Yeah, I didn't realize that it did until I looked at the results spreadsheet. There's an somewhat-anonymous kill count sheet that's not obvious, but I'm sure is used for the data. Makes me wonder what happens when we fight in wings and we're all using it - does the same kill get counted multiple times?
Thargoid Attacks: Where, When and What you can do about it
- Thread starter Factabulous
- Start date
Yeah, I didn't realize that it did until I looked at the results spreadsheet. There's an somewhat-anonymous kill count sheet that's not obvious, but I'm sure is used for the data. Makes me wonder what happens when we fight in wings and we're all using it - does the same kill get counted multiple times?
Oraon has been confirmed as the third target this week - Aegis transmitter was found in vicinity of station - OP updated
So now we have three targets, should be fine as long as no-one is distracted by the beta
So now we have three targets, should be fine as long as no-one is distracted by the beta
German Cmdr PetesGun stopped at Scout after killing it and took a photo
Maybe never done before but when Cmdr PetesGun had killed a Scout he made a stop and took a shot of the remnants - and discovered that the scout possesses intestines! Well, living beings, either...
Maybe never done before but when Cmdr PetesGun had killed a Scout he made a stop and took a shot of the remnants - and discovered that the scout possesses intestines! Well, living beings, either...
That just Thargoid cables. Finally found something with worse cabling than Faulcon DeLacy ships
What are the stars on this thread meaning? Is Factabulous a 5 star varmint now?
There's a thread in the admin section saying they've leaked out of the JWE forum
I think they were testing them and some admins can rate threads, so looks like one of the passed by - though I'd rather they had pinned it as I requested Edit: And nothing to do with me - I neither can rate, or have asked for a rating on this thread [up]
Ive been doing the same thing these last few days, examining the dead hulks of scouts. Ive been wondering if they are drones as there does'nt look like there's enough room for a living being inside.
Took a picture of the inside of a cyclops some time ago for much the same reason. No clues I could discern.
German Cmdr Mr.K did the same, you can watch them here! Well, that doesn't look like a mere machine at all. Stunning!
As far as we are aware at this point we have killed just short of 6500 scouts and 50+ interceptors at Oraon so far. The intention is to get over 7000 scouts and as many biggies as posssible. If this is not enough to save this system then we really need a clue as to what is good enough because there are a lot of CMDR’s investing a lot of time into this.
Just collecting your own group's kill count isn't going to be enough to determine this - you need to be able to estimate how many kills are being done by other pilots as well. Provided your group is contributing a reasonably high fraction of the kills - say 20% or more - this should be fairly practical.
Here's how I would go about it:
1) Ask your pilots to report, along with their kill count (K), the number of times they hyperspaced into the target system during the week (A). (If they're using an EDMC plugin to submit kills, it may be easiest to modify it to also submit jumps, and then they don't have to do any extra work)
2) Count the total number of FSDJump events recorded for the system on EDDN during the week (B). I collect this information anyway so ask me if you need any of them.
3) At 24 hour intervals throughout the week, read the Traffic Report from the station logs (C). Traffic reports update on the hour and cover the previous 24 hours, so you want to do this at the same time each day and *ideally* as close to 7am as possible.
(For A, B and C you want to keep the figures for separate days as well as for the entire week - this will be useful later)
This gives you a ratio A:C to estimate the proportion of system traffic that is contributed by your pilots. (The ratios A:B and B:C may be useful for more complex statistical modelling later, so record B for later but we don't need it yet)
You can now estimate the total number of kills (T) as
T = K + (K*(C-A)/(A*P))
where 'P' is an estimate of how much better your pilots are than the average visitor to the system. (i.e. if your typical pilot kills 30 a week but you think the average pilot not in your group would only get 10, P=30/10=3). You might be able to get some initial estimates for P by looking at the variation in kills among your own group. Based on what I've seen at CGs ... I'd start with something like 2.5 and go from there.
You'll find over the course of a few weeks that C/A is different in different systems so your estimate of T might vary quite a bit for similar starting K.
You can then mark "wins" and "losses" against the various T scores for each systems. If the initial guess at P was correct, almost all the wins should be on one side of a line, and almost all the losses on the other. If they're more jumbled up than that, then try increasing or decreasing P, then recalculating T, and seeing if that helps. Eventually you should be able to get a value for P which splits things either side of the line fairly well ... and now you have a good estimate for what T actually is (bearing in mind that Frontier like round numbers, so if you calculate 18,943 kills, it's probably 20,000 and your estimate for P is slightly off or inconsistent times for traffic data collection have meant your C values are a bit wrong, or similar [1])
Now comes the useful bit: you can use this information to predict how many kills your group personally needs to get in a system at the start of the week, given that other people are going to get some too. This number will vary from system to system and week to week - as you're seeing! - so that's what you actually want to know for motivating people and setting targets.
Let's say that you find that 10% of the traffic happens on Thursdays, that P is 3, the target kill count is 15,000, and the traffic report on day 1 was 50 ships of which 15 were yours, who got 1,000 Thargoids between them.
We can then predict the total kills for the day were
1000 + (1000*(50-15)/(15*3)) = about 1750
and the total kills for the week will be 1750 * 10 = 17500 which is probably going to be enough. You can guess that about 7500 of those kills will come from the other players, so you can set a kill target for the system of 7500 as well (of which you already have 1000, of course) ... or maybe a little higher, for safety.
