Powerplay PowerPlay and a basic example of how it's broken

[Martin Schou]

When I say "broken" I don't mean "I believe these rules should be changed". I quite simply mean "this does not work as advertised".

As usual, after a power has lost some systems (this cycle Yuri Grom has lost 5 control systems) the PowerPlay community discovers that previously contested CC income hasn't been returned to powers that exploit the previously contested systems.

According to one group those five systems contested 236 CC that should have been returned to other powers but hasn't been.

So me, being the curious bloke that I am, decided to check my own numbers. My numbers do not compare to the previous cycle, just what the game says to what we can calculate via resources out of the game like EDDB.

And from what I can tell there are 65 exploited systems missing out of 8,034, ranging from -2 (have more exploited systems than the game claims) to 21 systems. While that's less than 1%, but it should really be 0. This means 1,252 CC are missing across the powers, ranging from 33 CC to 308 CC (between 0.96% and 5.30% of each power's income). Notice that even though one power apparently has more exploited systems than I can calculate, no power has higher income than it should have.

With 8,034 exploited systems, this is not something that can easily be checked by hand. It is also not possible to take shortcuts in figuring out where the problem is.

And the cause of this? PowerPlay is broken. It's been broken since day one, and every single time we point out that CC is missing somewhere the broken part isn't fixed, the CC is simply added somewhere.

So far we're 347 weeks into PowerPlay, and the most basic rules of it remain as broken now as they were on day one. The only difference is that we didn't realise it at the time.

 

[cmdr Ramius007]

It's not example of broken power play, but rather shows level of bug infestation, and a lot of those horizons bugs come along with ody and all following updates

 

[Danieros]

It is really unfortunate... the game still cannot be trusted to perform its basic updates correctly - properly returning CC income to spheres which have systems no longer being contested.

A major event happens, it totally malfunctions and there's no Frontier announcement about it.

 

[Martin Schou]

It's not example of broken power play, but rather shows level of bug infestation, and a lot of those horizons bugs come along with ody and all following updates

It actually is though, because when the game numbers cannot be trusted, and we cannot trust the game to correctly reflect events, then it is broken.

Remember when NPCs could end up with beam lasers that fired plasma or railgun rounds, ripping anyone and anything to shreds within seconds? Sure, it's a bug, but it also meant that any and all NPC combat was broken.

 

[Mr.ziPPeR]

It is really unfortunate... the game still cannot be trusted to perform its basic updates correctly - properly returning CC income to spheres which have systems no longer being contested.

A major event happens, it totally malfunctions and there's no Frontier announcement about it.

They can't even deliver CG reward within 2 weeks or fix the NPCs not turning hostile in Odyssey, so don't have any kind of expectations about PowerPlay...

Even though I'd like they prove me wrong.

 

[Rubbernuke]

Powerplay has had loads of bugs from Day 1 onwards, as well as server issues that impacted play.

The issue is Powerplay is not and should have never been considered a 'one and done' feature but one that requires constant balance and maintenance. It was released incomplete and complex bugs left rather than tackled and here we are with a broken mess.

 

[Martin Schou]

Well, since I have absolutely nothing better to do, I decided to do a full review of Zemina Torval's exploited systems.
EDDB tells us that there are 244 such systems: https://eddb.io/system?p=4&o=32

This is the data I'm using for my checks, and it has two mistakes compared to the galaxy map: Scorpii Sector DB-X b1-8 and Wolfsegen. Both are detention centres and those are unexploitable. Fortunately I calculated their CC value to 0, so this has no impact on the apparent missing income.

That leaves 242 systems that the galaxy map claims are exploited by Zemina Torval. But the PowerPlay menu claims she has 231 exploited systems. That means there are 11 systems that Torval is not getting income from.

The PowerPlay menu says she has an income of 1,904 CC, but I calculate 2,005 CC - a difference of 101 CC. The game claims she's running a default deficit of 200 CC, but with 101 CC missing this should be 99 CC.

Let's see if I can figure out where the income is missing. This isn't easy, because I cannot know where the income from a system that is within 15 light years of more than one control system is supposed to go, and Torval has 18 such systems - more than the systems she's not getting income from. But let's give it a go anyway.

Control System	Game Income	Calculated Income without overlapping spheres	Missing
Kappa	128	131	3
Xuanduna	96	111	15
LTT 8517	69	93	24
HIP 98621	68	71	2
Trante	41	64	23
Tepertsi	58	59	1
Secoya	58	60	2
Nanabozho	27	28	1
Ross 429	23	24	1
Ullese	20	21	1
Herthans	64	75	11

That's a total difference if 82 CC of the 99 CC that are missing, which means the remaining 17 CC are somewhere in the overlapping spheres.

