This article is written from an in-character/roleplay perspective. If you wish to contribute to this page, please refrain from making out-of-character (OOC) references.
|Channel||"AJ: MD" & "Project Snapshot"|
|Contributors||Elisa Fir, Bairfhionn Isu|
Drafted by Faulx.
Project Snapshot is an endeavor to understand the nature of periodic wormholes in known space. In order to accomplish this goal, the project seeks to conduct a near-simultaneous scan of all wormholes in high security space. The data gathered from this "snapshot" of all high sec wormholes would then be analyzed to determine if there are any patterns in the wormhole formation and to measure an accurate count of all periodic wormholes in high security space.
Numerous observations gathered by Project Compass researchers, Brawyn78, capsuleer crowd sourcing, and Elisa Fir indicate that periodic wormholes will only form within certain sectors of space. These findings are summed up into preliminary charts here. Of particular interest to Project Snapshot are wormholes A641, R051, V283, and M555, which have been observed to only form in high sec (B041 is also expected to be found in high sec by the event). Furthermore, the data from the "snapshot" could be used to statistically model counts for wormholes O128, X702, R943, and Z971 which seem to form across high, low, and null sec. If the first snapshot is successful, a second experiment may be conducted in low security space, and, ideally, at some point in the future the results should be verified with a repeat experiment in high sec. For the time being though, Project Snapshot will focus on one enormous scan of high security space. The information learned about the nature of wormholes will then be folded into Project Compass and be made freely available to anyone interested.
Given that there are 1212 high security systems in known space, this undertaking is beyond the reach of any one individual. For that reason, Project Snapshot has reached out to Eve University and the Arek'Jaalan community at large for assistance. The project hopes to enlist the support of a 100-200 strong fleet of scanning ships. Each scanner would be responsible for scanning down and recording the statistics of wormholes in 10 or more systems in high security space within a coordinated period of a few hours. If sufficient numbers of scanning personnel are available, every system should be scanned independently by at least two individuals. This will allow cross verification of the data and also give a second sweep of each system a chance to catch wormholes which may have formed or moved within the "snapshot" period. If the numbers fall short, there will likely be only a limited verification process, consisting of a few systems in each group.
The proposed time for the start of the event is March 4th, YC 114 at 18:00 system time. As of Feb. 19 YC 114, it is expected that E-UNI will hold at least two classes in scanning in the week before the event as a lead up to the Snapshot.
- Contact Eve University and discuss the possibility of organizing an event (COMPLETE). Project was received with some enthusiasm. Thanks E-UNI.
- Finalize organizational details of the event (COMPLETE).
- Set a data and time for the event (COMPLETE). The event took place from 18:00-23:00 March 4th Eve Standard Time.
- Coordinate Event in concert with staff from Eve University (COMPLETE).
- Organize, Analyze, and Publish raw and processed data gathered during the event (COMPLETE).
Information to be Recorded by Scans
Organizational scale aside, the individual data collection is probably the most important part this project. Every wormhole in each system should be scanned down, warped to, opened in the NeoCom's "Show Info" window, and its information recorded. Recorded information should include the following (ideally in the same format):
- Time of scan (time left on wormhole) *
- wormhole Name (scan signature id) **
- System wormhole was scanned in -> System wormhole goes to ***
* "a day", "less than a day", "end of life" time left would be used to find wormholes likely to collapse during the snapshot
** scan sig id will be used to help cross validate results, this information should be in the scanner window
*** "unknown", "dangerous", "deadly", "high", "low", "null" mostly important in case of a K162, all of which should be visible in the NeoCom without traversing the wormhole
(( Please note: that copy/pasting text containing '<' and '>' symbols in the evemailer may cause text to be deleted, so "less than" should be spelled out. ))
for example (( OOC Note: For coordination's sake dates should be in system time, don't worry about the Yoiul Calendar )):
- 08:52 11/20/2011 (a day)
- wormhole M555 (NVK-773)
- Arlulf -> dangerous
With the exception of wormhole A641, participants will not be expected to traverse the wormholes, since this could prove dangerous and reshipping would cost valuable time. Wormhole A641, however, goes from high sec to high sec, and as such, it is of interest to the Project that it's K162 be located. In the case an A641 is found, the participant should traverse the wormhole, record its location, and launch probes to find its scan signature id (again, scan signature id will be used to cross validate data). The information should look like this example:
- 01:52 11/25/2011 (less than a day)
- wormhole A641 (FPH-058)
- Malukker -> Hebisa
- K162 (FUL-121)
The experiment will help to test several theories. One is that when wormholes collapse they immediately reappear somewhere else in the space in which they are able to form. Another is that there are a set number of each type of wormhole. A third is that certain periodic wormholes form exclusively in specific areas. A fourth is that the K162 will only form when a wormhole is traversed (or perhaps even merely observed... this may be some kind of wormhole related "Quantum Uncertainty Principle" or a causal effect of ships scanners pushing the other end of the wormhole into existence). Finally, the scan data may be useful in determining if each type of wormhole has an even more limited sector of space that they may form in than "high sec" (for example... perhaps M555 only forms in Heimatar and Metropolis... or perhaps B041 only appears in systems that start with an 'A', who knows?).
