< 1 2 3 4 > 
Shuttle missions AND slots update
Posted: 18 May 2017 09:36 PM   [ Ignore ]   [ # 31 ]  
Avatar
Ensign
RankRankRank
Total Posts:  221
Joined  2016-12-16
Nod - 18 May 2017 12:14 PM
[DB: Do Better] Principia - 18 May 2017 11:55 AM
Nod - 18 May 2017 11:47 AM
shahekee - 18 May 2017 11:42 AM

Shan,

Thanks for the update. Can you possibly clarify about the previous event’s success rate? Was the percentage showing on the crew selection screen off and shuttles were succeeding at an appropriate rate or was the success rate off?

Thanks!

So far there is no indication that the shown percentages and the actual success rates are off or incorrect. That said, we’ll continue looking into it, and if there is a bug, we’ll fix it. But right now there’s no indication on our end that it wasn’t working correctly. We’ve been reviewing a lot of the shuttles through our logging system, and haven’t seen any glaring errors, but investigation continues.

This would not be the first time you had a demonstrable client-end problem with one of your systems that could not be reproduced on your server end. It would behoove you to have people testing these things with the same live, real client and live, real game accounts when people are providing you with substantial proof that your system is not working as it should.

Regardless of what math you think is happening, the system is telling people for AND shuttles that the best people to send are people who don’t even have both of the AND skills listed. That’s a problem.

Sure. It’s a good point, and it’s how we’re continuing to investigate. In general when we say we’re investigating server-side, what we mean is we log exactly what a player sees on the client, and the results that were presented to them. We then look at and review the actual displayed chances, skills, and results as shown to players.

We’ll keep investigating; the feedback is not being ignored, but at this time we can’t say that there’s an issue with displayed chances. If we find something, we’ll work to identify the cause and address the issue.

I’ve had my runs of bad luck, but never a hint of systematic bias with shuttles. People get a run of bad luck and make up reasons why it isn’t their fault.

That being said, thanks for your hard work and continuing investigation into this perceived issue.

Profile
 
 
Posted: 18 May 2017 09:47 PM   [ Ignore ]   [ # 32 ]  
Avatar
Lieutenant
RankRankRankRank
Total Posts:  364
Joined  2016-02-20

So far my fleet has reported 2/4 fails at 97% percent. I can only verify my numbers. 16 runs, I’m 15/16 with an expected average success of 96.125% and currently at 2250 points. Only fail was the one below. I had the queen in the cmd and medical position instead and it was at 90%. Now I’m using more correct crew and its at 89%. We shall see how it goes. I’m doing more research this event vs just trying to be top 25.

Image Attachments
Screenshot_20170518-184348.png
Profile
 
 
Posted: 19 May 2017 02:20 AM   [ Ignore ]   [ # 33 ]  
Avatar
Lieutenant Commander
RankRankRankRankRank
Total Posts:  710
Joined  2016-09-15
Zombie Squirrel - 18 May 2017 06:32 PM

I am sorry if this has been mentioned before(didn t read back all AND SLOTS issue threads), and besides that 6/12 of my last 95%+ all AND slot missions failed(so 50%), i figured something i can t explain:

When i put my Borg Queen in a “MED AND CMD” slot, how can it be that percentages goes up the exact same way when i apply a boost, no matter if a CMD or MED boost? The Queen has no MED skill at all(and even if she has a small hidden MED skill, percentages shouldn t be the same)?!



 

Lethifold took a guess at the equation.
In it the boost is out side the skills. So it adds the same percent no matter which skill it is for.
And from your data it looks like he was right about it.

So your result is correct.
Also you would have seen the same result even if your crew has both skills.
So the skill is not adjusted first by boost. The boost is just added to the total after the multiplier.

Profile
 
 
Posted: 19 May 2017 02:26 AM   [ Ignore ]   [ # 34 ]  
Avatar
Lieutenant Commander
RankRankRankRankRank
Total Posts:  710
Joined  2016-09-15
Roonis - 18 May 2017 06:57 PM
Zombie Squirrel - 18 May 2017 06:54 PM

Thanks for the quick reply. My Queen is 5/5 LV 100 fe. , so max status.  CMD 1234

so, 1234(x3) 3702

 

5802 vs 2625

seems unlikely this would be the case, especially since I found with some testing that boosts are currently applying to slots without crew in them….

It could be top of the curve, or it more likely is a form of display issue with boosts

Lethifold came up with the following

(A+(B*(1/6))+S)*E   Though they did say they changed the multiplier.
A max skill
B min skill
S skill boost

Boost is not added to the skills it is added to total of max skill plus % of minimum skill.

All we need to do now is to determine new multiplier. 

You know I was making the same mistake. I to in checking my own example was changing the min and max skills because of boost. They do not change. Boost was outside. So boost is useless to try to get new data. We have to use two different crew to determine multiplier and third crew to make sure it checks.
Also they have to have both skills.
I am sure now Lethifold figured it out now when I originally doubted it.

Profile
 
 
Posted: 19 May 2017 03:04 AM   [ Ignore ]   [ # 35 ]  
Avatar
Commander
RankRankRankRankRankRank
Total Posts:  2370
Joined  2016-07-14
WaldoMag - 19 May 2017 02:26 AM
Roonis - 18 May 2017 06:57 PM
Zombie Squirrel - 18 May 2017 06:54 PM

Thanks for the quick reply. My Queen is 5/5 LV 100 fe. , so max status.  CMD 1234

so, 1234(x3) 3702

 

5802 vs 2625

seems unlikely this would be the case, especially since I found with some testing that boosts are currently applying to slots without crew in them….

It could be top of the curve, or it more likely is a form of display issue with boosts

Lethifold came up with the following

(A+(B*(1/6))+S)*E   Though they did say they changed the multiplier.
A max skill
B min skill
S skill boost

Boost is not added to the skills it is added to total of max skill plus % of minimum skill.

All we need to do now is to determine new multiplier. 

You know I was making the same mistake. I to in checking my own example was changing the min and max skills because of boost. They do not change. Boost was outside. So boost is useless to try to get new data. We have to use two different crew to determine multiplier and third crew to make sure it checks.
Also they have to have both skills.
I am sure now Lethifold figured it out now when I originally doubted it.

From testing the multiplier is .25, if they are behaving with the boost adding no matter what that would make sense but i hadnt looked at it yet. We were just using this example to prove the claim with how it should behave(not making a mistake), and as you can see the numbers are in question meaning it’s still not working as advertised, I.e. Adding full no matter which side gets the boost. I played with OR boosts briefly and saw some weird behavior as well that gave me the feeling this wont be easy to nail down.

Only nailed down the percent this last crop of shuttles, boosts were tomorrow lol.

 Signature 

This week Frank and Roonis talk about Picard Freaky Friday, a new Dixon Picard, how to correct when you accidently do the right thing, and Ezri whines about how hard things are.

Profile
 
 
Posted: 19 May 2017 03:42 AM   [ Ignore ]   [ # 36 ]  
Avatar
Lieutenant Commander
RankRankRankRankRank
Total Posts:  710
Joined  2016-09-15
Roonis - 19 May 2017 03:04 AM

From testing the multiplier is .25, if they are behaving with the boost adding no matter what that would make sense but i hadnt looked at it yet. We were just using this example to prove the claim with how it should behave(not making a mistake), and as you can see the numbers are in question meaning it’s still not working as advertised, I.e. Adding full no matter which side gets the boost. I played with OR boosts briefly and saw some weird behavior as well that gave me the feeling this wont be easy to nail down.

Only nailed down the percent this last crop of shuttles, boosts were tomorrow lol.

I tried using an event crew. Arik Song did not make it in my last batch of shuttles.
Was not having success trying to figure out what boost does. I think there is another multiplier on the boost depending on event crew. But they do not get full event multiplier boost.

I have no idea how the displayed success rate is calculated. I got close when I figured they just divide by the difficulty and that is the percentage.
But the difficulty changed when I had numbers with a larger then a 10% difference in success rate. I got no where when I tried using the actual sucess rate. But, I expected that, I just knew the actual success rate has nothing to do with this displayed success rate, I made some calculations, and I got Lethifold!s number almost exactly.
But, I do not have the actual math for displayed success rate so I cannot trust my data. Unless the difficulty changes with how good the crew is.

I had skill of 840 DIP and CMD 638. I put him in two different and slots by himself. One was DIP AND ENG the other was DIP AND CMD. First success rate was 29% and second was 33%.  It had a total of 3 crew slots. If you know anymore about how they display numbers maybe you can check this with your factor.

But as you said. All that will have been figured out is the behavior of the AND slots. I hope normal slots are not a problem.

I have to agree about this not going to be easy to figure the math.