That is 10 control systems that have missing income without overlapping spheres. Since Torval only has 34 systems that means the game is wrong in AT LEAST 29% of Torval's control systems. If we assume Torval isn't remarkable in this mess of errors, we can expect that 29% of the game's 699 control systems are reporting the wrong income. That's 202 control systems - more than the two largest powers combined (199 control systems) and more than the four smallest powers combined (175 control systems). This is an insane amount of control systems that we can expect to have the wrong income.

I am using Torval as the example here, because she is the smallest power by far. She has 20% fewer control systems than Archon Delaine (2nd smallest power) and 47% fewer exploited systems than Archon Delaine, making her the easiest power to check. Even though I "only" had to check 244 exploited systems and 34 control systems, this still took well over an hour for me to check, because EVERYTHING has to be done manually. I don't have the ability to pull data from Frontier's own data sets.

If I wanted to check everything I'd have to check 7,546 exploited systems and 699 control systems, and Torval only represents 3.4% of that.

Here's how bad it is - some powers are at a default deficit in game and they shouldn't be.

Power	Default Balance	Missing Income	Actual Balance
Aisling Duval	-17	143	+126
Archon Delaine	-728	63	-665
Arissa Lavigny-Duval	-368	308	-60
Denton Patreus	-29	155	+126
Edmund Mahon	-85	154	+69
Felicia Winters	-130	33	-97
Li Yong-Rui	-380	68	-312
Pranav Antal	-291	71	-220
Yuri Grom	-436 (Turmoil, so this is unreliable)	61	-375 (in turmoil, so this is still unreliable)
Zachary Hudson	-289	95	-194
Zemina Torval	-200	99	-101

Three powers that are currently running a default deficit SHOULD be running a default surplus: Aisling Duval, Denton Patreus and Edmund Mahon.

And all powers are far closer to a default surplus than the game claims, with some being so close they can pretty much taste it (Arissa Lavigny-Duval and Felicia Winters).

And as I said, this isn't new. This isn't something that's just happened on the last cycle tick on Thursday morning at 7 AM GMT.

And every time we've pointed out these types of problems, the "fix" is simply to do some kind of weird recalculation within PowerPlay but keeping the bugs that causes this to remain broken.

 

[Rubbernuke]

Well, since I have absolutely nothing better to do, I decided to do a full review of Zemina Torval's exploited systems.

I never thought I read that sentence in my lifetime.

 

[Misaniovent]

I don't really pay much attention to powerplay, but I was excited to see Gliese 900.1 in turmoil. I was not excited to see that after it revolted, Patreus was not refunded the significant amount of CC Gliese 900.1 contested. Unfortunately, powerplay is burdened by a series of unsuccessful attempts to fix calculation bugs:

In 2017, Noxa and I realized that there was a major issue with CC recalculations when powers dropped overlapping spheres: if CS A overlaps CS B and CS B revolts, CS A will receive the value of the shared CC no matter what — even if it was already receiving that value. Said differently, the calculation seemed to always think: "this is no longer shared CC, therefore the control system exploiting it should receive it." I submitted a pretty detailed bug report directly to BrettC (at his request) at this time.

This issue was introduced in February 2016 in response to an attempt to fix an issue discovered in September 2015, where exploited system income was not being reallocated at all after revolts. Unfortunately, my link to that original issue (discovered by ALD) is dead so I can't detail it further. Thanks @edvordo!

Frontier tried to fix the shared CC bug in late 2019 and I believe they may have inadvertently introduced the deflation bug we seem to be experiencing now at that time. When they attempted to fix this issue again last year, they either made a mistake or simply tried to band-aid it with a manual recalculation.

 

[Auronio]

When I say "broken" I don't mean "I believe these rules should be changed". I quite simply mean "this does not work as advertised".

As usual, after a power has lost some systems (this cycle Yuri Grom has lost 5 control systems) the PowerPlay community discovers that previously contested CC income hasn't been returned to powers that exploit the previously contested systems.

According to one group those five systems contested 236 CC that should have been returned to other powers but hasn't been.

So me, being the curious bloke that I am, decided to check my own numbers. My numbers do not compare to the previous cycle, just what the game says to what we can calculate via resources out of the game like EDDB.