- The first theory (wormholes immediately reappear) may be verified if a wormhole at its "end of life" near the beginning of the Snapshot in one system is found to have collapsed by the second scanner to scan that system (this is another reason why verification scans are important) and is then also found to have appeared in a different system (again by a verification scan). Sadly, a "double exposure" such as this is only likely to be captured in the data if all systems are scanned twice during the snap shot, which will likely require 200+ participants. Timestamps and scan signature IDs are also important for testing this theory.
- The second theory (wormholes have set numbers) can only really be verified by two separate snapshot events. Nevertheless, the first snapshot will give us a good sense of how many wormholes of each type might be in space at any given time and will lay the groundwork for any follow on experiment.
- The third theory (periodic wormholes appear only in specific "sectors"... e.g. "high sec") will be verified if only the expected types of wormholes (A641, R051, V283, B041, M555, O128, X702, R943, Z971, and, of course, K162) are found by all participants.
- The fourth theory (K162s do not spawn unless original wormhole is traversed or observed... i.e. warped to), like the first, will only be likely to be verified by a second pass through each system but also relies on participants traversing A641s. As each participant traverses a A641 (or warps to each wormhole to record its statistics), they will theoretically create a K162 which (if done early enough in the snapshot) may not appear on the first scan of the system but which will be present on the verification scan. Thus if a K162 appears on verification scan or did not appear on the first scan of the system (despite wormhole lifetime indicating that it should) then the K162 must have formed after traversal (or initial observation). Timestamps and scan signature IDs are critical here as well.
- The final theory (periodic wormholes might appear only in certain k-space regions or have other strange location limits) may not really be verified at all since such knowledge would require a constant stream of snapshot events. Nevertheless a huge pool of data will be made available for analysis. And some suggestive patterns may become apparent. And, if we get very very lucky, perhaps something unexpected will show up.
Any questions in regard to this project should be directed to Faulx via evemail. Eve University contact include Bairfhionn Isu. Furthermore, Elisa Fir is making inquiries on behalf of the project among the explorer community.
Project Snapshot fell short of its goal of simultaneously scanning all of high security space. The event had 51 participants who, on the whole, scanned 598 different systems in high sec (full listing here). This represents about 49.4% of the 1212 total high sec systems. There was good distribution of scanned systems, including several "high sec islands". The raw data for the project is available for any who wish to see it. Despite the incompleteness of the data, a few theories were put to the test and some of the underling mechanics behind periodic wormholes in high sec and beyond were revealed.
Of the theories Project Snapshot was to test, insight could only be gained in a few:
- Theory 1 (wormholes immediately reappear) [could not be discerned]: Because of the Snapshot did not cover all of high sec and because there were not enough participants for significant numbers of verification scans, it was not possible to definitively test theory. A Snapshot with 200+ pilots would be more definitive.
- Theory 2 (wormholes have set numbers) [second snapshot required]: This theory requires another Snapshot of high sec to be disproved. Nevertheless this Snapshot has laid the groundwork for the next.
- Theory 3 (periodic wormholes appear only in specific "sectors"... e.g. "high sec") [supported]: Only wormholes K162, R943, X702, Z971, A641, M555, R051, and V283 were discovered during the event. These were all in the set of wormholes expected to form in high sec. Because of the lack of completeness of this Snapshot event, this conclusion is less definitive than the project designer would have liked, but still quite strongly suggests that certain types of wormholes form exclusively in specific sectors of space. It is noteworthy that B041 and O128 were expected but were not found.
- Theory 4 (K162s do not form unless original wormhole is traversed or observed... i.e. warped to) [possibly supported]: The snapshot has found only one instance of a K162 being found before its opposite end.
- Craige Lordon's data
- 19:56 03/04/2012 (less than a day)
- wormhole K162 (RUK-654)
- Nein -> Vilinnon
- A641 (MSH-928)
- Bairfhionn Isu's data:
- 20:36 03/04/2012 (less than a day)
- wormhole A641 (MSH-928)
- Vilinnon -> Nein
- K162 (RUK-654)
- While it is possible that the A641 in question was found by another scanner before the snapshot began, Vilinnon is a very remote and unpopulated area of space, thus this evidence strongly suggests that K162s are present and traversable before their counterpart wormhole has been warped to. Unfortunately, there are no other instances of this contained in the data for this Snapshot, mostly due to lack of completeness and lack of personnel to perform verification scans.
