I'll suggest again, help contribute to the data I'm collecting. It includes filtering event vs non-event shuttles.
The intent is to do a percent-by-percent comparison, so eliminate skew from range comparisons.
I stopped reading, someone may have been more to the point.
But many of us are thinking the error in success occurs only when using boosts.
That is, you boost a skill the crew does not have and it increases success percent.
I remember one poster said they were treating the boost as not working if they did not have the skill. They made their own calculations and were getting numbers that were much closer to the mean.
I really do not worry about it myself, I just cherry pick my mission s for the event crew i have.
My 4000 + mission are all non event with no boosts used. The last 2000 missions show a deviation between actual and displayed percentages, where the first 2000 they were pretty close.
I discovered in this last event, regardless of displayed success rate, one of which failed displaying 99%, that all shuttles failed once each threshold was reached. I was able to predict how many were going to fail before I sent them. I was 100% correct for the entire event. Did anyone else notice this? I was running four shuttles at a time all were above 70%. Most were above 80%. One threshold was attained with 200 points over after claiming two shuttles and the reamaining two, both 85%, failed. First time I've seen this happen.
I wonder what the odds are of sending out four shuttles at over 95% and getting three failures. This has just happened to me.
I'm not stupid, I understand that they all had a 5% chance of failing, I understand gambling and am not trying to say "they should have won". Simply curious. What are the odds?
General rule of thumb, as far as STT is concerned, is to send shuttles having 80-85% chance to succeed. Those seem to do best, even if math rules don't seem to properly apply. Quite honestly, I lost more shuttles of 90%+ than 66-75% on the average.
I think there are two things going on here, two possible explanations, and one definite DB lie.
First going on, DBs RNG sµcks (sone of this marinating tofu B$). It sounds like hyperbole and an exaggeration, but is neither. This is the worst RNG I've encountered in my life.
Second going on, shuttle rules for actual success remain undisclosed, and it's likely that only one or the first slot of an AND and/or an OR slot gets counted for the full stat value. E.g. Med AND Dip dip always counts for 1/4. Likewise, Dip and Med med always counts 1/4, only dip gets full value.
First explanation, DB screwed up the coding and the client doesn't match the server. The client displays calculated success, but has to ask the server for results. The server doesn't use the same formula as the client to determine success. Exceedingly likely.
Second explanation, DB is handicapping shuttles on the server towards a mean. Like the "rubber band" effect on racing games that keeps the AI always within a certain distance from the player, or pulls the AI towards the player from behind or ahead, DB pulls the actual success rate towards 50%.
The Lie, DB refuses to tell us which. They know displayed doesn't match actual, there's no way after 2+ years and countless user interactions that they can not know. Weather it's by intentional design, or coding failure, knowledge with addressing the issue is a lie. Sometimes I'm not sure which lie is worse . . .
I honestly wonder sometimes if this is why DB won't disclose the rules and mechanics. They know that the coding is so bad that the intended mechanics don't match the actual mechanics, and if they told us what was supposed to happen, then there'd be written proof that they're wrong.
General rule of thumb, as far as STT is concerned, is to send shuttles having 80-85% chance to succeed. Those seem to do best, even if math rules don't seem to properly apply. Quite honestly, I lost more shuttles of 90%+ than 66-75% on the average.
after tracking several Faction Events, I found that the majority of shuttle failures happen between 2250VP and 2750VP regardless of success rates shown or boosts. I used the exact same crew in one faction event to prove it. I tracked (ad nauseum) every possible detail related to shuttle mission, and the same missions that failed repeatedly at the 2250-2750 mark, succeeded with no boosts at the 3500-4000 mark.
Was hoping to get another couple of McCoy's during this event but it looks like threshold and out for me as I'm not wasting my time to get another Vorik. Good luck to everyone else.
Was hoping to get another couple of McCoy's during this event but it looks like threshold and out for me as I'm not wasting my time to get another Vorik. Good luck to everyone else.
...on the other hand do you think DB would ever let your shuttle win a 5% success rate as much as they award that probability to themselves? Pure garbage
General rule of thumb, as far as STT is concerned, is to send shuttles having 80-85% chance to succeed. Those seem to do best, even if math rules don't seem to properly apply. Quite honestly, I lost more shuttles of 90%+ than 66-75% on the average.
after tracking several Faction Events, I found that the majority of shuttle failures happen between 2250VP and 2750VP regardless of success rates shown or boosts. I used the exact same crew in one faction event to prove it. I tracked (ad nauseum) every possible detail related to shuttle mission, and the same missions that failed repeatedly at the 2250-2750 mark, succeeded with no boosts at the 3500-4000 mark.