[1] This is one reason to collect B as well - some systems might be popular for other reasons (e.g. the attacks on Kamadhenu or Deciat probably had massively skewed A/C ratios because people go to those systems a lot anyway) - so you might want to start trying to estimate from B what C might have been the week before, and trying to filter out traffic which has nothing to do with the Thargoids. But this gets complicated and it might not be necessary most of the time.
What about the other 2 systems? Have we killed enough?
Heh - I used to do this - traffic reports were actually predictive where the EDMC plugin never was. People don't believe that though, because it's an estimate
It's nice that you're trying but I'd give up - this lot have their metrics, so it must be FD switching things, and it's going to change in a few weeks when Q4 drops anyway..
Not sure, both the Hive and AXI have forused on Oraon as far as I know. The system belongs to a player faction and they had asked for help. The Hive had over 6k scout kills in a system last week and we still lost it so we are not really sure if 7k will be enough.
Fingers crossed.
Fingers crossed.
That seems remarkably dismissive...Is...
Solid Assumption on no evidence A. (FDev haven't changed any of this single system kill totals) Plus Fact 1. Combined AXi/Hive Wing Kill counts have steadily increased (which are obviously only a sub-set of total kill counts) Leading to assumption X. Un-recorded kill counts are gradually going down...
Any MORE or LESS Likely to be true than...
Solid Assumption on no Evidence B. ("random" unrecorded scout kills are likely to broadly similar week on week) plus Fact 1. Combined AXi/Hive Wing Kill counts have steadily increased (which are obviously only a sub-set of total kill counts) Leading to assumption Y. FDev ARE changing the single system kill totals
In the absence of any evidence to the contrary?
Well, hence my suggestion on how data to determine the difference between those cases might be collected. It wouldn't be unreasonable for Frontier to revise the thresholds from time to time so that they remain relevant to the current player base - some CG thresholds have been lowered relatively recently so that it fits better with the reduced participation there - but without a decent statistical model it's very hard to tell.
One thing which does occur to me as a slightly ironic third possibility:
- AXI/Hive have been successfully recruiting and encouraging people to use their tools
- therefore while the total numbers of kills haven't changed much, the proportion of kills which is recorded has increased
- therefore the thresholds seem to be increasing.
The inclusion of a third system target each week will mean that even if unrecorded kill counts stayed roughly constant across the galaxy, the count per system would be expected to be on average only 2/3 of previously. Depending on what proportion they were that would be difficult to determine.
(I'm wondering if use of the Aegis Squadron Leaderboard when that goes live will also allow for a useful partial data sample for verification and validation - certainly there are some interesting possibilities there)
Targets this week : Ross 129, HR 3499 and LHS 6187
Dulerce & Oraon attacked this week
Eagle Eye targets this week:
Systems are saved by kills in the target systems- each kill counts the same, so a Scout is as useful as an Interceptor. Bonds do not matter.
Locations of the Aegis megaships Acropolis and Vanguard.
* NHSS will be found in systems within 10 Ly of target, the systems are not being targeted for attack - you can kill Thargoids from them, but it won't help save any systems.
Dulerce & Oraon attacked this week
Eagle Eye targets this week:
Eagle Eye this week:
EE4 mel 22 sector ue-g b11-0 a2
* Ross 129 Yang Hub
* Aries Dark Region db-x d1-64 - Thargoid Structure TS100
EE5 mel 22 sector ut-r c4-4 1a
* Synuefai UA-Y c17-4 - Thargoid Structure TS163
* HR 3499 - Goddard Hub
EE6 mel 22 sector wk-o c6-1 5c
* HIP 20644 - Thargoid Structure TS83
* LHS 6187 - Hooke Gateway
Three Thargoid Structures and three Targets this week
EE4 mel 22 sector ue-g b11-0 a2
* Ross 129 Yang Hub
* Aries Dark Region db-x d1-64 - Thargoid Structure TS100
EE5 mel 22 sector ut-r c4-4 1a
* Synuefai UA-Y c17-4 - Thargoid Structure TS163
* HR 3499 - Goddard Hub
EE6 mel 22 sector wk-o c6-1 5c
* HIP 20644 - Thargoid Structure TS83
* LHS 6187 - Hooke Gateway
Three Thargoid Structures and three Targets this week
Systems are saved by kills in the target systems- each kill counts the same, so a Scout is as useful as an Interceptor. Bonds do not matter.
Locations of the Aegis megaships Acropolis and Vanguard.
* NHSS will be found in systems within 10 Ly of target, the systems are not being targeted for attack - you can kill Thargoids from them, but it won't help save any systems.
Good news, that at least one system could be saved. I would have hated it to see the lovely trees in Thraskias tourism station burning. Thanks for everyone involved at fighting back!
An Alliance Joint Navy Operation (about 15 cmdrs) were active there and killed more than 2300 scouts (we kept track of the kills), I wonder how you get the numbers for your nice pie-chart? Are these kills included?
Bad news, that another Alliance system is investigated this week, Ross 129
Checking based on weekly EDDN traffic levels for the last few months it seems a general picture is
<150 ships: lose
>250 ships: win
150-250 ships: uncertain
This week, about 200 in Thraskias and 100 in the other two - not far off the Lab 69 ratios for Thraskias, but implies more even on Oraon/Dulerce. Not unexpected with different sampling methods.
The range doesn't seem to have changed with the transition between 2 and 3 targets, so it probably just shows the substantial per-system variations in EDDN/non-EDDN ratios ... and actually smaller variation than I was expecting there.
Last edited: