Getting to the bottom of ANDGate
![[BL] PShivers](https://us.v-cdn.net/6030345/uploads/userpics/013/nQE80BENLYES4.jpg)
[BL] PShivers
✭✭✭
in The Bridge
I wanted to start a thread for anyone interested in investigating the AND issue. I started doing missions and quickly realized that I am not sure which ones I should be running and how. If those who did the initial gathering of info would like to share which missions they ran and the criteria so others can help, that would be great. If those people would like to run the statistics again please let the rest of us know here where to send our data to be analyzed. Any help would be greatly appreciated.
0
Comments
But here is the foundational post, they specify which missions they ran and every thing, its a good writeup, See https://forum.disruptorbeam.com/stt/discussion/5097/edit-final-results-2nd-experiment-results-and-shuttles-are-absolutely-bugged/p1
Yes. Showing that specific bug isn't active anymore is actually pretty easy to do since you can get a display/actual difference on the order of 50%.
However, if your intent is to show that displayed success is accurate, it will take thousands of runs to have much confidence in the numbers.
That could be true, except for the fact that the displayed success rate is/was wrong.
The displayed expected % did not match the observed success rates by a significant margin. It is not possible that there wasn't a bug here.
They DID explain how it was supposed to work and the front end success % reflected their public description.
My wife takes me to task when I use the phrase "rule of thumb" even though there's no proof its origins are what she says. It's plausible though so I don't say it anymore. Domestic tranquility and all that.
HAHAHAH ... all i can think of is someone joining a conference call 10 minutes late and saying...
"i just joined... can you start again"
Captain Kathryn Janeway i suggest you look down in the forum and you will find multiple conversations about and describing it.
It’s probably best to start with the original thread, here: https://forum.disruptorbeam.com/stt/discussion/5097/edit-final-results-2nd-experiment-results-and-shuttles-are-absolutely-bugged/p1
Thanks. Dang, this is pretty serious...
The tl;dr version is that the displayed success percentage and the real success percentage for shuttle missions were not the same number due to a problem with how each crew’s skills were added together for two-skill AND slots (single-skill and OR slots were not affected). This caused countless shuttles to fail unexpectedly, unfairly depriving people of a litany of rewards over a months-long period.
DB denied the issue existed until a lot of people came together and generated statistically valid data proving that there was a problem, at which point DB ceased acknowledging the entire debacle and then slipped what appears to be a stealth fix into a server update some time later. To date, we have never received any sort of closure, never mind an apology or restitution...and some of us at this point don’t even care about refunds on failed transmissions or Faction-only items, we just want some honest, forthright communication on the error that helps everyone, DB included, move on.
WOW... that is bad. Affected players deserve an apology and compensation.
Not all of us think so.
Could you please continue the petty bickering? I find it most intriguing.
~ Data, ST:TNG "Haven"
Excellent summary, thank you, mind if I copy it for future use?
I would add - we don’t know if there’s been a stealth fix, there’s not been any serious testing. Subjectively I feel the problem still exists.
Go for it, and change whatever you need to change.
Personally, I still try to populate shuttles by first/second skills rather than high/low. It doesn’t always work out and sometimes a crew member’s skills are so close it doesn’t matter that much (Mirror Picard, Spider Barclay, and Mirror Cochrane come to mind)...but darn it all, it’s a matter of principle.
Seat #1 - SEC & MED
Seat #2 - MED & SEC
Option 1 - they have the same calculation, since order doesn't matter and it's 50/50 split
Option 2 - skill #1 + (skill #2 * 0.25) -- this is the generally accepted forum approach
Option 3 - (skill #1 * ???) + (skill #2 * ???) -- similar to option 2, but with unknown calculations for one/both skill(s)
Another issue that comes to mind is that the auto-sort of crew still seems to be based on option 1, which could very well lead people to select crew that are not optimized for the seat/shuttle.
Yet another issue is the impact of boosts. Depending on which option is correct, do boosts work differently when boosting skill #1 or skill #2? Does the visible boost in front line success rate correctly match the actual back end increase in success rates that the boost provides?
Soooooo many unanswered questions...
Once I switched to the forum approved method [1+(0.25*2)], I found my actual results were within 1-2% of the expected results, sometimes even slightly higher.
Fixed or not, I continue to the use that approach and rank my crew before each event. During last week's event all my shuttles were in the 91-93% posted success rate range, using no more than 3* skill boosts, and I think I only failed about 4 shuttles throughout the entire event. Until I get official confirmation, I'm just going to keep using the forum approach.
You don't think it's serious that DB was misleading players about percentages on things that people spent real money on? And denied that anything was wrong for months? And then when confronted with proof, they released a new set of graphics for shuttle completions after a few weeks, and may or may not have stealth fixed their error, without acknowledging it ever existed?
I mean, you be you and all that, but I find that position hard to understand.
What apparent stealth fix?
A while back they changed shuttles to say ‘mission completed’ instead of ‘failed’, but I can’t recall any explicit mention, verbally or visually, of a fix.
That's why it's a "stealth" fix.
Some of us noticed a decrease in shuttle failures related to AND slots after that change went live.
That is the point. There is no mention of it. The AND issue apparently stopped being a problem (it seems fixed to me, anyway) at the same time they put in new visuals.
But until DB announce it, or we spend hours of our own time performing another scientific study, it’s at best subjective. So I wish we, as those most aware of the issue in the first place, would stop assuming this has been fixed.
My own experiences in Shuttle events have been that it’s not fixed. I have a good-high depth of crew (379 immortals) and it feels like I’m failing slightly more missions than I should. I say slightly, because I think I’m tuned now to loading slots in the workaround way when I remember to (‘most of the time’). Other times I do not.
I care, mostly because I am anally retentive and like the displayed success rate to match the actual success rate. Otherwise my OCD kicks in and I start getting really angry.
If we get any compensation, I'd like a year's counseling for frustration issues and that includes the gauntlet
Check out our website to find out more:
https://wiki.tenforwardloungers.com/
Like I said, I realize this is not a popular opinion around here, which is why I haven't said anything up until now. (The only reason I finally mentioned it is because every post makes it seem like every single player is up in arms about it, and I wanted to share my point of view — that there was at least one outlier.
Could you please continue the petty bickering? I find it most intriguing.
~ Data, ST:TNG "Haven"
What sample size of something that is supposed to be 80% coming in below 30% do you need for it to have been conclusive.
While it was hard to get statistically significant results on event shuttles with three slots, one slot fed shuttles with an and mission had huge variations like this and were completely repeatable by all who tested them.