That’s the issue. Your perception of the activity may be different than the one of people who actually are targets of that activity.
I mean, you can’t say “I am not attacked so there is no attack”.
1 Like
I didnt say I was
But unless you know what I fly and how, you cannot say my view as someone who doesnt fly PvP ships regularly isnt relevent or valid, or whatever it is you seem to be trying to say.
If everything is relative, then nothing can be measured in granularity enough to be relevent and so all conversation is invalid.
So lets go with the definition of Rampant;
Ok for a start, its not unwelcome by the enactors of the act, so to say its rampant for everyone would be inaccurate, but the unwlecome part isnt binding, so we can skip that.
Flourishing or spreading. Ok, is it? My experience suggests it is not. Thats not proof that it isnt, but I certainly have no evidence to bring that is is spreading, my losses arent up, and I dont see any more wreck fields that have pricay tells (Destoryers, one other class and the destroyers all blue wrecks) than I did in the past. But thats anecdotal, so we can skip that too.
Unchecked. Well I guess that depends. CONCORD is one kind of check, as is AG, as is some other factors. But I have no data on thoise either.
So without data from either side, I guess we just dont know.
well try to understand what I’m saying before answering it ?
Your experience that ganking is not rampant does not invalidate the experience of other people which made them consider it is.
It does not. Your personal experience does not suggest anything outside of your personal activity. You can’t say anything general from anecdotal experiences, besides “this thing happened”.
Isnt that what I just said in my previous reply?
Youve managed to pull the sentences where you disagree, not the ones where I prove myself wrong in those sentences vOv
You should probably make it clear what youd like to hear, rather than just quoting what you dont.
Then we might get this marriage back on track.
Also, the plural of Lego is Lego, it never was and never will be “Legos”.
Thats not relevent but had to be said by someone.
you proved yourself wrong when claiming your personal experience suggests something general.
AKA the absence of the evidence is not the evidence of the absence.
I agree and pre-empted that in my previous comments.
Like I said, let me know what you want to hear, Ill get the lawyer on the phone, get the divorce cancelled, we can be having dinner before you know it.
2 Likes
What I want to hear is why people believe it is rampant, which I disagree with but could be wrong.
NOT the nitpicks and cherry picking from people who can’t accept to be corrected on an internet forum and as such will ignore what the people who disagree with them are actually meaning.
I also asked for something earlier about data gathering but some people are more interested in trolling than in actual conversing.
1 Like
Unfortunately that’s always going to happen when using online forums, you’ll just need to pick and choose who you engage with 
ok so now just looking at the kills registered by zkb from the police groups (sentry guns, police drones, concord drones) ; we assume each kill made by such an entity is the effect of a criminal activity.
Of course the issue is first that we have not all data from zkb : the killer or the victim must have registered the kill, if the killer is the entity it becomes more difficult when the victim did not register.
Then two entities(Caldari Police 3rd Lieutenant | Ship | zKillboard and CONCORD Police Captain | Ship | zKillboard) have at least more than 2k mails on a month, which is the limit that zkb can provide. I can try to avoid the issue with the finalblow-only/ but then I can only fetch like 60% of the kms and still have issues with those two entities.
| date | kills | systems | 10th | 100th |
|---|---|---|---|---|
| 2020-10 | 14364 | 595 | 260 | 20 |
| 2020-11 | 13709 | 568 | 187 | 22 |
| 2020-12 | 13321 | 582 | 246 | 19 |
| 2021-01 | 13048 | 576 | 340 | 16 |
| 2021-02 | 11351 | 576 | 234 | 17 |
| 2021-03 | 14304 | 601 | 286 | 24 |
| 2021-04 | 10213 | 471 | 258 | 14 |
from february (the lowest amount of kills) we can say that at least in average, a system in HS has witnessed 10.5 criminal activities ; and 52% (568/1080) of the systems have at least witnessed a criminal activity in the month.
edit : added the 10th highest system amount of criminal activities per month, and the 100th, which correspond ± to the 1st and 10th centiles
2 Likes
This is a big assumption, and a flawed premise.
this is a useless repeat of what I just wrote.
Please go take lessons into discussing.
Define useless
Please do not use undefined words
1 Like
And with average kill/system/month and the giny
| date | kills | systems | 10th | 100th | avg | giny |
|---|---|---|---|---|---|---|
| 2020-06 | 12484 | 616 | 217 | 15 | 11.559259 | 0.9037669 |
| 2020-07 | 12500 | 591 | 267 | 17 | 11.574074 | 0.9085706 |
| 2020-08 | 13253 | 588 | 253 | 14 | 12.2712965 | 0.91855216 |
| 2020-09 | 11582 | 563 | 243 | 16 | 10.724074 | 0.90287286 |
| 2020-10 | 14364 | 595 | 260 | 20 | 13.3 | 0.9157118 |
| 2020-11 | 13709 | 568 | 187 | 22 | 12.693519 | 0.9081377 |
| 2020-12 | 13321 | 582 | 246 | 19 | 12.334259 | 0.9141136 |
| 2021-01 | 13048 | 576 | 340 | 16 | 12.081482 | 0.9120549 |
| 2021-02 | 11351 | 576 | 234 | 17 | 10.510185 | 0.90047896 |
| 2021-03 | 14304 | 601 | 286 | 24 | 13.244445 | 0.90139276 |
| 2021-04 | 10214 | 472 | 258 | 14 | 9.457407 | 0.92733973 |
I try to take the values for 10years instead of one but zkb takes ages to answer.
