Naruto Online Forum

5.0 GNW Treasure Minato

LOGIN

Register

Forgot Your Password?

Nickname: PraiseLuka
User Group: User
Statistics: Topics 29 ｜Posts 2271

ACTIVITY: Registered: 2017-07-24 19:08:02; Last Login: 2018-03-09 12:56:33; Last Post: 2018-03-07 04:03:03

TITLE SECTION REPLIES/VIEWS LAST POST

[Updates] IMPORTANT - Mystery skills delay problem
"
To fix the problem, one must first understand the problem. There are THREE different reason delay occurs. 1. Connection issue. Self explanatory enough. This is USUALLY not an issue within a server, but can be problematic in ANY Xserver event. 2. Flash rendering. The server assumes that flash is rendered perfectly at the speed set. Unfortunately, this isn't always the case. On more modern systems, rendering can be kept up at 1x, but falls behind on 2x as the animations aren't sped up as much. And on older systems, it is possible that the client can't even keep up rendering at 1x speed, thus falling behind. And finally, #3, chase and animated mystery. This, in a way, is an offshoot of #2. The problem here is that the server does NOT wait extra time for those to resolve. Instead, it simply allot the same amount of time as it would with an attack/mystery that does NOT cause chase or have animation. But as the client need to render these animations, it falls behind. It is even possible for longer standard attack animations, such as that of Minato, to also contribute to this, as I suspect the server waits for the exact same amount of time for every standard attack. As a result of any of the above, client state can fall behind server state, either because client didn't go as fast as it should or because server wait time is not properly allocated for chase/mystery. And once there is a difference in state, we will see the delay problem. Specifically, when we issue a command, the command is queued according to the server state, which is "the future" compared to client state. While it might still be in the middle of the round on the client side, the actual server state, and thus when the command will result, is either later in the round or even after the round, thus causing the command to be queued toward next round. There are, IMO, two methods toward solving this. 1. Logic path: Give extra mystery cast options of "reactive casting"(only when target cast something) and "queued cast" where it will be cast when able(if currently unable) and "cast at last opportunity"(where it will be cast before the turn end for prompt and right before its standard for non-prompt). Generally speaking, those three will be all that player wants, so even if we are not seeing the correct state, we can still issue command that we wanted all along. 2. Configuration path: configure the server to wait properly wait for the client. To be clear, problem #1&2 are mostly the player's problem, the server should NOT wait for those(except Xserver issues, that's more of a stability issue, and animation should be sped up more on 2x). But for #3, the server need to allot sufficient time for every mystery, chase and, should it be a problem, extra long standard attacks to resolve. Once the server actually wait the correct amount of time, client side will no longer fall behind due to having to display all those. Comparing the two proposed solution, #2 seems simpler while #1 is more complex and in, in a way, only a workaround. So we should do #2, right? The thing is, #1 will work now and always, with ONE change. Whereas #2 will require constant maintenance. Because to know how long the server SHOULD wait for, one must take into account just how long the mystery/chase/standard attack actually takes. This means that you will essentially need a time chart for every single action in the game that needs to be kept up to date with every ninja introduction and skill change. This might have scaling issues. Also, I'm not entirely sure how easy it might be to have it properly calculate the time of a chase chain, which in theory is simple addition but depend on the coding, might be complicated.Or, of course, every action that can be taken will wait for the maximal amount of time an action of that category needs to take. (Say all mystery waits like Garras, all standard waits like Minatos, thankfully I think chase all have the same amount of time per chase). But this could make the game slower than before, which might negatively impact GNW, which does time out every now and again.
"

Bugs & Support

65

33027

newgate446@gmai
2018-04-10 02:51:55

[Suggestions] modify the rewards for top3 people in swb
"
What color are your 8 gate? Because higher color cost MUCH more. I have 5 purple(4 of which being quick, ofc), 11 gold, all lv 11, and 16 reds, 12 @lv 10 and 4 @lv 9. 40-50 million is hardly a lot of coins, that's about what I get in less than 2 weeks. And it's not even enough to get one red to lv 10. As for wheel, right now we are at 14 spins/day. I get, on average, about 30 coupon from it daily, and about as much "value" in magatama, thou I don't care much for magatama now. Are you saying almost 1k coupon/month is not useful?
"

Bugs & Support

63

15140

dddds2
2018-03-15 17:36:15

[Other] About Treasure Tools
"
Exactly that. It refunds you the exact same amount of exp, just in different form. It's actually better this way, since it allow you to spread the love.
"

