run out of players and need to wait at least 30 minutes to find a match (just one, not five!).
    Of course this calculation is nonsense in both directions:
    Reality is better, because people don't always stick to a specific bracket every day.
    But reality is also worse, because people are not equally distributed in time brackets and the weakly populated brackets would run out of opponents much much faster. Some brackets will encounter the problem on the first day already.

    So yeah, this is a solution for cheating, but it comes at a very high price. Is it really worth it to essentially make Clash! impossible to play for large parts of the community only to fight a few cheaters? Is that worth it? I would expect a massive shitstorm if we implemented a system like that.

    And that doesn't even involve the fact that Clash! is much more fun if you play against people you know. So this solution would not only increase queue times massively, it would also eliminate the most fun part of Clash!
    Is that still fun?

    Gathering and analyzing more information can only help and improve the recognition of bad practices.

    Not necessarily. This is a common cognitive bias, the "Information bias". It's the false beliefe that information is always useful, even if it does not affect the action that is taken.

    In order to influence the action that is being taken, the additional information needs to contain something that doesn't already exist and that something needs to be valuable.


    If a player reports someone, that report contains two things:


    A) An opinion based on objective data ("This player is a cheater because he is obviously losing intentionally, as can be seen with his stats").
    B) An opinion that is purely subjective (Example: "I think this players is a cheater, because people from country XY are all cheaters").

    The information contained in A is info we already have. We have the data, with or without report.
    The information in B is info we do not have, but it's not valuable. We should not make decisions based on subjective, inaccurate or straight out wrong information.

    So punishing a player that also got reported a lot might FEEL better, because you feel like people are on your side, but nevertheless the reports do not provide any additional valuable information and therefore do not increase the quality of the decision.

    PS: This applies to this situation specifically. There are situations where subjective opinions are valuable info, for example when you have to decide if a certain insult is actually an insult or not. That is subjective by definition, so in cases like that, it matters and subjective reports do provide some value.

    I suggest that after each game played, players be able to alert other players to bad behavior.

    This can help detect bad players earlier.

    That doesn't change the fact though that we still need to use the same methods we are already using to evaluate the report. Given the fact that a player needs to make a decision (report or don't report) based on just ONE MATCH, these reports will be wildy inaccurate, because players who fail in one match for whatever reason will be reported just as much as actual cheaters.
    Just because a player reports someone this doesn't make the accused player any more guilty.


    So essentially, this doesn't make anything better, it just puts in an extra step.

    This should be very easy to look in to, with your 'tools' or just by looking at how many times those two accounts have been in the same match together. If it's more than once and if one of then suicides every time on the 3rd player, it's pretty clear what is going on.

    For obvious reasons I can not get into details, but our solution is pretty similar and even a bit more sophisticated.

    There are three reasons why this isn't a holistic solution:
    1) Family & friends. It's pretty common that players try to be matched with people they know. Just being matched with someone frequently is not an unusual pattern that identifies cheaters. Also, given the skill difference, frequently losing to the same person is also very unusual. A prime example would be me. I simply suck in Clash!, I lose a lot. Often to my colleagues. So my pattern essentially also looks a bit like a cheater.
    2) Frequency. It's not like all cheaters cheat all the time. Most of them just cheat sometimes. This also makes them very similar to normal players mentioned in the first point.

    3) Cheaters are smart, at least some of them. If radical behavior (like winning ALL THE TIME against the same person) gets you banned, they adapt and simply behave less obvious.

    So, I do not even try to find out intentions of others to punish them, I try to avoid punishments by making thinkable and observed intentions impossible to perform

    Yeah, that is in general the most effective approach to prevent cheating: Don't make it possible in the first place.


    Your thoughts about this are far from uncommon or unusual, it's essentially the basic idea of the "player dynamics" mindset: Make bad interactions impossible/unattractive by design.


    But as I wrote before: That's a goal. A very good goal, no doubt. But finding a goal is not the hard part. There is no need to discuss about that goal, we already agree on it 100%. We should leave the clouds of ideology and goals but get down to earth instead, discussing how to actually reach those goals.

    So my question to you is: How would you actually reach this goal? Not in some other game or in general, but specifically in the actual game Clash! in Rail Nation.
    This is the kind of discussion we need to have if we actually want to address the problem. I really enjoy the fact that we apparently share the same mindset concerning fighting cheaters, but it doesn't really bring us any closer to an actual solution for Clash!, doesn't it?
    Real world problems require real world solutions, so let us talk about those.


    Only if they know

    Yeah, and it's not that hard to figure out. After a few attempts you know whats going on and you adapt.

    I cannot really believe in the idea that the suicide players are multi accounts.

    Did I say they are? I mean, some probably are, but I agree that many of them are not.

    See, WE play clash for a few hours and as one result we KNOW the names of the cheaters, both, the suicidal avatar and the winner avatar. We happen to watch them 3 times, 10 times or more ... and we KNOW the cheaters, both.

    Well, that is pretty much the definition of a pattern. That is precisely what I was talking about. One match...fine, maybe he answered the door. Two games...sure, whatever, might be. 3 games? Suspicious. Etc.

