Statistical probability in the Neath seems to work according to the same surface-world principles we were once so accustomed to. That is, until chance pokes its head in the lower 90’s. I seem to fail a statistically implausible amount of times when "chance" to succeed hits 90-96%.
For example, I have failed no less than 6 (SIX!) times in a row during a seduction storylet Persuasive test in which the succeed rate was supposed to be 94%. My reaction was as follows;
1st fail: Oh Well.
2nd fail: Raised an eyebrow
3rd fail: Chuckled and shook my head
4th fail: Checked my turns candle and concluded this IS actually the 4th time I failed a 94% succeed rate challenge.
5th fail: Facial expression spasms in utter disbelief whilst making yelps of bewildered astonishment
6th fail: Surprise has now been replaced by pure, undiluted rage. Words I swore to never speak in front of my kids and I frankly forgot I even knew, pass my lips in a deluge of profanity. My wife stares at me with tears in her eyes while my 6 year old son weeps uncontrollably, his sobbing face buried in her skirt.
I am no genius in calculus but doesn’t failing a 94% challenge 6 times in succession equate to (.06*.06*.06*.06*.06*.06)*100= 0.0000046656% chance of that happening? Is that right or is that calculated in a totally different way?
And just now, I have failed a 92% challenge reading poetry to the Curate and his sister 4 times in a row and 6 times over 10 turns.
It is becoming clear to me there is enough circumstantial evidence to suggest there is indeed a 16th Master called Mr. Chance who derives some sort of depraved pleasure from lulling the residents of the Neath into a false sense of security when the odds start to hit the low- and mid-90’s.
My guess is that the Master uses a kind of quantum reality flux field trapped into a infinite cycle of self-creation and destruction which ruptures all conventional boundaries of probability and derived statistics. Not entirely unlike a black hole, but instead of using mass as a medium it rather uses the improbability of its own existence to power itself. In fact, I have once read that improbability itself may indeed even harnessed as a stardrive as long as the actual design defies all probability of functioning in the first place.
Does anyone else have similar tales of the Bermuda Triangle of Chance which resides in the shady 90-96’s, or is Mr. Chance just picking on me? edited by dharthoorn on 8/9/2014 edited by dharthoorn on 8/9/2014
I seemed to fail a certain 99% challenge about 5% of the time, but when I did a couple of other 99% challenges hundreds of times it evened out to 1% failures, so my suspicions were allayed for the moment. And I’ve never had as bad luck as you. I think the closest I’ve ever gotten to ‘impossible’ was 0.075% (link). Though with all the players of Fallen London playing for all this time, I suppose one of them is bound to have impossibly bad luck at some point. Or at several points. Have you been disrespectful to any weird old ladies lately?
Here’s a discussion of the RNG being rigged from six months ago. And here’s one from six months before that. Long story short: the RNG is a completely standard model; any random system is going to produce odd patterns; people are always going to vehemently insist that particularly evil tricks have been programmed into it to catch them out.
Oh my GOD. It’s even worse than I expected. Mr. Chance has gotten its improbability tendrils into Microsoft’s [color=33FF00]System.Random[/color] function as well.
That does indeed explain a lot. I personally recommend moving to a more secure, and more encrypted form of RNG function like for instance [color=33FF00]System.Security.Cryptography.RandomNumberGenerator[/color] or perhaps have a custom engine developed like the NSA and CIA have done. Pehaps it’s already too late and we’re all DOOMED.
With the string of highly improbable circumstances you’ve been having, the only logical course of action is to stop discussing it and instead use that time to purchase a ticket to every lottery you can find. The odds are slim, but clearly utterly in your favour.
In case, if you check your e-mail, you may have already won the Spanish lottery. It begins!
[quote=]With the string of highly improbable circumstances you’ve been having, the only logical course of action is to stop discussing it and instead use that time to purchase a ticket to every lottery you can find. The odds are slim, but clearly utterly in your favour.
In case, if you check your e-mail, you may have already won the Spanish lottery. It begins! [/quote]
Unfortunately, that would be the predictable, logical course of action, which is why it probably wouldn’t work.
The odds of failing a specific six times in a row are quite low, but if you’re doing a long string of the same action the odds of a long string of failures popping up are fairly high. (And it’s worth noting five failures in a row would probably stand out to you just as much, for example, or if it had come several actions later or earlier.)
In fact if this sort of thing didn’t happen occasionally it’d be a sign that something screwy is happening with the random number generator. (which some games will rig them to not do, since people always complain when it happens.) edited by WormApotheote on 8/9/2014
It’s my impression, which is not based on any hard data, that sometimes if a single action is performed repeatedly in rapid succession the result does not vary for 5 or 6 attempts (the RNG does not reset?).
The "System.random" function which EB-FL apparently uses as their RNG should be called and executed for each challenge individually. I would be truly amazed if it was not.
Streaks do stand out in and of themselves. Apparently I have been extremely unlucky because I have had multiple 90~96% losing streaks and have yet to succeed even 2 times in a row on a rare success in item conversion let alone 4,5 or 6 times. Having said that, I do not know for sure if the odds are the same. I believe rare conversion successes are about 5%, aren’t they?
I personally think long streaks of either winning or losing should be curtailed in the code. At the end of the day, I get much more upset from an extremely unlikely losing streak than I get happy from against-all-odds winning streak. And whether or not people in general (myself included) LIKE the game ultimately decides its Fate, not statistical or mathematical "correctness".
The fact that there is a reasonable-sized player-base does not inherently mean the RNG system (or any game feature, for that matter) is beyond improvement. Many games have tweaked their probability rates beyond the naked math, and are more enjoyable for it. edited by dharthoorn on 8/10/2014
I’ve noticed this happening a lot, but I’m not really convinced that’s not just a perspective bias (ie remembering when it happens but not when it doesn’t).
If it is happening it’s probably a bug somewhere that makes it think it’s just returning data on the previous attempt again. (ie it’s not a problem with the random number generator being not random, but the game getting confused about whether it should call it again.) edited by WormApotheote on 8/10/2014
If you repeat a storylet very fast via the "Try again" button the RNG sometimes seems stuck to one "result". Playing another storylet or using the "Onwards" button solves this "problem".
I also noticed something similar with the card deck. If you flip through the cards using the discard button and drawing new ones very quickly you have a pretty high chance to draw the same cards again and again. I now alternate between drawing cards und playing storylets which in my view gives me more "random" results.
Thanks all to those who’ve commented on my observations! I’ve taken just enough stats to be dangerous (i.e., to know that I may only remember bad luck strings and not good luck strings…not that this would stop me from only remember bad luck!). Still, it sure feels like the RNG/call function gets “stuck” sometimes. Nowadays I take a pause after two or three consecutive fails, and keep quickly repeating on successes…just in case it matters.