Just saw this, great work. I only did a formal sheet once. Rest was pen and paper for a few events and posting in our line chat. Otherwise I would update it. But typically it's 10 points lower. One event it was only 5. I was very happy
Sorry, I didn't keep specific data for this event, but if these anecdotes mean anything, I was 81-91% success rates all through the event. I had a single 4/4 group, and over all remaining groups 2 out of 5 groups came back 2/4.
For my last 8 shuttles, 7 out of 8 failed, very strongly suggesting your observation of non-rng intervention to cause failures to be accurate. I dropped from 550 with less than 4 hours to go in the event to finish at 1028. Sorry DB, there is no way I believe you that this is random or accidental.
The good news, I'm never purchasing another event pack, and with the voyage nerfs upcoming, my monthly card just got more likely to get be canceled.
overall less or equal 90% between 40% and 60%
missions 662 470 117
successes 462 296 61
failures 200 174 56
theoretical success 77.58% 70.84% 55.87%
actual success 69.79% 62.98% 52.14%
difference total success rate -7.79% -7.86% -3.74%
relative failure increase +34.77% +26.95% +8.46%
Find them in a proper table manner in the above link.
- It seems the higher the displayed success rate, the higher is the relative failure increase. So if you have 50% displayed success rate, you can expect to get something nearer to that number. If you have 90+ you can expect to have a MUCH higher increase in relative failure rate.
- The previous event (Da Vinci) was the first where I got exactly the numbers that were displayed on average. It was very slightly in the negative, but a single success more instead of a failure and it would have been slightly positive. So it was exactly what you would expect for the very first time. My guess is that there a lot of OR missions instead of AND. I think the boosts will be particularly wrong for AND missions. I did only the 2-slot missions and practically all boosts were only applied to one crew member, not both.
- The current event was a joke again. Out of a 76% displayed success rate, I was able to get out a 62% success rate. This is a relative failure increase of 57%. If you calculate the missed out VP over approx. 120 shuttles we are talking more than 50,000 VP I missed out on. That would have been the top 1000 easily. This is becoming laughable. Overall (all 662 tracked shuttles) we are at slightly below 35% increased relative failure.
- I am convinced now that there is a mechanism that tries to get you out of a reward bracket at the very end in order to make you force to spend dilithium so you will be able to still hold it. I was in around rank 600 the entire event. 5 hours before the end I sent out 4 shuttles and planned to send out 4 more with 3* Time Bonus. On average I had a 72% displayed success rate. Of those 8 shuttles, only one (1 !!!) came home successfully. I was kicked out of top 1000 about 20-30 seconds before the event ended. I was not outside the top 1000 the entire event. In the last minutes my rank was falling very quickly, which shows that DB's strategy works. Many people were spending dilithium like crazy in the final minutes to make it to the top 1000. If you do not pay up, you fall out. Of course I did not pay, since I don't own any dilithium. If you read my OP, you will see that those insane increased failing rate in the final hours happened not for the first time. It defies any logic that it happens practically every event that you spent within a rank bracket with any meaning.
I think the statistical proof is already quite stunning and revealing. I will try to consolidate some data others provided over the holidays, such that we have an even higher sample size.
What you can do is put pressure on DB not to ignore the problem any longer, this would help the cause a lot.
I am not able to say how long I will still be able to do the tracking, since I already received threats from DB to shut my voice in the forums if I don't go easier on them. They called it a constant violation of the community guidelines, which I find really funny. All I do is presenting facts as they unfold within the game. So after I am gone (which will happen, I am convinced) it is vital you keep up the fight for a game with integrity!
For my last 8 shuttles, 7 out of 8 failed, very strongly suggesting your observation of non-rng intervention to cause failures to be accurate. I dropped from 550 with less than 4 hours to go in the event to finish at 1028. Sorry DB, there is no way I believe you that this is random or accidental.
Sorry, just saw that post now.
Well, isn't that exactly what I experienced at the end of this event?
How about it DB? Will you finally do something or will denial through silence remain the order of the day?