    And exactly like you, we still don't know, we just suspect and are very shure about it. The clearer the pattern, the higher the chance to be right and the lower the chance of false positives.
    But although you wrote it in capslock, by just watching you never KNOW. You suspect.

    So, you need an automatic system to detect the cheaters, both of them, and to be 100% sure. [...]

    I made sure, that any bug in it will NEVER EVER punish an innocent, fair playing player or customer.

    I think you misunderstand the problem of avoiding punishment of innocent players. It's not at all about bugs. It's about the inability of being omiscient. The only way to "NEVER EVER" having false positives when punishing people for intentionally losing is to read their minds. Otherwise you are never actually sure. How would you be sure about someones intentions if you can't read his mind?
    You can be VERY SURE if you wait long enough to see a pattern that can't realistically be explained any other way...but that has the disadvantage of requiring you to wait. While you wait, the cheaters continues to wreak havoc. But if you don't wait that long, your accuracy goes down and you will also punish players who just look like they did it intentionally but are actually innocent.

    So the challenge is not to develop a bug free automation. It's about having absolute knowledge about what's going on in other peoples heads.
    In the past I had the privilege to work with a very big gaming company with virtually unlimited amounts of resources and talent. Their team to tackle this problem is bigger than the entire RN team. They still haven't found a real solution after more than 10 years.

    This is also a central topic in the Fair Play Alliance and last year me and many other members (who are 10 times smarter than me) met at Google in London to discuss this problem (and others). Essentially, all the biggest gaming companies in the world (and a variety of NGOs, universities and even nations) are trying to figure out a way to tackle this general problem of disruptive behavior on the internet. And "intentionally losing" is one of those gaming specific problems where no one has found a real solution yet.

    And that is, why I think more about cheating out the cheaters immediately

    That is a nice and noble goal. But I think figuring out the goal is not the challenge here, as with many other noble goals. The goal is clear. The challenge is to find a way to actually get to said goal. HOW would you do this?
    If you happen to know that way and you figured out how to identify someones intentions with 100% (or at least acceptable) accuracy, let me know.


    ne possible (part of a) solution would be a database table, that stores the names or account ids of the last 2-4 competitors and not to let them run up together till the table contains fresh names ... this way the cheaters would not be able to work together in two consecutive games.

    Easy workaround for cheaters: Just involve a third account.
    We also prevent that? Np, involve a fourth account. And so on...


    I definitely dislike bans, cause there's always the danger of misinterpretation

    Well, depends. Naikes example is ONE MATCH. Of course we don't make decisions based on individual games, but patterns.

    I prefer solutions that make special ways of cheating impossible (or at least very difficult) and take the control from the cheaters ... for example by a way to shuffle competitors.

    Maybe my idea will increase the waiting time at periods of low customer load ... but those cheaters would have to wait even longer to get together again.

    Yeah, I had a similar approach/idea: We could limit participation in Clash to certain times of the day, making sure that there are always so many participants that it's simply unlikely you will be matched with someone you want to be matched with in order to trade wins. However, that idea has the downside of not considering different lifestyles or "life schedules". In order to group participants densely enough, the time frames to participate in Clash! would have to be rather small. And if players happen to be unable to participate in that time frame, they are screwed. That's not cool.
    Your idea has a similar problem: The waiting time at some times of the day would be insane. Not everyone has that much time. If I come back home after a hard day of work, I don't want to spend hours in a waiting queue, just to be able to play my Clash matches.

    I want the players who have been caught for 2 years in such scams to get a complete ban on participating in Clash

    That is also my personal goal ( possibly including a temporary ban from RN). We still need to check how to do that, but at least that's the goal.

    From my old army days we had a rule about complaints. If you bring a complaint forward, you had better have a solution. Else wise it is not worth listening to. If your solution includes self promotion, then it again isn't worth much either.

    We have a very similar guideline within our company. And just in general I think these are very good words to live by, not just in a work context. Complaining is easy, it requires no effort, skill or knowledge, so you shouldn't think you deserve something (like a solution) for simply complaining.
    That being said, in the context of Rail Nation, I think it's a bit different. It's very honorable and useful if you complain only if you have a solution, but that's not necessary. It's not your job or obligation to do that, especially because figuring out a solution from the outside is often not even possible. For example I could also complain that we don't have any settlements on Mars yet, but I probably don't know enough about the topic to combine that complaint with a solution that is actually useful.

    So even though I fully agree with what you say, I would still say that, in the context of being a customer, simply complaining is still okay. Even if it's not always that useful, you can still voice your opinion. Just hearing your opinion is already useful, especially if you explain your opinion. Then we can try to figure out a solution. Maybe we won't be able to, depending on the problem, but it's good to at least know about the problem.

    Coming back to Clash!: I don't have a solution. We can't prevent people from intentionally losing, we can't control peoples minds. We also can't just punish everyone who simply loses hard. Might be a lack of understanding, might be internet problems, might be someone afk because he needs to answer the door. We can't just ban innocent players. Of course we could catch players who have done it a lot of time; so often that it can't be a coincidence anymore....but then they already did it. That doesn't prevent anything. The same applies to announcing draconic punishment. We already announced that cheaters might get banned and we do it. They don't seem to care. Also, in general, research confirms that announcing harsh punishment doesn't stop people from commiting crimes (or rule violations in this case). Sure, we will punish them eventually, but that doesn't undo the cheating.