It's in the coding folks.
I experienced that yesterday & thought I had done something to anger RNGesus. It set me back about 9 hours, but I can live without another 1/5* belowdecks. 🖖🏻
#crewslotsale
#freeStellar
"In the short run, the game defines the players. But in the long run, it's us players who define the game." — Nicky Case, The Evolution of Trust
I discovered in this last event, regardless of displayed success rate, one of which failed displaying 99%, that all shuttles failed once each threshold was reached. I was able to predict how many were going to fail before I sent them. I was 100% correct for the entire event. Did anyone else notice this? I was running four shuttles at a time all were above 70%. Most were above 80%. One threshold was attained with 200 points over after claiming two shuttles and the reamaining two, both 85%, failed. First time I've seen this happen.
After reading this during the week I remembered to make a note of whether this was correct and certainly for me it was, first shuttle pushed me over the limit for Surak other 2 both failed in spite of them supposedly each having >85% chance of success. Though far from conclusive certainly enough data to make me suspicious...
Can anyone else confirm??
Was hoping to get another couple of McCoy's during this event but it looks like threshold and out for me as I'm not wasting my time to get another Vorik. Good luck to everyone else.
Same f*ing thing here. I've sent out 3 shuttles at over 95% and 90% of the time the 3rd one fails. I've been stuck at 2250. I think since reaching 2250 the 3rd has failed around 7 out of 7 times. I wish I kept track. Also a lot of failures at 1500 too. Those saying it's just "chance" are 100% wrong, although in DBs case that would mean 0%.
I discovered in this last event, regardless of displayed success rate, one of which failed displaying 99%, that all shuttles failed once each threshold was reached. I was able to predict how many were going to fail before I sent them. I was 100% correct for the entire event. Did anyone else notice this? I was running four shuttles at a time all were above 70%. Most were above 80%. One threshold was attained with 200 points over after claiming two shuttles and the reamaining two, both 85%, failed. First time I've seen this happen.
This is the most likely reason I've seen in months. I'm going to pay close attention to this. If this is the case, it is not maliciousness by DB, it is a programming issue that can be found and corrected. Good work, and thanks for sharing.
My 3rd shuttle is still failing around 90% of the time. This is with the chance of success being 90% or more. I usually reach 4k vp but this event I've only reached 3500 sunday when my chance of success for all 3 shuttles range from 87% to 90%. I can count on 1 hand how many success the 3rd shuttle has gotten.
My 3rd shuttle is still failing around 90% of the time. This is with the chance of success being 90% or more. I usually reach 4k vp but this event I've only reached 3500 sunday when my chance of success for all 3 shuttles range from 87% to 90%. I can count on 1 hand how many success the 3rd shuttle has gotten.
How about posting the mission and crew configuration so we can see if the prevailing theory on what is wrong explains the fails
I had 1 failure the entire event. And last faction only event i had 5 total failures. All works as intended, as long as you stick to the AND slot strategy.
I had 1 failure the entire event. And last faction only event i had 5 total failures. All works as intended, as long as you stick to the AND slot strategy.
Yeah similar here, 2 fails total so far out of around 100 4ks
What is the AND slot strategy? I went 4/4 exactly once after reaching 4k shuttles and my average chance to succeed was shown at 88%
Treat and slots as first stat plus 1/4 of second stat.
This is different than the on screen display which treats them as high stat plus 1/4 of low stat.
I did this this time, many times slots even required not using bonus crew even though non-matching bonus crew looked like a higher chance to succeed. I had MUCH higher success rates than by believing what DB shows in the client.
I'm willing to grant back-end access to anyone who has large data sets to contribute; PM me.
So far, the data is far from extensive enough to draw firm conclusions, but the current average difference between displayed and real is <0.2%
.....and went 3/4 again on my 92.0%, 94.3%, 93.8%, 93.2% set of shuttles
I'd be glad to dump in the data from this trial, (not too thrilled on manually entering 87 data points). That said, according to a a binomial distribution in excel, the probability of my data occurring naturally is 0.0000000423321980, or less than 1 in 1,000,000,000
edit, formula used:
=BINOM.DIST(63,87,0.92,0)
This screams to me that there is some kind of bug or distribution occurring.
Comments
You're well aware that this result is completely insignificant.
The intent is to do a percent-by-percent comparison, so eliminate skew from range comparisons.
https://goo.gl/forms/aeeVQviXahWFGc7q1
It's as significant as any other 42 shuttle runs in the data set.