I'll try and focus it more in excel so I can post. The formula has changed 2 times since we are aware. The first, with the expansion in crew slots and the addition of 'AND' missions. The second is a formula change with 'AND' missions that gave weight to the second skill. This is when my event shuttles went from 70 percent range to 90s. And since it's in the 90s I expected more success but had the same rate of fails as 70s. That is what got me interested.
overall less or equal 90% between 40% and 60%
missions 728 509 121
successes 507 316 63
failures 221 193 58
theoretical success 77.85% 70.91% 55.74%
actual success 69.64% 62.08% 52.07%
difference total success rate -8.21% -8.82% -3.68%
relative failure increase +37.05% +30.33% +8.31%
Overall the overall numbers are a bit worse than last time.
The faction part of the event was terrible. I had more than 80% displayed success rate but was only able to bring about 68% home (all those red shirts, lost forever). Altogether I had about 65% more shuttles failing than what I could expect. This made me think, and I enunciated the theory that the faction part of a hybrid event has always a higher relative failing rate than the (usually) following galaxy part. After all, it was the reality-defying performance I had during the Streak of Stardust hybrid event (Oct 5th - 10th) that ultimately motivated me to start my data collection. The following event, First Impressions (Oct 12th - 16th) was again a hybrid event and my performance was once again dismal. I had a higher than 130% (!!!) increase in failed shuttles there, with an average displayed chance of almost 90%. The following five pure shuttle events had all except The Daystorm Award (Nov 29th - Dec 4th) a much lower than expected rate of success (not that it was positive at the Leonardo Event, it just happened to be exactly what one could expect from the displayed numbers), but none of those event was as bad as the hybrid event. Now we had another hybrid event and once again the numbers are really bad. So, is it that in the hybrid events the bias against your odds are worse than in regular faction events? That was my first thought…after all, the big money for DB is in the (usually) following Galaxy part of the event. But I looked at the data and something else caught my attention, something I have mentioned before.
It happened to be that both hybrid events I was able to get the best displayed percentages on average of all the events I monitored so far. The first one (First Impressions) I was able to pull off almost 90% on average (my best), but got only 74% home successfully. The current event was with more than 80% not bad either, and actually my second best. All regular events I had worse average percentages displayed. And what I observed so far overall is a strong correlation between high percentages displayed and higher overall failing shuttles. Weirdly enough, the Leonardo Faction event was my best event so far (the only one I was able to finish in the top 1,000, even deep in there), although it was the event where I had the lowest overall displayed success chance.
What could be the reason for that? You should note two key numbers in my statistic. The total and the relative difference and the relative difference in shuttle failing chance. When you have a displayed success chance of 90% but are only able to pull off 80% successfully, you have a total difference of 10%. But the relative failing chance difference is 100%; You should theoretically fail 10/100 shuttles, but you are failing 20/100, which is an increase of 100%.
Lets look at our usual game plan. We try to send out as many shuttles as possible (ideally at 4,000VP/shuttle) and hope the gods of probability are on hour side such that the end result will be enough for the top 1,000.
What is DB’s agenda: They try everything possible to make the experience as pleasant and fun as possible for us, so that in the end we will all talk about what a fun event it has been, how great the background story was, how awesome the event crew is, and so forth…
…I’m just kidding of course…
…their goal is to make as much money as possible from those events and this includes (obviously, when you look at the data) manipulating and frustrating us. A certain level of frustration is actually a vital aspect for a successful pay to win game development strategy. This theory is shared by many experts on the matter (for a start check out this article: https://www.gamasutra.com/blogs/RaminShokrizade/20161123/286132/The_Price_of_Frustration.php ) Many people would probably pay up a lot less money if they could just plan out their events beforehand (at least on average).
