I liken this to how it seems that the Dilithium mining gets twitchier during the Bonus Dilithium Events.
'But to be logical is not to be right', and 'nothing' on God's earth could ever 'make it' right!'
Judge Dan Haywood
'As l speak now, the words are forming in my head.
l don't know.
l really don't know what l'm about to say, except l have a feeling about it.
That l must repeat the words that come without my knowledge.'
The "story" that there was a 'bug' inadvertently hiding the failures is a particularly odoriferous accumulation of equine excreta.
Failure's have always shown for every one of my 23 toons.
Every time a toon gets it's first bunch of DOff's and you run missions with white's only, fully half the missions would fail. And right there in the results window, was the little read "failure" or even 'disaster' result.
As others have said, now however even assignments with Ultra Rare DOff's assigned are failing, something which almost never happened before.
My guess is that the so called "bug" was actually a change (or patented "Cryptic Stealth Nerf") to the RNG, so that players were less able to use DO'ffing alone to level up characters.
The problem is a noticeably very high failure percentage on Confiscate Contraband Missions.
Trust me, I know that failures show now, I get that. I also know that my failure rate on this particular assignment as more then tripled.
Ah. Guess I didn't read the whole thing.
But yeah, honestly, I've always felt that the odds for the "Confiscate Contraband From Crew" missions were rigged somehow: for me, they fail way too many times for what you'd statistically expect. In my case, that usually means 9 out of 10 missions should succeed. In reality, I'm lucky if I don't get a failure in like 50% of the 'rolls'.
I don't think its odd that you succeed a mission often and fail it sometimes given the probabilities.
It's not easy to get past perception biases.
Which can be true. But the idea that I did both of those missions yesterday and they both failed... very odd.
Sure, it is entirely possible I was just unlucky, and my assignment log does not go further back. But I do want to say that both of my characters failed that mission the day before yesterday, as well. Which again, does not necessarily mean anything and I could be remembering it incorrectly.
When the failure rate is that low (12% and 9%), and given the similar experiences mentioned elsewhere in the thread? I think it's suspicious at the very least.
I'm not jumping to conclusions or think there's any stealth nerfs or anything malicious.
I am saying that I find it extremely unusual, and worth mentioning.
Which can be true. But the idea that I did both of those missions yesterday and they both failed... very odd.
Sure, it is entirely possible I was just unlucky, and my assignment log does not go further back. But I do want to say that both of my characters failed that mission the day before yesterday, as well. Which again, does not necessarily mean anything and I could be remembering it incorrectly.
When the failure rate is that low (12% and 9%), and given the similar experiences mentioned elsewhere in the thread? I think it's suspicious at the very least.
I'm not jumping to conclusions or think there's any stealth nerfs or anything malicious.
I am saying that I find it extremely unusual, and worth mentioning.
Your whining about the fractured state of the Romulan race has displeased the RNG and therefore you have been punished for it. :P
"The clustering illusion is the tendency to erroneously consider the inevitable "streaks" or "clusters" arising in small samples from random distributions to be statistically significant. The illusion is caused by a human tendency to underpredict the amount of variability likely to appear in a small sample of random or semi-random data."
Put another way, the presence of streaks or clusters is to be expected from randomness. If there were no streaks or clusters, it would suggest that the outcomes weren't really random, since with random outcomes it is very probable to see streaks and clusters.
I've mentioned this phenomenon before on these forums, but I don't think anyone understood what I was talking about at the time. It's interesting that several people are bringing it up now.
I'm aware of this fallacy. It is most relevant when dealing with small sample sizes of truly random outcomes. This phenomenon is precisely why "true" random number generation is generally undesirable in game development--it's more important to ensure that a player perceives the results as random than that they conform to an expected Poisson distribution.
Getting four or five outcomes in a row that each have (e.g.) a % chance of occurring is statistically inevitable with true randomness, but from a player's perspective it is a design defect.
Are we talking about random sampling from a uniform distribution or random sampling from a Poisson distribution? It's news to me if STO uses a Poisson distribution anywhere.
What you are describing is correct - when people see clusters they tend to see a pattern where none exists because they are looking at a small group. They think random but what they really want to see is uniformly distributed.