- Edit: 19.41 Mar 7, YC 114: Seamus Donohue reports behavior consistent with Theory 4 during his experiences in scouting; he suspects the Vilinnon A641 was indeed previously located. A fuller account can be found here. This is a very difficult theory to prove or disprove, since it requires fore-knowledge of wormhole destinations and/or trust of accounts from scanners at your destination both of which are hard (or perhaps even impossible) to come by. Nevertheless, observation of a consistent pattern of behavior is compelling support for the theory.
- Edit: 23:58 Mar 11, YC 114: A conversation with Qvar Dar'Zanar, who makes a habit of finding the occupancy status of wormholes, has revealed a pattern in both our private data sets showing that the abnormally low number of K162 found by the snapshot may be an effect of theory 4. It seems that (in our combined private data sets) out of about 75 class 1 systems found which have an exit to high sec, all but 2 of those found via K162 were occupied (these two were not scanned in enough detail to be certain of occupancy), and unoccupied systems were found only via incoming wormholes (i.e. not K162s). Given that there are 215 systems with statics to high sec out of 348 class 1 systems (and that in class 1 systems there is only one static, and in class 1 systems only static wormholes lead to k-space), this data shows extreme statistical bias, since if K162s were always present, one would expect to find several unoccupied systems via a K162. Since none at all were found out of so many data points, this supports the theory that the K162s which would otherwise lead to class 1 space are not present because the system other other side is unoccupied (no one is there to "observe" the K162 into existence). This means that during the Snapshot, wormholes from unoccupied systems that would otherwise have K162s leading to high sec, did not generate a K162. This explains the suspiciously low K162 count found by the project.
- Theory 5 (periodic wormholes might appear only in certain k-space regions or have other strange location limits) [could not be discerned]: No easily detected patterns were discernible. The distribution of wormholes and number of wormholes able to spawn in a given system are consistent with systems being randomly selected at the time of wormhole formation. Nevertheless, patterns may exist and the raw data is available for further analysis by others.
Assuming that there are a set number of each type of wormhole, the total counts of each of kind of wormhole may be gleaned from the data. This then, can be used by explorers to get a sense of how often they might expect to find each wormhole, and thus, how often they might expect to be able to reach a particular destination.
Following are the numbers of unique signatures found by the snapshot sorted by wormhole type:
*note the odds of finding 0 wormholes of a given type on this dataset, given that there are at least 5 in existence, are less than 5% (((1-.494)^5) = ~3.32%), suggesting that there are less than 5 of these wormholes. This also suggests that as a group the number of B041s, B449s, and O128s combined is likely less than 5.
While they do not form in high sec, D792 were identified as K162s leading to dangerous or deadly space. This is because "dangerous" space includes only class 4 and class 5 w-space systems, and "deadly" space includes only class 6 systems. Since no known wormholes from class 4 reach high sec and since D792s have been observed to only form in class 5 and class 6 w-space systems, D792 was the only K162 which could possibly lead to "dangerous" unknown space. No K162s were spotted leading to "deadly" space. Sadly there is some skew to this data since a small number of K162s' destinations were not properly recorded.
By a similar logic, B449s would have been found as K162s leading to null sec. This is because B449s have only been observed to form in null sec. However, none were found, suggesting their numbers to be less than five. Again, there is some skew here as small number of K162s' destinations were not properly recorded.
B041s and O128s were and still are expected to form in high sec, however none were found by the Snapshot, suggesting their numbers to be less than five.
With 598 out of 1212 systems scanned, it can be expected that about 49.4% of each type of wormhole was found. By multiplying by the reciprocal we can estimate the full numbers of each type (with decreasing statistical certainty for those with fewer numbers):
1212 / 598= 2.0267558528428093645484949832776
Thus we can project estimated total counts for each type of wormhole (in high sec):
|Wormhole||found in||leads to||estimated
|R943||high sec||class 2||95||these numbers for high sec, (more in null and possibly low)|
|X702||high sec||class 3||81||these numbers for high sec, (more in low and null)|
|Z971||high sec||class 1||67||these numbers for high sec, (more in null and possibly low)|
|A641||high sec||high sec||51|
|M555||high sec||class 5||36|
|R051||high sec||low sec||18|
|V283||high sec||null sec||8|
|D792||class 6||high sec||6||inaccurate count, see #Continued_Analysis|
|B041||high sec||class 6||1-4|
|B449||null sec||high sec||1-4||inaccurate count, see #Continued_Analysis|
|O128||high sec||class 4||1-4|
**note that there should be at least 654 K162s in high sec from static wormholes departing class 1-3 alone (as well as an unknown number from periodic wormholes from null sec, low sec, c5, & c6), so this estimate is suspiciously low, suggesting that other estimates may need to be revised upward.