Song in ENG AND SCI got 63 % and then with 5* boost did not matter which skill it was 94%. His base stat on SCI was 726 no ENG skill. 
The event victory points for all of these is 1800. It is the same shuttle this was the other slot.

I just figured what if difficulty does not change. I did a calc. for Arik and I get a multiplier of 2.5. The event crew bonus is not 3.
But there is still not full boost effect. I wonder if it is 2.5/2 or 1.25 That is it. Lethifolds factor is right. The boost for ands is half of its value.
But wait he had one skill. If he had two would it have been 2.5.
I do not know if this is happening on your data for OR.

Trying a new mission four slots CMD AND SCI. Arik’s SCI 726 CMD 214. 1800 event victory points. 52% unboosted 85% does not matter which skill is boosted. My numbers are not checking.

My results show as skills get higher the % does not go as high. Something similar to the actual success rate. But the reason I gave up on actual success rate being displayed, is I used someone with skill about 168 and the percent shown was 17% calculated it with the formula we have, and it comes out to 27% with actual. So I stopped trying to use it. I actually thought it was just simply use one or two calculations to determine difficulty and then use that for difficulty and just do skill formula fo one crew divided by that difficulty will yield displayed succcess rate actually worked on three crew slot example I had ( but all were and slots) . The 4 slot mission is gone now, I do not remember its configuration. However, probably does not matter which 4 slot mission is tried.
The Turning Point numbers just changed yet Event reward still shows 1800. Now my numbers are 40 and 69 percent when boosted (do not get it I just did it though it was 64 before) Arik as above.

Image Attachments
IMG_7541.PNG
Profile
 
 
Posted: 19 May 2017 04:50 AM   [ Ignore ]   [ # 37 ]  
Avatar
Lieutenant
RankRankRankRank
Total Posts:  372
Joined  2016-04-10

Thanks Shan and Nod for giving us an update. Appreciated.

Profile
 
 
Posted: 19 May 2017 09:28 AM   [ Ignore ]   [ # 38 ]  
Avatar
Ensign
RankRankRank
Total Posts:  120
Joined  2016-07-23
Grant77 - 18 May 2017 09:36 PM
Nod - 18 May 2017 12:14 PM
[DB: Do Better] Principia - 18 May 2017 11:55 AM
Nod - 18 May 2017 11:47 AM
shahekee - 18 May 2017 11:42 AM

Shan,

Thanks for the update. Can you possibly clarify about the previous event’s success rate? Was the percentage showing on the crew selection screen off and shuttles were succeeding at an appropriate rate or was the success rate off?

Thanks!

So far there is no indication that the shown percentages and the actual success rates are off or incorrect. That said, we’ll continue looking into it, and if there is a bug, we’ll fix it. But right now there’s no indication on our end that it wasn’t working correctly. We’ve been reviewing a lot of the shuttles through our logging system, and haven’t seen any glaring errors, but investigation continues.

This would not be the first time you had a demonstrable client-end problem with one of your systems that could not be reproduced on your server end. It would behoove you to have people testing these things with the same live, real client and live, real game accounts when people are providing you with substantial proof that your system is not working as it should.

Regardless of what math you think is happening, the system is telling people for AND shuttles that the best people to send are people who don’t even have both of the AND skills listed. That’s a problem.

Sure. It’s a good point, and it’s how we’re continuing to investigate. In general when we say we’re investigating server-side, what we mean is we log exactly what a player sees on the client, and the results that were presented to them. We then look at and review the actual displayed chances, skills, and results as shown to players.

We’ll keep investigating; the feedback is not being ignored, but at this time we can’t say that there’s an issue with displayed chances. If we find something, we’ll work to identify the cause and address the issue.

I’ve had my runs of bad luck, but never a hint of systematic bias with shuttles. People get a run of bad luck and make up reasons why it isn’t their fault.

That being said, thanks for your hard work and continuing investigation into this perceived issue.

hmm, yes .. 3 months of tracking data .. 3 months of data showing ~50% return rate on a displayed amounts of 65% and above ?  Yep, “perceived” ... ooookay.  O.o

 

Profile
 
 
Posted: 19 May 2017 09:57 AM   [ Ignore ]   [ # 39 ]  
Avatar
Ensign
RankRankRank
Total Posts:  159
Joined  2016-06-02

Sent out 12 shuttles in the past 15 hours, all with 90-99% and all with boosts, failed 6 of them, 1 of which was at 500vp (99% success apparently)!!  I know we all go through bad RNG and the odds are never what they think they are, but surely there is something wrong here, what are the odds of failing so many 99%‘ers?!! Never had such bad results. Just makes me feel like I have wasted the last couple of weeks with working to FF/FE Drone seven, Hugh, Borg Queen and now Icheb. No matter how well you do or try your still governed by the poor excuse this game uses for RNG…..