So why is the relative fail rate higher if the displayed percentages are on average higher? Because DB wants you to lose points. If you have a theoretical success rate of 99%, but bring home only 98% of shuttles, that is actually a relative increase of 100%, but it amounts only in an additional loss of 3,200 points per 100 shuttles (@ 4,000VP/shuttle), which is negligible. If your displayed average success chance is 90%, but you bring home only 80%, that is a relative increase of 100% as well. But the actual additional loss of points per 100 shuttles amounts to 32,000 points, which is significant. This is why with a higher percentage displayed the actual relative increase of failed shuttles has to be higher than with a lower average percentage, in order to make any loss of points significant enough to frustrate you to a certain extend.
This is, I believe, why my data consistently shows a higher increase of relative failure for shuttles with high success rates compared to low success rates.
Well, of course one could come along and brush this off as conspiracy mumbo-jumbo, and sure as hell, people will. But as far as I am concerned, data speaks louder than words
With this sentiment in mind: Happy holidays everybody!
50 missions, this was a rare one when I had a higher success rate. Got some info that alot of the typical difference is due to rounding. The base crew is rounded, then an additional bonus for event is rounded, each slot has different crew with different rounding. Throw in a skill boost, it's rounded again. So a typical 3 or 4 slot mission could have 6 to 8 different rounding errors with the display. Don't know if fully true but makes sense. DB is well aware of the issue.
I too have noticed that event shuttles fail strangely often. I ran four overnight with double rewards on (regular reward 1500-1800), the percentages shown were 97, 96, 93, 89. Three failed.
Glad to know it's not just me who's noticed that. Genuinely irks me that I'll have a shuttle go out with percentages in the 90's or even high 90's only to come back as fail.
I can understand if it's at 70% or even 80% is pushing it for it to come back as a fail, but up in the 90's? SERIOUSLY!? If it wasn't for the faction events and easy freebies they offer, I would stop even doing those.
I'm glad this information is being pressed more and more. There is no doubt in my mind that there is something very very wrong with shuttle failures, whether the % chance is incorrect or there is a non-refreshing RNG seed that fails far more that it should (however this should work itself out in the long term, and it doesn't)
The number of failures I've had that are statistically highly unlikely that happen frequently is disturbing. Things like 3 of 4 failing when the lowest chance to win is 94%, this happened twice in a row in this event, its broken beyond reason. Anyone that is playing should be able to see this pattern, it is statistically significant without the need of actual statistics
Like the Gauntlet, I stopped doing the shuttle events for very long. I take some early gimmes and then abort. I'd be interested most in seeing the changes to these systems from a before/after perspective from the sale of DB. New owners of gaming properties tend to go for as heavy a monetization as possible in order to recoup as much of their outlay as they can. Nearly every game I've ever played online has degenerated after a sale precisely due to that practice. To put it another way, this isn't the same game since the sale. So whether or not the new owner lies or not to me is immaterial. They're a corporation. Lying is in their DNA. I'd be most shocked if we *weren't* getting ripped off.
This is data combined with hopes. 939 event shuttles over the last 6 months. Not every event was tracked but when it was it's almost to 1000 ran. 7.7 overall point difference. 80-89 percent had the worst difference followed by 90-94 percent shown. So it's enough data for me to officially declare its broken and we need an official DB response or fix. (I know, going on a limb saying it's broken lol)
Comments
Edit, found the one I detailed in the forums. All details in here
https://forums.disruptorbeam.com/stt/viewthread/68900/P15
This was my final ranking that event
For my last 8 shuttles, 7 out of 8 failed, very strongly suggesting your observation of non-rng intervention to cause failures to be accurate. I dropped from 550 with less than 4 hours to go in the event to finish at 1028. Sorry DB, there is no way I believe you that this is random or accidental.
The good news, I'm never purchasing another event pack, and with the voyage nerfs upcoming, my monthly card just got more likely to get be canceled.
https://docs.google.com/spreadsheets/d/1h3wLNzSDOiu-OIIUhP9OlaKPqRH2YKT83BOPHpFxFCg/edit
New numbers are:
overall less or equal 90% between 40% and 60%
missions 662 470 117
successes 462 296 61
failures 200 174 56
theoretical success 77.58% 70.84% 55.87%
actual success 69.79% 62.98% 52.14%
difference total success rate -7.79% -7.86% -3.74%
relative failure increase +34.77% +26.95% +8.46%
Find them in a proper table manner in the above link.