Bugs & Support

5

2122

Tobei
2018-03-03 02:43:19

[Suggestions] modify the rewards for top3 people in swb
"
Problems: 1. Freezing does have negatives. If nothing else, you are missing coin from fox, which is used in treasure, 8 gate, donation into wheel spin into coupon. And depending on how hard you freeze, potentially other things like jin rewards, daily quest rewards etc. Generally, the pro is greater than the con, but you do have to sacrifice something. And there is potential for con to outweigh the pro, see #2 2. Freezing creates horrible game experience for newer players. If they don't freeze, they will not win on their way growing up. If they do freeze, they will still lose to player who have frozen longer. So they will lose all the time either way. For server cluster that are relatively new or small, this might not be a problem. But for somewhat older, growing clusters, this could be a problem. Fact of the matter is, there are only 9 "winners" between all ptw and freezers for a partciular bracket in a specific cluster. More than 9 people freezing means someone will lose, and that tend to be the "fresh meat". Proposed solutions: 1. Nerf top 3 rewards for non-100 fields. For one, this only affect sage, which is not the only reason people freeze. For another, this is working off the assumption that only the freezers win. But while rare, a "normal" player could win as well, why exactly should the be punished? This is, imo, a very spiteful suggestions as it helps no one and only hurt some. And not everyone of the "some" are "guilty" of whatever behavior we do not approve of. 2. Give exp to winners. This would work for lower brackets, but for higher brackets, say the 80s or even low 90s, it doesn't work so well. For smaller amount of exp, it would take too long to have any effect. For larger amount of exp, say 10% of a level, that's literally more than what a player can normally gain in a day. This could potentially make ptw even stronger in newer servers. So while it might help in the Xserver prospective, it can cause problem within a specific server. Granted, the impact might be limited, because if a ptw spend enough to keep winning sage, he'd probably win w/o the 10 level advantage over his ftp/ptp peers. But it's still a potential problem to have lv 90s in a server that otherwise barely have any 80s. 3. Power over level This, IMO, is what it should have always been. But it does have the problem of potentially harming ptw, which the game is probably unwilling to do. Because bracket one in each cluster is going to be filled with those who spend money to win but are now unable to, regardless of their level. Also, while I dislike level freezing, it is, objectively speaking, an "investment" that people work for. Taking away the fruit of the investment is questionable. For player who have already harvested some fruit, it might be fair, but for those whose fruit have yet to ripen, it can be unfair, especially if you consider many of them are just following the advice provided by more seasoned players. Furthermore, while the top ptw are pitted together and "lose", the lower paying player with purchased ninja might end up dominating the rest of the fields and free player might have no recourse whatsoever against them due to the similar power, whereas level freezing at least offers some options for f2p players. 4. Capping wins This is an interesting suggestions. If there is a finite number of time one can win, to maximize one's own points, one need to try to terminate streaks. This would encourage the strong to fight each other rather than picking on the weak. The problem with this is that it does cause sage result to be partially based on luck rather than power/skill. Because you might never get to catch someone with streak even if you can beat them. Also, we would have potential problem of players using alts to feed a strong opponent, getting them to the win limit but with minimal amount of points for that number of wins (which, btw, is probably why they kick out the people who keeps losing, so people can't use alts to feed themselves) As much as I dislike bullies, I hate cheaters more. On a side note, you might noticed that I said capping win as opposed to capping "morale"(win streak). Because one c*ually terminate their own streak easy enough, so capping streak won't do much. 5. Shrinking sage size This would be good, but unlikely. Because this essentially means more reward per capita for free. It's essentially asking for free hand out and we know how oasis feel about that. But if done, at least in the short term, it would help, as it's less likely for there to be 9 ptw/freezers taking up all the winner slots in each bracket. That is, at least, until problem #2 noted above catches up and they become saturated again. 6. Winners bracket This is an idea I had before, where the winner(top 3 of each side, so 9 per bracket) gets tossed into one bracket the next time they do sage. This, in a way, is a cross between #1 and #5. It will punish the winners, except the top bracket winners, as no one can win twice in a row. It will allow more player a chance to win top reward (becomes 18 per bracket instead of 9, as alternating sets). It does avoid the "extra free reward" bit of #5 which oasis won't like. But at the same time, it carries the same potential problem. It may or may not be fair to punish the winners. And eventually, even 18 winner per bracket might get saturated as more and more people level freeze. So as you can see, there is no "perfect" solutions. Every solution have problems, either long term or short term. And there is always the question of fairness as well as the status quo oasis might wish to preserve in the amount of reward to hand out as well as protecting the "rights" of whale which they do need to cater to. So sadly, there is no cut and dry simple 100% positive solution. Also, I'd like to note any modification would have to be implemented by tencent, not oasis. And then take months before propagating here. The chance of these suggestion even reaching tencent is rather unlikely. And who knows what would be lost in translation in between, with the reasoning or wording twisted in such a way that puts the devs off.
"