- Edit: 23:58 Mar 11, YC 114: Alternatively, this low K162 count may be due to the effects of Theory 4, which states that a wormhole will not generate a K162 until the moment it has been observed or traversed (see note under Project_Snapshot#Theory_Results, Theory 4).
In addition to the totals above, the Snapshot allowed the opportunity to find and measure the base scan strength of two wormholes types which had not previously been measured. This data can be used to discern what a signature may be without having to scan it to above 75% signal strength:
X702, 1/20 base scan strength, measured in high sec by Faulx
R051, 1/20 base scan strength, measured by Faulx
The data on wormhole mechanics learned from Project Snapshot has been added to Project Compass's wormhole charts detailing known information on all kinds of wormholes. Several wormhole types continue to remain mostly theoretical and anyone sighting wormholes C248, K329, C391, N290, B520, S047, B041, U319, L614, M609, or O883 are encouraged to contact Faulx or Elisa Fir so that they may be properly documented.
Upon further analysis of the data, I find I must revisit my conclusions on Theory 4. The theory states that a wormhole's K162 will not form until the wormhole has been observed (i.e. warped to) or possibly traversed. Initially scant evidence was found which suggested this was not the case. However, a review of accounts (such as this one) by capsuleers like Seamus Donohue, describing how static wormholes would consistently lie unused by those on the other side until warped to and traversed, seemed convincing (if anecdotal) evidence that the static's K162 was not able to be located by those on the other side (assumedly becuase it had not yet formed). Moreover, Qvar Dar'Zanar pointed out that unoccupied class 1 w-space systems seem only to be found via an incoming wormhole in w-space (i.e. never through a K162 in high sec).
These facts led me to question the reason for the "suspiciously low" number of K162s found by the Snapshot, so I made a simple statistical analysis of the data. Using a list of of w-space systems constructed using data from Brawyn78's datacore, combined with some of the data from Project Snapshot, I estimated the lower limit for the number of K162s that one should expect to appear in all of high security space at 682 (assuming that K162s are always present, scannable, and traversable).
w-space statics to high sec:
- N110 (215 systems) [class 1 to high sec] w-space list
- B274 (334 systems) [class 2 to high sec] w-space list
- D845 (105 systems) [class 3 to high sec] w-space list
- A641 (at least 25 wormholes) [high sec to high sec] snapshot data
- D792 (at least 3 wormholes) [class 5 & 6 to high sec] snapshot data
- B449 (as few as 0 wormholes) [null sec to high sec] snapshot data
Total: at least 682 K162s should be found in high sec
Given, that there are at least 682 K162s in the 1212 systems in high sec, and that 598 different systems in high sec were scanned by Project Snapshot and assuming that the location where these K162s appear in high sec is completely random. The odds of finding a single K162 within those systems scanned is 598/1212=49.34%. So using a binomial distribution with n=682, p=.49334, and q=(1-p)=.5066, the expected value for the distribution is E=n*p=336.5 and the standard deviation is sigma=sqrt(n*p*q)=13.06. Thus in 68 out of 100 snapshot events we should expect to find 337 K162s give or take 13 wormholes (i.e. between 324-350). Since we found only 251 unique K162s, our snapshot seems to be 6.548 standard deviations below the expected value. The odds of being more than 6 standard deviations away from the expected value are worse than 0.0000001973%, which means that 1 in more than 506,797,345 snapshot events that scan 598 systems in high sec should produce only 251 K162s. This is statistically compelling evidence that one of the assumptions made is incorrect (or that my math is wrong... but lets hope that's not it).
The only unverified assumptions made were that K162s locations are random and that K162s always exist. Therefore, we must assume that one or both of these assumptions is wrong (or that the Snapshot scanning data is wrong... but let's not go there). In lieu of any obvious pattern in K162 locations and in light of the anecdotal evidence about K162s, I'm inclined to reject the latter. Thus, this lack of K162 is support for Theory 4. It seems that a great number of wormholes which should have lead to high sec remained "unobserved" during the snapshot event, meaning fewer K162s were formed, and consequently fewer K162s were found by the event participants.
This has some consequences for the project's periodic wormhole counts for wormholes D792 and B449, since these numbers depended on counting K162s in high sec that lead to "dangerous", "deadly", or "null sec" areas. Since the K162 would not have formed unless observed from the other side, an unknown number of "unobserved" wormholes would have not been found by the project. Therefore the counts for D792 and B449 are completely inaccurate and should be dismissed.