This is where one has to be careful with technical language. Clusters or streaks tend to appear in random samples from a uniform distribution. It's surprising to people that random samples from a uniform distribution don't look evenly spread out. The standard meaning of "uniformly distributed" is that each outcome is equally likely. It doesn't refer to the appearance of an i.i.d. sample from such a distribution.
Interesting read. I had not realized perceptions were SO far off base in so many people. But then, I have been familiar with statistics for many years so I guess my perceptions are not the norm across groups that did not take the classes I did. That said, I would think that education in these areas would help people to understand. Not everyone went to a high school that offered Stats as a full-fledged course or needed it for their college degree. But I am sure a lot of people would get it very quickly if shown.
I'm trying to think of a way to explain the appearance of clusters. I can think of an intuitive explanation, but I'm not sure how technically correct it is. Intuitively, in order for a sample to appear evenly spread out, the members of the sample have to "actively avoid" each other and thus can't be independent.
I suppose that depends on what 'odd' means. If it means something that is normal but isn't expected to happen often, then sure, it's odd. It would also be odd to go outside and see a shooting star.
There's absolutely nothing suspicious about it. Nothing at all. It's exactly what you'd expect to happen if the mission outcomes were random and hewed to the probabilities.
To reiterate, this thread was started in response to a change which made it so that failures were no longer invisible. That alone should be enough to make anyone laugh at it.
"After they made it so I could see which missions I failed, I suddenly noticed that i was failing a bunch of missions that I didn't notice before."
It's comedy.
The rationalizing of voodoo perception biases is less comic.
Alright, humor me. How many of these failures in a row (with the data I've posted) would you consider normal before getting suspicious over the RNG?
<snip> I'm trying to think of a way to explain the appearance of clusters. I can think of an intuitive explanation, but I'm not sure how technically correct it is. Intuitively, in order for a sample to appear evenly spread out, the members of the sample have to "actively avoid" each other and thus can't be independent.
I think the easiest way to visualize it for me, was brought up by a programmer responsible for creating the "random playlist" feature of some music playback program. (could have been itunes or some such)
Basically he said that they originally created a true "random" outcome playlist feature, but whenever it completely randomly put 3 or more songs from the same artist together, people complained that it wasn't random.
So they had to add code to the 'randomizer' that would stop it from putting too many songs from the same artist (or whatever other variable you chose to create the playlist - genre, length, title etc.) together in a playlist ... In other words, it was no longer truly "random'.
If you toss a coin 20 times, you might not be too surprised if you got 15 heads and 5 tales.
If you toss it 200 times, you may be a little surprised if you got 150 heads and 50 tales. Toss the coin 20000 times and get 15000 heads and 5000 tales, you'd probably "perceive" it as rigged somehow, but the percentages are exactly the same each time.
I'm not being superstitious. Just suspicious. I do understand everything you're saying, and I do agree with it.
To take your shooting star analogy another way... if you walked outside and saw the same shooting star night after night, you wouldn't at least consider looking up it up on the interwebs to see if there were any reports of a meteor shower or looking up observational data for that particular place in the sky you're looking at?
That's kind of where I'm at. I think I've seen some sort of unusual behavior, and now I'm cross-referencing it. It's entirely possible it really is just bad luck.
However, it is also possible it's the Freemasons using the Illuminati's RNG software to rig the game so the Lizardpeople can take over the globe with chemtrails.
Or maybe somebody just dropped a decimal point somewhere in the last patch a la Michael Bolton in Office Space.
If you toss a coin 20 times, you might not be too surprised if you got 15 heads and 5 tales.
If you toss it 200 times, you may be a little surprised if you got 150 heads and 50 tales. Toss the coin 20000 times and get 15000 heads and 5000 tales, you'd probably "perceive" it as rigged somehow, but the percentages are exactly the same each time.
No, I'm sorry, but this is wrong. Here's the output from Maxima.
The last probability is not really zero, but small enough that Maxima returned zero for the output. A better explanation is that if you toss a fair coin 20,000 times, you'd expect at least 19 streaks of 10 heads in a row.
Just reporting in that I suddenly suffer a lot of doff assignment fails as well. Noticed it during last weekend. Even assignments who at least succeed 350 Days a year tended to fail in a bunch. Perhaps just a bad streak of luck but I found reading this thread which was created the same time very interesting.
