Yes, and the data showing reverse long term trends between Wormhole and Low is especially fascinating, as they have strong anti-correlation. A possible narrative is that small groups, frustrated with being beaten down in Wormhole space, try their luck in low sec instead. So as small groups move out of wormhole space we see the destruction rate decreasing and as the small groups try their luck in low we see the destruction rate increasing.
I fear once they’re beaten out of Wormhole and Low, theyll just quit or merge into a blob group instead of try to live in high sec. The data is already pretty clear that the “utility provided to the big groups” is BS, as nullsec really didn’t show appreciable long term growth to address this alleged “structure spam” problem by @Brisc_Rubal . But honestly I don’t expect this data to convince him, given how adamant he was about the changes.
Either way, a fantasy of what I would like to do next – because I’m not made of infinite time and energy and I’d rather play Eve than do these analyses – is revamp the “Initiative and Brisc detector” to detect whether these destructions are being done by “big groups”. It is hard to detect if the defenders are “small groups” because of the “holding corp” meta.
I hope they are able to share more precise data than my zKill data scrape and I will be very interested to see what conclusions CCP Games reaches from their data.
Considering literally no reply from anyone. Even CSM I think I need to stock up on some popcorn for that stream. It’s gonna be “fun” to watch what type of narrative they will spin this time.
I like it when people start detailing destruction rates to indicate that there is not a problem, however the issue is not so much that the destruction rate went up and then stabilised, it was more that the structures that were blown up were not replaced.
I don’t know anyone here that is arguing that. Where are you seeing people say that?
So, I want to be really precise with my claims. You cannot claim that without either:
Having the detailed data CCP Games has; or
Making arguments by analyzing subsequent destruction rates
For the latter, it is a rate relation kind of problem, to make an argument. Experience with calculus helps here. Absent the CCP data – which if they share it will trump all of the following, it is still one I am making now with the one-sided data I have at this time. I will basically duplicate the argument here from my earlier post where I initially posted the data, to perhaps better show the limits of the claims.
It goes like this:
Setup:
Ignore the short-term effects of the patch on destruction and examine only the long-term trends, as initially the patch garners excitement and initial fireworks.
There are two related rates: Destruction, and anchoring of new targets. These both affect a “small structure quantity” we’ll call N, and that N affects the first-derivative decrease (destruction) & increase (anchoring of new targets). The more stations available, the more destruction, and the less likely someone will want to plop a duplicate structure nearby. That’s the related rate relationship modeled here (this model is extremely simple – which can be argued against, but that’s perhaps a digression). I have the numbers of the former related rate. I do not have the numbers for the latter related rate.
The patch (by definition) makes the underlying process that we’re measuring more prone to destruction. That is, if there were two copies of the same universe, one pre-patch and one post-patch, the system post-patch has higher destruction rate. This is often annotated as a scalar factor in the derivative. It should be uncontroversial to say that if the patch suceeded, this scalar factor increased because of its changes.
Analysis:
Destruction rate goes down: If the structures blown up were not replaced at its pre-patch rate, N decreases, and you’d expect subsequent destruction rates to decrease, simply because there’s a lesser pool of available structures to destroy. This is despite the scaling factor increase.
Destruction rate goes up: Two possible outcomes:
If the structures blown up are replaced at its pre-patch rate, N stays the same, you’d expect subsequent destruction rates to be elevated due to the nature of the patch because of the scaling factor increase.
If N increases (people are anchoring more structures) then destruction goes up no matter what, patch or no-patch.
Destruction rate stays the same: This can be caused by two outcomes, either:
People are replacing small structures regularly so N stays the same, but the patch failed in increasing the playerbase’s eagerness for destruction, so the scalar factor didn’t change! Or,
People are not replacing small structures regularly, so N decreases, but down proportionally to the playerbase’s eagerness for destruction equivalent to the patch’s scalar factor increase.
So the argument goes:
The only way the data shows structures are replaced, maybe, is if destruction goes up.
Long term:
High sec is flat.
Wormhole space is going down.
Nullsec is going down.
Lowsec is increasing.
Pochven not enough data.
Applying the above “analysis” section to the preceding points:
High sec defenders people are not replacing structures anymore or attackers don’t care about structure bashing post-patch any more than pre-patch
Possible Narrative: small group presence disappearing or patch failed to make structure bashing palatable to attackers
Wormhole space on the whole are not replacing small structures there
Possible Narrative: small group presence disappearing
Nullsec space on the whole are not replacing small structures there
Possible Narrative: large groups don’t care about small structures as much
Lowsec space people are replacing structures or new folks are trying to move in
Possible Narrative: Small groups evicted from WHs are trying to make a life in lowsec instead
Pochven not enough data for analysis.
If you’re a small group, the data available now certainly is not preventing the above narratives from forming. But it is not conclusive either.
I guess it is public structures only, to be published on ESI but gives at least a consistent picture over time.
In your zkill analysis would be interested to look into rate of abandoned structures being destroyed as subs have declined, IE someone quits the game without unanchoring their stuff. Not sure how exactly, perhaps on damage the structure takes. In high sec there are a few groups doing this where I feel they’ve increased activity (although unclear if they’re just more effective now or there are for them to destroy).
I am glad that you are not doing that, my point of view is that Medium structures with this change are no longer fit for purpose.
Observations:
A small group lost 20+ structures in NPC nullsec, most of them after the change was implemented. Most of them medium, I know that only 3 of them were replaced.
Several allied groups in hisec give up on a myriad of medium structures and concentrate on large ones only.
You put a lot of effort into that post, I hope CCP is analysing it on the basis you detailed.
I have not, thanks for sharing! I’ll need more time to give it a thorough look-at.
Sounds like a neat angle to look at. Unfortunately I’m about burnt out of putting any more effort into data collection/analysis, but the code I shared should be a huge jump-start for yourself or any other person wanting to go further or in related directions.
I was informed this did not occur – I didn’t have the opportunity to watch the source stream. The retrospective was postponed, likely indefinitely, by CCP Games.