- It seems the higher the displayed success rate, the higher is the relative failure increase. So if you have 50% displayed success rate, you can expect to get something nearer to that number. If you have 90+ you can expect to have a MUCH higher increase in relative failure rate.
- The previous event (Da Vinci) was the first where I got exactly the numbers that were displayed on average. It was very slightly in the negative, but a single success more instead of a failure and it would have been slightly positive. So it was exactly what you would expect for the very first time. My guess is that there a lot of OR missions instead of AND. I think the boosts will be particularly wrong for AND missions. I did only the 2-slot missions and practically all boosts were only applied to one crew member, not both.
- The current event was a joke again. Out of a 76% displayed success rate, I was able to get out a 62% success rate. This is a relative failure increase of 57%. If you calculate the missed out VP over approx. 120 shuttles we are talking more than 50,000 VP I missed out on. That would have been the top 1000 easily. This is becoming laughable. Overall (all 662 tracked shuttles) we are at slightly below 35% increased relative failure.
- I am convinced now that there is a mechanism that tries to get you out of a reward bracket at the very end in order to make you force to spend dilithium so you will be able to still hold it. I was in around rank 600 the entire event. 5 hours before the end I sent out 4 shuttles and planned to send out 4 more with 3* Time Bonus. On average I had a 72% displayed success rate. Of those 8 shuttles, only one (1 !!!) came home successfully. I was kicked out of top 1000 about 20-30 seconds before the event ended. I was not outside the top 1000 the entire event. In the last minutes my rank was falling very quickly, which shows that DB's strategy works. Many people were spending dilithium like crazy in the final minutes to make it to the top 1000. If you do not pay up, you fall out. Of course I did not pay, since I don't own any dilithium. If you read my OP, you will see that those insane increased failing rate in the final hours happened not for the first time. It defies any logic that it happens practically every event that you spent within a rank bracket with any meaning.
I think the statistical proof is already quite stunning and revealing. I will try to consolidate some data others provided over the holidays, such that we have an even higher sample size.
What you can do is put pressure on DB not to ignore the problem any longer, this would help the cause a lot.
I am not able to say how long I will still be able to do the tracking, since I already received threats from DB to shut my voice in the forums if I don't go easier on them. They called it a constant violation of the community guidelines, which I find really funny. All I do is presenting facts as they unfold within the game. So after I am gone (which will happen, I am convinced) it is vital you keep up the fight for a game with integrity!
PS.
Sorry, just saw that post now.
Well, isn't that exactly what I experienced at the end of this event?
How about it DB? Will you finally do something or will denial through silence remain the order of the day?
The new numbers are as follows:
overall less or equal 90% between 40% and 60%
missions 728 509 121
successes 507 316 63
failures 221 193 58
theoretical success 77.85% 70.91% 55.74%
actual success 69.64% 62.08% 52.07%
difference total success rate -8.21% -8.82% -3.68%
relative failure increase +37.05% +30.33% +8.31%
Overall the overall numbers are a bit worse than last time.
The faction part of the event was terrible. I had more than 80% displayed success rate but was only able to bring about 68% home (all those red shirts, lost forever). Altogether I had about 65% more shuttles failing than what I could expect. This made me think, and I enunciated the theory that the faction part of a hybrid event has always a higher relative failing rate than the (usually) following galaxy part. After all, it was the reality-defying performance I had during the Streak of Stardust hybrid event (Oct 5th - 10th) that ultimately motivated me to start my data collection. The following event, First Impressions (Oct 12th - 16th) was again a hybrid event and my performance was once again dismal. I had a higher than 130% (!!!) increase in failed shuttles there, with an average displayed chance of almost 90%. The following five pure shuttle events had all except The Daystorm Award (Nov 29th - Dec 4th) a much lower than expected rate of success (not that it was positive at the Leonardo Event, it just happened to be exactly what one could expect from the displayed numbers), but none of those event was as bad as the hybrid event. Now we had another hybrid event and once again the numbers are really bad. So, is it that in the hybrid events the bias against your odds are worse than in regular faction events? That was my first thought…after all, the big money for DB is in the (usually) following Galaxy part of the event. But I looked at the data and something else caught my attention, something I have mentioned before.