Who knows, but I bet this is just a prelude to Gekos announced normalizing of the doff system.
The word probably means bringing it more in line with the current active game play experience of Star Trek Online. At least failure is a word most players who survived Delta Rising should feel accustomed to by now (bugs, lag, nerfs, crashes, crafting mods lottery, leveling, advanced pve fail criteria ec ).
We shall see and hope for the best.
Looking for a fun PvE fleet? Join us at Omega Combat Division today.
I'm not being superstitious. Just suspicious. I do understand everything you're saying, and I do agree with it.
To take your shooting star analogy another way... if you walked outside and saw the same shooting star night after night, you wouldn't at least consider looking up it up on the interwebs to see if there were any reports of a meteor shower or looking up observational data for that particular place in the sky you're looking at?
That's kind of where I'm at. I think I've seen some sort of unusual behavior, and now I'm cross-referencing it. It's entirely possible it really is just bad luck.
However, it is also possible it's the Freemasons using the Illuminati's RNG software to rig the game so the Lizardpeople can take over the globe with chemtrails.
Or maybe somebody just dropped a decimal point somewhere in the last patch a la Michael Bolton in Office Space.
But for that to be true the displayed odds would have to be different from the actual odds and there's no reason to believe they are.
If you toss a coin 20 times, you might not be too surprised if you got 15 heads and 5 tales.
If you toss it 200 times, you may be a little surprised if you got 150 heads and 50 tales. Toss the coin 20000 times and get 15000 heads and 5000 tales, you'd probably "perceive" it as rigged somehow, but the percentages are exactly the same each time.
If you tossed a coin 20,000 times (or 200,000 times, however large you want to make that number), heads and tales will have come up about equally (and certainly will towards infinity). If it were any other way, the coins would be faulty.
So, out of 200,000 coin tosses, when you got 150,000 heads, you'd practically know something is fishy. :P
If you toss a coin 20 times, you might not be too surprised if you got 15 heads and 5 tales.
If you toss it 200 times, you may be a little surprised if you got 150 heads and 50 tales. Toss the coin 20000 times and get 15000 heads and 5000 tales, you'd probably "perceive" it as rigged somehow, but the percentages are exactly the same each time.
If you toss a coin 20000 times and get 15000 heads and you have any sense at all, you realize the weight of the coin is off and you stop using it in that way.
This isn't a perception problem, it's an unevenly weighted coin. The only real question is was it damaged on purpose or by accident.
I finally figure the cause of the increased failure of DOFF missions, especially as how even missions with 0% failure are failing. The probability matrix has been reprogrammed by a Ferengi consortium along the lines of their business strategies "Rule of Acquisition 5000: There is no such thing as 0%"
Comments
l don't know.
l really don't know what l'm about to say, except l have a feeling about it.
That l must repeat the words that come without my knowledge.'
Because that's not the problem.
The problem is a noticeably very high failure percentage on Confiscate Contraband Missions.
Trust me, I know that failures show now, I get that. I also know that my failure rate on this particular assignment as more then tripled.
Either a change was made here that we weren't told about, or it's a bug.. but either way something is wrong.
Failure's have always shown for every one of my 23 toons.
Every time a toon gets it's first bunch of DOff's and you run missions with white's only, fully half the missions would fail. And right there in the results window, was the little read "failure" or even 'disaster' result.
As others have said, now however even assignments with Ultra Rare DOff's assigned are failing, something which almost never happened before.
My guess is that the so called "bug" was actually a change (or patented "Cryptic Stealth Nerf") to the RNG, so that players were less able to use DO'ffing alone to level up characters.
Ah.
But yeah, honestly, I've always felt that the odds for the "Confiscate Contraband From Crew" missions were rigged somehow: for me, they fail way too many times for what you'd statistically expect. In my case, that usually means 9 out of 10 missions should succeed. In reality, I'm lucky if I don't get a failure in like 50% of the 'rolls'.
Which can be true. But the idea that I did both of those missions yesterday and they both failed... very odd.
Sure, it is entirely possible I was just unlucky, and my assignment log does not go further back. But I do want to say that both of my characters failed that mission the day before yesterday, as well. Which again, does not necessarily mean anything and I could be remembering it incorrectly.