And from what I can tell there are 65 exploited systems missing out of 8,034, ranging from -2 (have more exploited systems than the game claims) to 21 systems. While that's less than 1%, but it should really be 0. This means 1,252 CC are missing across the powers, ranging from 33 CC to 308 CC (between 0.96% and 5.30% of each power's income). Notice that even though one power apparently has more exploited systems than I can calculate, no power has higher income than it should have.

With 8,034 exploited systems, this is not something that can easily be checked by hand. It is also not possible to take shortcuts in figuring out where the problem is.

And the cause of this? PowerPlay is broken. It's been broken since day one, and every single time we point out that CC is missing somewhere the broken part isn't fixed, the CC is simply added somewhere.

So far we're 347 weeks into PowerPlay, and the most basic rules of it remain as broken now as they were on day one. The only difference is that we didn't realise it at the time.

I already was happy if the user unfriendly merit acquisition mechanic was changed to the way we can buy/sell commodities.

Of course that doesn't translate into endorsing the mentioned CC calculation bug.

 

[Martin Schou]

I already was happy if the user unfriendly merit acquisition mechanic was changed to the way we can buy/sell commodities.

I can't remember how long ago it was, but I did some timed testing to see how long it takes to fill a Cutter with merits compared to goods. It took about 4 seconds to fill it with goods, and I think about 4 minutes to fill it with goods if you only had a 10 merit allocation.

 

[Auronio]

(I can't remember how long ago it was, but I did some timed testing to see how long it takes to fill a Cutter with merits compared to goods. It took about 4 seconds to fill it with goods, and I think about 4 minutes to fill it with goods if you only had a 10 merit allocation.

It's ridiculous.
They could simply make it buy (and sell) how many you want - with commodity mechanics - and subtract the value of the 30' credit allocation.
Utmost opposite of QoL mechanic.

 

[edvordo]

I don't really pay much attention to powerplay, but I was excited to see Gliese 900.1 in turmoil. I was not excited to see that after it revolted, Patreus was not refunded the significant amount of CC Gliese 900.1 contested. Unfortunately, powerplay is burdened by a series of unsuccessful attempts to fix calculation bugs:

In 2017, Noxa and I realized that there was a major issue with CC recalculations when powers dropped overlapping spheres: if CS A overlaps CS B and CS B revolts, CS A will receive the value of the shared CC no matter what — even if it was already receiving that value. Said differently, the calculation seemed to always think: "this is no longer shared CC, therefore the control system exploiting it should receive it." I submitted a pretty detailed bug report directly to BrettC (at his request) at this time.

This issue was introduced in February 2016 in response to an attempt to fix an issue discovered in September 2015, where exploited system income was not being reallocated at all after revolts. Unfortunately, my link to that original issue (discovered by ALD) is dead so I can't detail it further.

Frontier tried to fix the shared CC bug in late 2019 and I believe they may have inadvertently introduced the deflation bug we seem to be experiencing now at that time. When they attempted to fix this issue again last year, they either made a mistake or simply tried to band-aid it with a manual recalculation.

I don't have anything useful to add, just wanted to give a working link to the "dead" forum post mentioned in third paragraph.

 

[Misaniovent]

I don't have anything useful to add, just wanted to give a working link to the "dead" forum post mentioned in third paragraph.

Updated, thanks!

 

[Auronio]

I don't have anything useful to add, just wanted to give a working link to the "dead" forum post mentioned in third paragraph.

Don't know whether I should laugh or cry considering that was 2015.
Telling clearly how much attention Powerplay receives.
Anyway Lads - enjoy your rest of the weekend and next week.
o7

 

[PhoebiousZ]

When I say "broken" I don't mean "I believe these rules should be changed". I quite simply mean "this does not work as advertised".

You sound surprised; you shouldn't thought. Typical frontier culture here, nothing new about it. Galnet is broken since I remember playing this game, even the starting balance of some powers you mention is wrong (not your fault).

 

[Oraki]

And every time we've pointed out these types of problems, the "fix" is simply to do some kind of weird recalculation within PowerPlay but keeping the bugs that causes this to remain broken.

What makes this especially funny is that when the inflation bug was "fixed" last year, in the respective patch megathread about said topic many of us @Ed @sallymorganmoore and other CMs, asking whether deflation would be fixed as well. Their response was something like "yes". Anyway... if we look at the Original Post:

Income values will be updated for many control systems so that they are the sum of their own income PLUS all the systems they currently exploit.