It happened to be that both hybrid events I was able to get the best displayed percentages on average of all the events I monitored so far. The first one (First Impressions) I was able to pull off almost 90% on average (my best), but got only 74% home successfully. The current event was with more than 80% not bad either, and actually my second best. All regular events I had worse average percentages displayed. And what I observed so far overall is a strong correlation between high percentages displayed and higher overall failing shuttles. Weirdly enough, the Leonardo Faction event was my best event so far (the only one I was able to finish in the top 1,000, even deep in there), although it was the event where I had the lowest overall displayed success chance.
What could be the reason for that? You should note two key numbers in my statistic. The total and the relative difference and the relative difference in shuttle failing chance. When you have a displayed success chance of 90% but are only able to pull off 80% successfully, you have a total difference of 10%. But the relative failing chance difference is 100%; You should theoretically fail 10/100 shuttles, but you are failing 20/100, which is an increase of 100%.
Lets look at our usual game plan. We try to send out as many shuttles as possible (ideally at 4,000VP/shuttle) and hope the gods of probability are on hour side such that the end result will be enough for the top 1,000.
What is DB’s agenda: They try everything possible to make the experience as pleasant and fun as possible for us, so that in the end we will all talk about what a fun event it has been, how great the background story was, how awesome the event crew is, and so forth…
…I’m just kidding of course…
…their goal is to make as much money as possible from those events and this includes (obviously, when you look at the data) manipulating and frustrating us. A certain level of frustration is actually a vital aspect for a successful pay to win game development strategy. This theory is shared by many experts on the matter (for a start check out this article: https://www.gamasutra.com/blogs/RaminShokrizade/20161123/286132/The_Price_of_Frustration.php ) Many people would probably pay up a lot less money if they could just plan out their events beforehand (at least on average).
So why is the relative fail rate higher if the displayed percentages are on average higher? Because DB wants you to lose points. If you have a theoretical success rate of 99%, but bring home only 98% of shuttles, that is actually a relative increase of 100%, but it amounts only in an additional loss of 3,200 points per 100 shuttles (@ 4,000VP/shuttle), which is negligible. If your displayed average success chance is 90%, but you bring home only 80%, that is a relative increase of 100% as well. But the actual additional loss of points per 100 shuttles amounts to 32,000 points, which is significant. This is why with a higher percentage displayed the actual relative increase of failed shuttles has to be higher than with a lower average percentage, in order to make any loss of points significant enough to frustrate you to a certain extend.
This is, I believe, why my data consistently shows a higher increase of relative failure for shuttles with high success rates compared to low success rates.
Well, of course one could come along and brush this off as conspiracy mumbo-jumbo, and sure as hell, people will. But as far as I am concerned, data speaks louder than words
With this sentiment in mind: Happy holidays everybody!
I can understand if it's at 70% or even 80% is pushing it for it to come back as a fail, but up in the 90's? SERIOUSLY!? If it wasn't for the faction events and easy freebies they offer, I would stop even doing those.
>90% 2/6=33%
80-90% 8/12=67%
70-80% 5/7=71%
60-70% 3/6=50%
DB wake up and fix this anoying Bug!!!
The number of failures I've had that are statistically highly unlikely that happen frequently is disturbing. Things like 3 of 4 failing when the lowest chance to win is 94%, this happened twice in a row in this event, its broken beyond reason. Anyone that is playing should be able to see this pattern, it is statistically significant without the need of actual statistics
Alot of the data was also brought over from here
https://forum.disruptorbeam.com/stt/discussion/3467/do-they-understand-what-95-chance-to-succeed-actually-means#latest
About 5-7 shuttles are falsely lost per each event per the data I looked. I know others have better info including crew share breakdowns.