    So, before my morning coffee gets cold, let me wrap this up: This is far from being a trivial problem; it's hard to solve. The good thing is that only a tiny percentage of players is foolish enough to cheat in Clash! and since every active player is playing a lot of matches, due to the law of big numbers, everyone is affected in the same way and it doesn't really matter on the long run. It's very frustrating in the particular match where it happens, but it won't stop your progress in general.
    If anyone has a realistic solution to get rid of these frustrating Clash! matches, I'm happy to hear them. If you don't have a solution and simply want to vent or complain, that's fine too. I just ask you to consider that the reason we don't have a solution right now is not because we don't care, but because a solution is simply not that simple.


    Precisely.
    I think it's important to understand the difference between the big picture, defined by the law of big numbers and statistics and an individual game. Clash! overall, the big picture of it, is essentially perfectly balanced. Each side will win roughly 33,3% of the matches on the long run. However that does not at all mean that an individual match can not be unbalanced. It's random after all and that includes the possibility that you get extremely lucky or extremely unlucky in an individual match. That's not in contradiction to the general fairness of the game.

    So my recommendation to guard yourself from being angry about "bad luck" is focus on the big picture. Play as skillfully as you can in every individual match, regardless of how Fortuna favours you (or not) and in the big picture, your performance will benefit you, regardless of those unlucky matches.
    Doesn't just apply to Clash! but to essentially all competetive games or sports that have a certain aspect of randomness that gets compensates by the law of big numbers.

    According to Google, XXL is not a proper Roman number. 300 is CCC, unless you are talking about and extra large update!

    It is indeed a reference to the clothing size, not a Roman number.

    Would definitely be interesting to know the stats per colour for each player, agreed.

    I don't know right now, but I can tell you last years win distribution stats after 9 days:

    Unten = Bottom
    Rechts= Right
    Links= Left

    Can we please come back to the original topic instead of discussing who is "best"?

    Rail Nation is a game that supports multiple game goals and play styles, so there is no "correct" way to play. And even if there were, the individual skill of a player does not improve the quality of his arguments in a discussion, so there is no value in discussing about that either way.

    It's not an IP scan, it's a bit more elaborate than that. But anyway, don't worry, any player who tries to win a top position in Clash! with unfair methods will be sorted out and sanctioned.

    Fun fact: We did actually play around with concepts of futuristic scenarios. My personal idea was dystopian scenario called "Dark Future" (a not so witty play on our studios name "Bright Future").
    And my all time favorite concept art will always be this:

    I'm talking about rogue players, or one-man association players, who commit such acts of sabotage without any obvious motivation.

    I disagree a little bit on this statement: I think if the motiviation is not obvious, the usual rule of "innocent until proven guilty" applies. In my personal opinion, it's not about about "no obvious motivation", but "obvious motivation to simply being a jerk".
    Maybe that is what you meant, maybe not. But in my opinion a player should not be punished just because his intention is not clear (because that also applies to players who just don't know better), but he should only be punished if his intention is clearly "evil" and there is proof for it.

    Sabotage, connections, prestige hunters, industry bombing make the competition much more interesting and it's fun.

    I agree on that. Rail Nation in general is an extremely cooperative game and quite peaceful, given that you can't really lose anything you earned and any competition is more like a race about who does something best, not about taking something away.
    So to keep things interesting, a certain threshold of gameplay-induced friction is a good thing in my opinion. The "salt in the soup", so to speak.
    This only gets too much once players use these mechanics for the sole purpose of ruining the fun for others.
    Being upset about some things others are doing is part of a competitive game, which Rail Nation still is.
    However, doing something ONLY to upset others is not okay.
    Distinguishing between the two is not always easy though :/

    About the topic at hand:
    The game rules acutally will soon be updated and will include a general rule about fair play that covers...well, pretty much every intentional harmful behavior that is not covered by other rules.
    We will also be introducing account bans and increase our efforts concerning multis.
    Will this solve all problems? Definitely not. No rule and no punishment in the world can fully prevent harmful behavior, some people simply do it anyway. Punishments are not really that effective at preventing anything.
    But it will help.

    Required reading - "How to lie with Statistics"

    Except that we never share the results of these surveys, so we have no need to fake any wished result.

    Questions, in a survey, or when collecting customer requirements must be unambiguous or they produce worthless data.

    You seem to have ignored or avoided reading my previous post:
    You seem to think that we were asking for specific changes. Instead of asking for that, we asked a very general question about any change at all. In that case, this would indeed be a worthless question, since it's not measuring what we want to measure.
    But, as I explained in my previous post, what we wanted to measure is the acceptance to change in general. And the question is asking about precisely that. So it's measuring what we want it to measure.


    In other words: You claim this question is measuring A, while it's asking for B.
    But it's actually measuring B while also asking for B.
    If we are interested in A, we will ask for it.