I guess "many" did not mean "all" of them. How about we simply do a recalculation of the entire PP bubble and address this issue. There is a difference between "not working as we want" vs "not working as advertised", as already pointed out by Martin. Some people I know also did some recalcs and we are atm finalizing our issue report (is this the third, fourth, or fifth Issue report in 16 months regarding this issue?). Some of your numbers also seem off @Martin Schou; do you work with Horizon population numbers or odyssey ones; because old CS use Horizon numbers still. Also Galnet likes to flip +/-, so some of your numbers are going into the wrong direction (Torval for example is positive)

 

[Martin Schou]

What makes this especially funny is that when the inflation bug was "fixed" last year, in the respective patch megathread about said topic many of us @Ed @sallymorganmoore and other CMs, asking whether deflation would be fixed as well. Their response was something like "yes". Anyway... if we look at the Original Post:



I guess "many" did not mean "all" of them. How about we simply do a recalculation of the entire PP bubble and address this issue. There is a difference between "not working as we want" vs "not working as advertised", as already pointed out by Martin. Some people I know also did some recalcs and we are atm finalizing our issue report (is this the third, fourth, or fifth Issue report in 16 months regarding this issue?). Some of your numbers also seem off @Martin Schou; do you work with Horizon population numbers or odyssey ones; because old CS use Horizon numbers still. Also Galnet likes to flip +/-, so some of your numbers are going into the wrong direction (Torval for example is positive)

I'm using the population numbers that EDDB provides. And I remember having a conversation with Themroc about it, and he pointed out that weird population changes wasn't an Odyssey issue, as he had examples from July 2020 where it happened.

Database Change History - ROSS

ross.eddb.io

| 2020-07-16 08:24:26 | {"government_id":{"value":"Corporate","before":"Democracy","label":"Government"},"allegiance_id":{"value":"Independent","before":"Federation","label":"Allegiance"},"population":{"value":100000,"before":420000,"label":"Population"},"primary_economy_id":{"value":"Terraforming","before":"Agriculture","label":"Economy"}}

| | 2020-07-16 08:42:37 | {"government_id":{"value":"Democracy","before":"Corporate","label":"Government"},"allegiance_id":{"value":"Federation","before":"Independent","label":"Allegiance"},"population":{"value":420000,"before":100000,"label":"Population"},"primary_economy_id":{"value":"Agriculture","before":"Terraforming","label":"Economy"}}

| | 2020-07-16 14:20:47 | {"government_id":{"value":"Corporate","before":"Democracy","label":"Government"},"allegiance_id":{"value":"Independent","before":"Federation","label":"Allegiance"},"population":{"value":100000,"before":420000,"label":"Population"},"primary_economy_id":{"value":"Terraforming","before":"Agriculture","label":"Economy"}}

| | 2020-07-16 14:25:58 | {"government_id":{"value":"Democracy","before":"Corporate","label":"Government"},"allegiance_id":{"value":"Federation","before":"Independent","label":"Allegiance"},"population":{"value":420000,"before":100000,"label":"Population"},"primary_economy_id":{"value":"Agriculture","before":"Terraforming","label":"Economy"}}

And it doesn't just change population. It can be economy and security type as well, multiple times a day

And for those not in the know, the population difference here would change the CC value for the system by 1 from 6 to 7.

 

[Oraki]

I'm using the population numbers that EDDB provides. And I remember having a conversation with Themroc about it, and he pointed out that weird population changes wasn't an Odyssey issue, as he had examples from July 2020 where it happened.

Isnt that a separate issue entirely? If you have a DB dump pre Odyssey and post Ody deployment you will notice lots of changes across the galaxy; new settlements were added that altered the population numbers. That is iirc the main reason why so many powers are now "missing" so much CC from your point of view. It was also mentioned by Foursyth in the original missing CC "patch" discussion thread.

 

[Danieros]

I'm using the population numbers that EDDB provides. And I remember having a conversation with Themroc about it, and he pointed out that weird population changes wasn't an Odyssey issue, as he had examples from July 2020 where it happened.

Database Change History - ROSS

ross.eddb.io

And it doesn't just change population. It can be economy and security type as well, multiple times a day

And for those not in the know, the population difference here would change the CC value for the system by 1 from 6 to 7.

Populations can also be changed as a result of a community goal reward or any other reason for Frontier to add or update assets in a system, and that may be enough to entirely change the economy of the system. In the Sothis case, I could find a mention of three stations being added in 3304 (4 years ago), which explains the jump from 100k to 420k. Then the Odyssey assets bumped it to it's current 440244. Population changes like this, I don't know why, can be seen as "noise" in the EDDB change history even months/years later.