When the failure rate is that low (12% and 9%), and given the similar experiences mentioned elsewhere in the thread? I think it's suspicious at the very least.
I'm not jumping to conclusions or think there's any stealth nerfs or anything malicious.
I am saying that I find it extremely unusual, and worth mentioning.
I haven't had it fail yet this week.
My character Tsin'xing
I've mentioned this phenomenon before on these forums, but I don't think anyone understood what I was talking about at the time. It's interesting that several people are bringing it up now.
Are we talking about random sampling from a uniform distribution or random sampling from a Poisson distribution? It's news to me if STO uses a Poisson distribution anywhere.
This is where one has to be careful with technical language. Clusters or streaks tend to appear in random samples from a uniform distribution. It's surprising to people that random samples from a uniform distribution don't look evenly spread out. The standard meaning of "uniformly distributed" is that each outcome is equally likely. It doesn't refer to the appearance of an i.i.d. sample from such a distribution.
I'm trying to think of a way to explain the appearance of clusters. I can think of an intuitive explanation, but I'm not sure how technically correct it is. Intuitively, in order for a sample to appear evenly spread out, the members of the sample have to "actively avoid" each other and thus can't be independent.
Alright, humor me. How many of these failures in a row (with the data I've posted) would you consider normal before getting suspicious over the RNG?
I'm willing to test this for the long haul.
I think the easiest way to visualize it for me, was brought up by a programmer responsible for creating the "random playlist" feature of some music playback program. (could have been itunes or some such)
Basically he said that they originally created a true "random" outcome playlist feature, but whenever it completely randomly put 3 or more songs from the same artist together, people complained that it wasn't random.
So they had to add code to the 'randomizer' that would stop it from putting too many songs from the same artist (or whatever other variable you chose to create the playlist - genre, length, title etc.) together in a playlist ... In other words, it was no longer truly "random'.
If you toss a coin 20 times, you might not be too surprised if you got 15 heads and 5 tales.
If you toss it 200 times, you may be a little surprised if you got 150 heads and 50 tales. Toss the coin 20000 times and get 15000 heads and 5000 tales, you'd probably "perceive" it as rigged somehow, but the percentages are exactly the same each time.
To take your shooting star analogy another way... if you walked outside and saw the same shooting star night after night, you wouldn't at least consider looking up it up on the interwebs to see if there were any reports of a meteor shower or looking up observational data for that particular place in the sky you're looking at?
That's kind of where I'm at. I think I've seen some sort of unusual behavior, and now I'm cross-referencing it. It's entirely possible it really is just bad luck.
However, it is also possible it's the Freemasons using the Illuminati's RNG software to rig the game so the Lizardpeople can take over the globe with chemtrails.
Or maybe somebody just dropped a decimal point somewhere in the last patch a la Michael Bolton in Office Space.
No, I'm sorry, but this is wrong. Here's the output from Maxima.
binomial(20,5) * 0.5^20;
0.014785766601563
binomial(200,50) * 0.5^200;
2.8243676198016359*10^-13
binomial(20000,5000) * 0.5^20000;
0.0
The last probability is not really zero, but small enough that Maxima returned zero for the output. A better explanation is that if you toss a fair coin 20,000 times, you'd expect at least 19 streaks of 10 heads in a row.
Who knows, but I bet this is just a prelude to Gekos announced normalizing of the doff system.
The word probably means bringing it more in line with the current active game play experience of Star Trek Online. At least failure is a word most players who survived Delta Rising should feel accustomed to by now (bugs, lag, nerfs, crashes, crafting mods lottery, leveling, advanced pve fail criteria ec ).
We shall see and hope for the best.
Looking for a fun PvE fleet? Join us at Omega Combat Division today.
My character Tsin'xing
If you tossed a coin 20,000 times (or 200,000 times, however large you want to make that number), heads and tales will have come up about equally (and certainly will towards infinity). If it were any other way, the coins would be faulty.
So, out of 200,000 coin tosses, when you got 150,000 heads, you'd practically know something is fishy. :P
If you toss a coin 20000 times and get 15000 heads and you have any sense at all, you realize the weight of the coin is off and you stop using it in that way.
This isn't a perception problem, it's an unevenly weighted coin. The only real question is was it damaged on purpose or by accident.