Jump to content

Tyche Shepherd

Resident
  • Posts

    1
  • Joined

  • Last visited

Everything posted by Tyche Shepherd

  1. Someone called me to this thread I've been collecting Grid Concurrency Stats every 5 minutes for the past 10 years ( with a couple of breaks of up to 4 weeks) Concurrency is falling, slowly but the trend is a decline - This can be seen with both the Maximum Daily Concurrency and the Median daily concurrency - See the attached chart Overall the Grid has grown in size since 2018 but that growth has mainly been Linden Projects such as Bellisseria, The Lab has added an additional Linden Owned 454 regions in the past 14 months but Private Estates have decline by 89 - From a growth perspective 2019 was more or less flat for private Estates - Most of the net loss has been in the first two months of 2020 The total signup figure is growing at a similar rate over all 10 years - We know its just supposed to be a count of signups to the Second Life Web Page and the figures have little relationship with the concurrency figures which do exhibit a decline over time I have also audited the concurrency figures in the past and believe they are an accurate representation of the number of accounts logged into Second Life at any one time - The total number of Green Dots seen on the Grid Maps are within 1% of the reported concurrency during the audit process (its takes about an hours elapsed time to run the census)
  2. Concurrency as reported on my gridsurvey site is the number of user accounts logged into Second Life at any particular time . The data is provided by Linden Lab via the llsd feed (its one of the few public stats feeds still running). I sample and store the data from the feed at 10 minute intervals. Daily I calculate the Median and Mean values for the past 24 hours and report these on the site. The same feed also provides total signups to date which I also store and report The concurrency figures are an accurate measure of the number of accounts logged in at any one time. I have used my survey bots to confirm that reported concurrency is more or less the same figure as if you counted every green dot on the map . Of course logged in accounts does include bots, idle accounts and people who may be logged in more than once. Nevertheless it's probably the best public measure of SL usage we have. As several have pointed there is little evidence of any increase in the overall decline since Sansar was announced and as with private estate declines which I also regularly report on , the past 12 months or so have shown a general slowing down in the rate of decline than previous years . There is a general decline overall , but it doesn't seem to been exacerbated by more recent Linden Lab announcements I've also copied in my median daily concurrency time series decompostion chart which identifies overall seasonality and and trend over the past 7 years , As you can see , if anything the more recent trend has almost flattened out 
  3. Hi AlexandreLois1 I'm the owner and maintainer of http://www.gridsurvey.com database There is no method in the Gridsurvey API which allows bulk downloading of anything from the main database and there are rate limits in place to deter large number of queries. However on the Home/Region Page of gridsurvey.com the "Current Region List" button allows the user to download a list of current regions along with their SL Map Coordinates in Tab Delimited Text. This is designed to be used externally to SL by developers who maintain their own external databases and unlike the main API cannot be downloaded direct via LSL. This data is based on my regions surveys which are currently only updated weekly so you should only require one or two downloads of this data in any one week if its useful for your project . Good luck with your project
  4. Based on past 12 months to yesterday Saturdays and Sundays have higher concurrency then the rest of the week (this is significant at the 95% confidence level) 13:00 and 14:00 SLT (1 - 2 pm) are the busiest hours - again this is a statistically significant finding at the 95 confidence level These charts illustrate the mean concurrency over day of week and hour of day. This analysis is based on sampling SL concurrency at 5 minute intervals over the past year.
  5. Medhue Simoni wrote: ..... So, the marketplace, in my opinion, doesn't answer why LL got back nearly 1,000 regions, in a matter of months. Plus, the rate of abandoned regions seems to go in steps, which would coincide with search engine updates. Of course, the search dropping parcels from it's engine affected every1. Not just merchants, which is why it shows on the graph so distinctly. The changes in this chart are not due to abandoned land , it shows the size of the Linden Estate in regions rather than the ownership of parcels within those regions.The rate of mainland abandonment will have no impact on this chart as The Lab doesn't take Mainland regions off line, nor does the Estate owner change. The growth in this chart during the highlighted period was due to the roll out of the first Linden home continent, its subsequent expansion then the arrival of a second cloned continent followed by a third & fourth continent. There was also a roll out of a large number of welcome islands at the time, many of which were then taken offline at the end of March which explains the small dip at the end of the period. During the same time there was also a corresponding increase in Premium Accounts, which had been in decline up until then.
  6. The very first Mainland Census I ran back in November 2009 found that abandoned land made up between 3.5% and 5.0% of mainland and this remained fairly consistent through the first part of 2010. However since December 2010 I have been running a quarterly census and the amount of abandoned land has generally increased and last month I reported that it was between 9.8% and 10.9% of all mainland Here is how the amount has changed in the past two years I do monitor the amount of mainland for sale in each Census but currently the method I use doesn't differentiate between Land for Auction or Land for direct sale , they are both classified as Land for sale so I can't say whether there has been an increase in land specifically put aside for Auction or not. However the amount of Total Mainland for Sale has steadly increased each quarter and since June this year the majority for sale has been by the Lab rather than by private owners. This chart illustrates this very clearly, the Jun 2011 census was the first one after the automatic putting on sale of abandoned land was implemented, prior to that, Linden Owned Land for sale was predominantly Auction only . My next Census is due to be run in December and reported on SLU , I'll have a look at seeing if I can differentiate between Auction and Direct Sale Land. Tyche
  7. Currently each Private Estate region generates on average US$215.27 Tier per month (factoring in known grandfathering , distribution of Full vs. Homestead vs. Openspaces and using list prices) . A full region on Privately owned mainland (including Linden Homes) generates an average of US$261.07 Tier per month. However if you include the 44.8% of land which is Linden Owned Mainland as an overhead this drops to US$144.11. So that's Gross income,ignoring any costs such as support, infrastructure. Figures are derived from my Mainland Census at the beginning of June and my Private Estate survey at the end of June. So depending on whether or not you want to include Linden Owned mainland in the equation - it goes either way P.S. I suspect this is my first post in the (now not very new) forums .
×
×
  • Create New...