“Knowledge is the death of research” – Walther Hermann Nernst, Chemist
12.04.yc118 J163408 < E-C00264 < Region E-R00026
Rather than lead you through the data analysis of Phase II, let’s cut to the chase and reveal the results.
The null hypothesis: Based on region, known wormhole types are randomly connecting to other regions of space within the known expected distribution by type to the destination region using a significance level of 0.05.
Conclusion based on Phase II data: Since the p-values are greater than the significance level of 0.05, we accept the null hypothesis. The observed distribution is from the same population as the expected distribution.
TLDR: Known wormhole connections are equally random.
Project Coordinator: Katia Sae
Project Liaison: Merkato Cesaille
Technical Lead: David Louis
Project Specialist: Ashlar Maidstone, Soul Darkshade
Research Team: Forcha Alendare, Alek Azam, Lucas Ballard, Triffton Ambraelle, enkidu nagata, Sanibel, Lucas Ballard, Mushroom Greene, Theo Fugger, Caleb Wolfram, Akatsuki Hikage, Pileto, Mako Koskanaiken, Earthling Jaer, Vladimir Gengodov, Gorgan Fullsail, Caille Sinclair, Jen Outamon
Read on if you’re interested in some of the details. I’ll specifically target our exceptions from Phase I.
Check out these post for more information about Project “W”, how it came about, and the Phase I results.
Get on with it!
From September to the end of November yc118 (2016), there were a total of 15,305 connections observed. From that data set, I used the known connection types for the analysis which gave me a total of 4,902 connections to analyze. Compared to the 300 connections from Phase I, we increased our sample size by a factor of 16.34. With this data set, we were able to meet the following conditions of the Chi-square Goodness of Fit test as follows:
- Sampling method is simple random sampling. Our observed connections are equally likely to occur in our expected destination population (Regions). Passed.
- Variable under study (connection type) is categorical (Regions). Passed.
- The expected value of the number of sample connections in each level of the variable is at least 5. Passed.
From Phase I, our anomalies concerned High Sec, specifically the Genesis and Molden Heath regions, and Class 5 wormholes, specifically region E-R00024. Let’s compare them.
As you can see, our High Sec observations by region went from a range of 0 to 9 to a range of our minimum of 5 to 120.
Our Class 5 observations by region went from a range of 0 to 7 to a range of 5 to 50.
Comparing our Chi-square ranking, you can see our anomalies normalized with the other regions with the additional data collected.
That’s really all there is to it. The research is a good lesson that conclusions can’t be drawn until all of the conditions of a given test are met. In this case, our Phase I data we didn’t have the minimum of 5 observations for each region of New Eden. Our Phase II data met that requirement and we were able to show that, at least for known wormhole connections, the destination region is equally random.
Where to from here?
I’m going to look at K162 connections and see if there are any abnormalities to be found there. My thought process is this: A K162 connection should randomly connect to anywhere, be it High Sec, Low Sec, Null Sec, or W-Space. The caveat is this, when reviewing the K162 connection, it will give you an indication of what type of space it leads to, just like our known wormhole connections did from our previous two analysis. I’m going to go with the assumption that until you look at the connection, it could lead anywhere. I’ll call this Phase III and use the same data set we just collected. Stay tuned…