Bugs & Support

63

15140

dddds2
2018-03-15 17:36:15

[Help] Level Freezing
"
FYI: Jin reward comes in the mail, one half (participation) comes with exp, the other (top 10) does not. You can receive one without the other. So you should do Jin either way. Nine tail exp reward scales with damage, with minimal amount of coupon and the magatama paid out as long as you participate. You can poke it gently with a lv 1 ninja to get the minimal reward for minimal exp gain. It may or may not be worth it but is a thing. On the other hand: You will, even if you do not skip exam, miss out on some reward from exam. As you gain power slower w/o leveling due to the other parts stated. As such, you will pass roadblock levels later than normal, thus missing out the amount you can farm out in between. This amount will only grow over time until such point where you should have cleared them all w/o freezing, at which time the gap begin to shrink, finally catching up as you, with freezing, clear it all. Also to keep in mind, with increasing number of player freezing, the benefit is diminishing. Because you will lose to any one who started freezing in the same range before you did. And you will lose to ptw regardless. With in a particular level range, there is only a finite number of slots for "winners", at this point, you might not be able to grab a spot even if you did freeze. Unless, of course, you go ptw yourself. This will seriously depend on the level range you plan to freeze at as well as the server cluster you are in. You should investigate whether you will actually get the wins you want if you do freeze before hand.
"

General Discussion

5

3530

ZWen
2018-02-27 14:59:07

mine steal
"
Yes, it will be in the logs (on top)Not sure... maybe.... But not many would really care for the ore, since you are mostly coin bound. The thing for stealing is the card flip for essence which doesn't come out of the loser. Also, they can avenge themselves (with, afaik, infinite attempts until success) which I think also give them back some ores?
"

General Discussion

3

1419

MoeDmC
2018-02-25 07:59:40

Basic Rewards X Server Ninja War
"
1 Yes 2. Yes but no. You get pack as long as you are in the Saturday fight, for both GNW and XGNW afaik. And in case of XGNW, there are only 8 selected so you can't be disqualified in Wed. Unless you expect to at least win one round on Sat thou, it might be better to opt out, you will get more seal scrolls that way. But if you think you can win at least one round, Xserver might be worth it.
"

General Discussion

3

1842

Arora-Fireball
2018-02-25 07:18:21

[Help] Ok that's it
"
That they don't have evidence of anything because basic observation is filled with bias and misinterpretation. They need to collect real unbiased data to prove anything. Edit: most of the post is explaining how bias can come about.
"

General Discussion

52

940

PraiseLuka
2018-02-24 09:09:24

[Events] guide about rng events.
"
And welcome to the echo chamber! please clearly define what your "experience" is, including all data, observation and methodology. What did you do to prevent selection/observer bias etc.
"