Which is to say, insignificant.
On that note, I'm out.
I'll miss the community but certainly not the frustration.
Live long, prosper, and have fun (what demented Vulcans say)
My 4000 + mission are all non event with no boosts used. The last 2000 missions show a deviation between actual and displayed percentages, where the first 2000 they were pretty close.
I'm not stupid, I understand that they all had a 5% chance of failing, I understand gambling and am not trying to say "they should have won". Simply curious. What are the odds?
Higher than winning at Dabo?
First going on, DBs RNG sµcks (sone of this marinating tofu B$). It sounds like hyperbole and an exaggeration, but is neither. This is the worst RNG I've encountered in my life.
Second going on, shuttle rules for actual success remain undisclosed, and it's likely that only one or the first slot of an AND and/or an OR slot gets counted for the full stat value. E.g. Med AND Dip dip always counts for 1/4. Likewise, Dip and Med med always counts 1/4, only dip gets full value.
First explanation, DB screwed up the coding and the client doesn't match the server. The client displays calculated success, but has to ask the server for results. The server doesn't use the same formula as the client to determine success. Exceedingly likely.
Second explanation, DB is handicapping shuttles on the server towards a mean. Like the "rubber band" effect on racing games that keeps the AI always within a certain distance from the player, or pulls the AI towards the player from behind or ahead, DB pulls the actual success rate towards 50%.
The Lie, DB refuses to tell us which. They know displayed doesn't match actual, there's no way after 2+ years and countless user interactions that they can not know. Weather it's by intentional design, or coding failure, knowledge with addressing the issue is a lie. Sometimes I'm not sure which lie is worse . . .
after tracking several Faction Events, I found that the majority of shuttle failures happen between 2250VP and 2750VP regardless of success rates shown or boosts. I used the exact same crew in one faction event to prove it. I tracked (ad nauseum) every possible detail related to shuttle mission, and the same missions that failed repeatedly at the 2250-2750 mark, succeeded with no boosts at the 3500-4000 mark.
It's in the coding folks.
Nine failures.
It's just not logical.
Was hoping to get another couple of McCoy's during this event but it looks like threshold and out for me as I'm not wasting my time to get another Vorik. Good luck to everyone else.
...on the other hand do you think DB would ever let your shuttle win a 5% success rate as much as they award that probability to themselves? Pure garbage
I experienced that yesterday & thought I had done something to anger RNGesus. It set me back about 9 hours, but I can live without another 1/5* belowdecks. 🖖🏻
#crewslotsale
#freeStellar
Can anyone else confirm??
Same f*ing thing here. I've sent out 3 shuttles at over 95% and 90% of the time the 3rd one fails. I've been stuck at 2250. I think since reaching 2250 the 3rd has failed around 7 out of 7 times. I wish I kept track. Also a lot of failures at 1500 too. Those saying it's just "chance" are 100% wrong, although in DBs case that would mean 0%.
Nah completely wrong. There's a bug in their code somewhere.
This is the most likely reason I've seen in months. I'm going to pay close attention to this. If this is the case, it is not maliciousness by DB, it is a programming issue that can be found and corrected. Good work, and thanks for sharing.
How about posting the mission and crew configuration so we can see if the prevailing theory on what is wrong explains the fails
Yeah similar here, 2 fails total so far out of around 100 4ks
Treat and slots as first stat plus 1/4 of second stat.
This is different than the on screen display which treats them as high stat plus 1/4 of low stat.
This was the outcome of my 4k shuttles, and I will tell you, I about lost my mind at these odds:
https://goo.gl/forms/arKyTBjQGMNa63fW2
I'm willing to grant back-end access to anyone who has large data sets to contribute; PM me.
So far, the data is far from extensive enough to draw firm conclusions, but the current average difference between displayed and real is <0.2%
I did this this time, many times slots even required not using bonus crew even though non-matching bonus crew looked like a higher chance to succeed. I had MUCH higher success rates than by believing what DB shows in the client.
.....and went 3/4 again on my 92.0%, 94.3%, 93.8%, 93.2% set of shuttles
I'd be glad to dump in the data from this trial, (not too thrilled on manually entering 87 data points). That said, according to a a binomial distribution in excel, the probability of my data occurring naturally is 0.0000000423321980, or less than 1 in 1,000,000,000
edit, formula used:
=BINOM.DIST(63,87,0.92,0)
This screams to me that there is some kind of bug or distribution occurring.
There is bad luck and then there is this......