1 Like
The Punisher is bait.
1 Like
The database that underpins zkillboard was developed in 2013 and they don’t have accurate data that goes back 10 years. Anything prior to mid-2013 is very innaccurate. They are missing lots of data from back then.
There was a comparison of the kills in Battleclinic and zkillboard done a few years ago that I kind find right now; and even accounting for a percentage of kills in Battleclinic being fake, manually posted killmails, there was a lot of kills missing from zkb. Changes in the structure of killmails, the addition of a dedicated killmail server, changes in the API, etc. all contribute to the older data being very much on the low side, to the point of being useless for data analysis.
If you mean November, then the percentage would be 568/1212, or 47%.
Regardless, the only real conclusion from that data is that the number of police group kills is relatively stable, not growing. There are a lot of errors in the data, but if we assume that the sources of the error are reasonably constant, then it still shows that piracy isn’t rampant in highsec.
No matter which way the available data has been cut so far, everything is suggesting that the claim of rampant piracy can’t really be validated.
it’s not about the accuracy, but about the sheer amount of time. It takes 2min to fetch a month of kills for only those NPCs from zkb.
I take all the kills from kzb which takes 1H for 2.5years and get the actual mail from ESI.
But yeah I should stick to 2018+. Already 3years.
I assumed the numbers of HS systems were 1080. Actually it’s 1192. So the values I gave are wrong especially the giny, which measures the concentration.
In every case I take the minimum value since I am talking about a minimum. All we have are minimum values as the zkb can’t be assumed to hold all the interesting values.
No, it’s not.
The only conclusion is that the criminal activities are concentrated, with a giny above 90% (0% = perfectly spread, 100% = all in one system).
The number of kills was not the data I was trying to get, only the indication of spread was important, and what’s more the evolution of the spread. You can’t use data that was gathered for an experiment, in another experiment - that’s a stretch.
Indeed though, we can’t find in those data any indication that the number of criminal activity is increasing, neither could we deduce a trend from the rigorous evaluation of “concentration” of criminal activity, therefore this corroborates our first idea, that is that criminality is not rampant in Eve HS, even though more people went from NS to HS.
Possible reasons why people are led to believe it’s rampant are :
- the location of criminality has switched from locations they were not previously, to locations they are playing in. If criminals are moving in HS, then it’s expected to happen, and what’s more people who were used to criminality but are no more used to it won’t come and complain about it while people newly target of that migration will : this will mechanically lead in an over representation of the criminal act increase. (cf participation bias)
- the activity that are victims of criminality switched and led to more people being victims, example by switching from 20 account for freighters to 5 accounts for barges means that the victims are 4 times as many.
- the rate of criminal actions recorded by zkb is decreasing, leading mechanically to the observed criminality rate being constant while the effective criminality rate is increasing.
- the data that is not present in zkb follows a pattern that misleads us (basically it’s not homogeneous). This is totally possible but in the absence of evidence and the lack of action to remediate this issue, this is not a possibility that should be considered. Therefore we assume the rate of registering of criminal activities is homogeneous both in time and in activities.
- They don’t know what “rampant” means.
1 Like
Yeah, whatever has flipped by trigs, etc. has reduced it from the original number.
I remembered it was 1090 last time I checked and thought 10 were removed. Turns out it was 1190 and I already had taken removed into account.
So yes there is a small issue, in that there are 10 missing before the invasions. Therefore the data should be considered prior to invasion, and posterior to invasion : no comparison can be made between data of the two periods.
Yeah that’s a good point. So it would be back to 1212 HS sytems, but the error is only a couple of percent on your numbers, so doesn’t really change what you’ve put.
And yeah:
This is true (I have no clue what “giny” means so didn’t quote beyond that bit).
As much as I salute your quest for and presentation of data to prove a point, I am sorry that it seems I was the one who led you on this quest to prove or disprove a completely false premise. I tried to explain before and it seems I failed so let me try again by first making a vital point.
I AM AMERICAN. Moreover, I come from a certain part of America. Even before I was 10 years old I started globalizing my English, with a first step of eliminating my local accent from my speech patterns. But despite even now living abroad with more British speakers around me than Americans, there are STILL fine points I learn all the time. TO ME, and I don’t know how many other Americans, the word rampant can and often does mean simply “a lot” and can have NOTHING to do with any increase whatsover.
That said, lets say some shipping lane on Earth suffered from pirate attacks, and some country’s navy vowed to protect a portion of that shipping lane by patroling it, and declared that portion to be high security for all their patrolling. Yet, every month, 10 pirate attacks happened, with 5 resulting in the loss of the ships of the targets of those attacks.
Three questions: 1) Would you say that is a lot of piracy for a place that was deemed “high security”? 2) Would YOU feel secure when traveling through that area? 3) If there was less piracy OUTSIDE of the “high security” area, would you think that would be a decent metric for saying there is A LOT of piracy in the high security area?