Profile
 
 
Posted: 19 May 2017 10:45 AM   [ Ignore ]   [ # 40 ]  
Avatar
Cadet
Rank
Total Posts:  42
Joined  2017-03-22

The failure chance for high success missions is even higher than last event. Most missions over 85% chance fail miserably, with Borg Queen in crew.
Last night, before tucking in, I sent 3 shuttles on 9 hours missions, all of them with over 97& chances of success, in the morning two out of three failed. Disgusting.

I am simply tired of percentages in this game. They lie. Big time. Your RNG is Redacted, please refer to our Community Participation Guidelines before posting, thank you. ˜Shan

- over 20 critical success nodes for one rare reward - EPIC FAIL!
- shuttle missions that lie worse than a politician about success rate - MEGA EPIC FAIL!
- Gauntlet that respects no common sense rules and keeps fisting you until you bleed - FAIL!
- PvP ship arena, where nothing makes sense and you blow out of the sky just because - FAIL and shameful, walk naked until I’m satisfied!

I understand most of the reasons and why you do this, but you are out of line.
I will vote with my wallet from now on.
Shame on you.

Profile
 
 
Posted: 19 May 2017 12:31 PM   [ Ignore ]   [ # 41 ]  
Avatar
Lieutenant
RankRankRankRank
Total Posts:  364
Joined  2016-02-20

So far I’m trying to avoid using “and” missions if a crew doesn’t have the other skill. I have sent out some that only had one but am trying to avoid that. This event my shown percentage is 95.04% and my actual is 95.8. I just sent out my first 4000 point shuttles after 7 runs.

Profile
 
 
Posted: 19 May 2017 12:43 PM   [ Ignore ]   [ # 42 ]  
Avatar
Lieutenant Commander
RankRankRankRankRank
Total Posts:  942
Joined  2016-04-13
Nod - 18 May 2017 11:47 AM

So far there is no indication that the shown percentages and the actual success rates are off or incorrect. That said, we’ll continue looking into it, and if there is a bug, we’ll fix it. But right now there’s no indication on our end that it wasn’t working correctly. We’ve been reviewing a lot of the shuttles through our logging system, and haven’t seen any glaring errors, but investigation continues.

What about the part where the success chance formula has been determined by community members, and it’s been shown mathematically that it’s not possible to get below 15% by way of how the equation works, yet the success chance shown on screen can still say 0%? 99.5%+ showing as 100% is not a big deal, but 15% vs 0% is a pretty big difference.

That doesn’t give me any confidence in the rest of the values being displayed.

 Signature 

My crew spreadsheet

Profile
 
 
Posted: 19 May 2017 01:32 PM   [ Ignore ]   [ # 43 ]  
Avatar
Lieutenant Commander
RankRankRankRankRank
Total Posts:  815
Joined  2016-06-15
(FF) The Sisko1 - 19 May 2017 12:31 PM

So far I’m trying to avoid using “and” missions if a crew doesn’t have the other skill. I have sent out some that only had one but am trying to avoid that. This event my shown percentage is 95.04% and my actual is 95.8. I just sent out my first 4000 point shuttles after 7 runs.

Same experience.  I only put crew in AND slots that have both skills.  Then I pay attention to predicted success rate after that.  My overall success rate is MUCH higher this event.

Profile
 
 
Posted: 19 May 2017 02:20 PM   [ Ignore ]   [ # 44 ]  
Avatar
Yeoman
RankRank
Total Posts:  57
Joined  2017-01-28

Well I was testing it out and it doesn’t look like any changes to the AND slots. See attached pics for proof.

Image Attachments
IMG_1630.PNGIMG_1631.PNG
Profile
 
 
Posted: 19 May 2017 02:26 PM   [ Ignore ]   [ # 45 ]  
Avatar
Lieutenant
RankRankRankRank
Total Posts:  364
Joined  2016-02-20

The one you showed is or. I assume she would be the same in both.

Profile
 
 
   
 < 1 2 3 4 > 
3 of 4