General Discussion

6

199

TheWakkaSage
2018-02-24 07:48:03

[Help] Ok that's it
"
First, we need to define what "rigged" is. "manage or conduct (something) *ulently so as to produce a result or situation that is advantageous to a particular person." So there are two criteria to this, one it needs to be *ulent. and two it needs to be a result that's advantageous to a particular party (in this case, oasis, or tencent if you want to go there) Take TI reward, for example. It IS *ulent in that the card flip is fake. However, there is thus far no proof whatsoever that it is not "fair", i.e. the pre-generated flip result is any different than if you were to choose at random. And as such, there is no proof that it is advantageous to a particular party and therefore it is not "rigged" as it does not fit the definition. There are many theories on how RNG events might be rigged, but there is a significant lack of proof for any such theories. The closest one was the "expose" on "one per day" thing. But, as evident by the SA adv thread fiasco, the mods do NOT necessarily know what they are talking about, so take it with a grain of salt. (And tbh, I expect the mods to continue to insist that SA did drop it before 4.0 so... you know... whatever) And now, I will proceed to give you a short lecture on RNG and selection bias. On why even if it isn't rigged, you could see what you think you saw. (To be clear, I am NOT saying it isn't rigged, just saying that you have inconclusive evidence) Take, for example, this statement "Kaizer spent 25k for Minato and got nothing. Even at 1%, after rolling 1000 times you should freaking get him at least once." It is simply false. Now, on average, vast majority of people would get it within 1000 rolls and stop when they do. But it isn't a guarantee. Specifically, at 99% chance of not getting one in a single roll, the chance of not getting it in 1000 roll is (99%)^1000 or 0.004% A small number to be sure, but it's not 0. Now, it's rather unlikely for this number to be actually reached. Because it's rather unlikely that it would occur out of the few people who would go that far. So while not concrete, this does begs for some questions. But that's on the basis of the winning chance is 1%,which is where selection(or observer, if you will) bias comes in. Given 100 player, for example if: 30 do not spend to spin, only 7 free, 1 wins 50 spend small amount to spin say ~50 times each, 10 wins 19 spend a larger amount, say 100 spins each, 9 wins. (do note, some of those might be willing to spin more, but got it and stopped) 1 spend huge amount, say the 1000 spin and doesn't win. What does people see? 1. A bunch of lowie got it (the free-only and maybe some small spending ones) 2. "most" people who spin a good amount got it. (And let me tell you a story here. A couple month back, water main was really dominating. And at the time, there were frequent complaint how "everyone is a water main" and the "water cancer spreading everywhere" etc all over the place. But when I actually did a head count in space-time ranking, water main only come up to a bit over 1/3. Still a significant over-representation to be sure, but VERY far from "everyone". So when you see something often enough, you might over estimate it. So despite only a MINORITY of the spinners "won", the perception will be that most have) 3. The expected amount to get it is around 100 spin, as large number of player get it at that point. Because NOBODY pay attention to the losers, at least not unless it's a big loss like 1000 case. What is the reality? even if we ignore the 1000 spin player, who would obviously drag down the average, the actual winning chance is slightly below 0.5% And what does that change about the chance of spinning 1000 times and not get it? It becomes 0.66%, 154 times bigger than our previous result. And given that we do have a least a dozen or two whales who would be willing to go that deep, the chance of this happening to one of them is about 10-20%. That, my friend, is hardly an unlikely scenario. Especially if you think a step further, that this is not the only time such event happens. On average, two such event exist every month. And people will spin for win during them. Over the course of 3-4 month, this is almost guaranteed to happen to one of the whales. And that's just with 0.5% chance of winning. If it were 0.3%, the chances would be vastly greater and frankly, the difference would be hard to tell to the average joe. (by the way, that would be 5% for a single player, 28% for a single event and happen once every 2 month or so) And then we come to the second part of the selection bias, that is the self selection bias of complaining about bad luck. How many % of people do you think would come to the forum and make a post about getting minato with around 100 spin? Given how many people got him and how few post there are on the subject (most of which are youtubers who will post about ANYTHING anyway), I'd say less than 0.1%. Similarly, how many people who spin 100 times, did not get him and stop would come here? probably not a lot, most people won't come here until a higher count is reached. You yourself didn't until about 200+ spins per event in MULTIPLE events. But, should the "injustice" of spinning 1000 times and not getting it happen, how many would report? I think it's a safe to say most will. So when it does eventually happen, as I've pretty much proven it will above, here we are-- with "proof" of event being rigged pasted all over the forum with little to no counter-examples nor sufficient data provided to actually find out what the actual rates are. So does that mean you can't prove the game is rigged? Of course not. You just have to eliminate the problem in selective reporting and reactive behavior. You need concrete unbiased data. Here is how it can be accomplished. 1. All participant (be it your own alts or other players of whichever server) must agree to follow through a specific set of actions, regardless of result, at the next lucky wheel and report it. This participation MUST be committed BEFORE the event actually starts.and the reports must be made no matter what. If more than 5% does not follow through, the data is un-usable. 2. a. You need about 1000 characters of all level range (and preferably power and paying status) who will ONLY do the free spins. b. You need about 100 characters of all..... who will spin 100 times (you can choose if that include or exclude the 7 free) no matter WHAT. they are NOT allowed to stop even if they win and they are NOT allowed to continue if they don't. c. you need about 10 characters who will spin 300-500 times no matter what. Each character must commit a specific number before hand and are not allowed to stop/go beyond no matter the results. And then, once the event is over, the result of those 1110 can be collected, and we can examine the data to see if there exists any type of objective bias of sufficient significance and confidence. If there is, then we can say the event is rigged. P.S. This, by the way, is why "public opinion" on many issues are so polarized. Because only those with a strong opinion one way or the other (similar to those who had extremely good luck and extremely bad luck in RNG stuff), would voice out. The moderates, the normies mostly just sit silent. If you want actual correct counts, you have to "force" people to vote/make their opinion known. Which is why, in this case, prior commitment toward contribution of the data collection must be made as well as that it requires most players following through for the data to be valid. I am, by the way, willing to contribute as a 7-free-spin player if you want to gather actual data. So now, go find the other 1109
"

General Discussion

52

940

PraiseLuka
2018-02-24 09:09:24