I wanted to see for myself how bad housing situation really is, so I've tried to make some data analysis. Keep in mind I'm not experienced in it, so take all of this with grain of salt, however, I do believe that data I've collected have high enough accuracy, so if you want to make your own conclusions, go ahead and use that data. But I do not claim that data are 100% accurate.
Way of acquiring data:
Because SE doesn't make it easy to get this kind of data, I've used most popular, player made solution for checking availability of housing, which is Paissa. Paissa's data are crowdsourced, this means that data are never 100% accurate, but it's the best we can do. Those inaccuracies are mainly in form of plots which have not been updated since last roulette cycle. Smaller outliers are those plots which have not been updated in a long time, those seem to be only on Materia DC. However, from all of the problems Materia has, lack of plots is certainly not one of them. Main downside is that you cannot reliably check number of bidders on each plot. However, it's very unrealiable if you want to see specific bids for each plot, but that is not main focus.
There was a total of 3465 permutations to get all data I wanted - 81 servers x 3 plot sizes x 5 districts x 3 tenants modes (individual only, FC only and unrestricted). I was not getting all the individual plot data(I did that in the end, but ended up not using it for anything), instead I used website's filters to get all corrensponding sums of available plots. I have also filtered out all the FCFS plots - those plots are not FCFS, they're (to my knowledge) just new wards 25-30 on new servers, which have not been actived yet. Even though they're technically there, they're still not available to purchase, therefore I think it's fair not to count them. This seems to be case for 4 new EU servers and all servers of Materia and Dynamis.
Because website has funky filter for tenants (you cannot filter out exclusively unrestricted plots), I needed to recalculate values to get FC exlusive, private exclusive and unrestricted plots. First sheet is exactly that, 2nd sheet has corrected values.
Download link for excel file with everything in it
Known issues:
Sheet "General - DCs" - ~E251, ~Q251, ~QE251, ~AQ251. In the tables, Materia (OCE) is under NA category. (Doesn't affect data)
Main table with all the data (you probably want to open it in new window, so you can actually see it):
Main takeaways:
- Housing situation is considerably worse on EU/NA than on JP datacenters
- JP has higher demand for FC housing than EU/NA
- EU/NA has higher demand for Private housing than JP
- If situation on the server is bad enough, people will settle even on undesirable plots like smalls in Goblet
- Popularity of Dynamis' servers is Seraph >>> Halicarnasus > Maduin = Marilith
- It's healthier to add new servers to existing DCs than to create new DC with all new servers if possible
- To match JP's population density, NA would need 15 new servers, and EU 8.
- Vastly different population density on EU, NA and JP servers, unsurprisingly, affect housing availability
- Right now, only realistic chance to get large is in OCE - even Dynamis servers have similar amounts of large plots available as every other server
- Best housing situation (excluding Materia and Dynamis) is on Elemental, worst on Crystal
- JP isn't really using a lot of new unrestricted wards
- District popularity is The Lavender Beds > Mist > Shirogane >>> Empyreum > Goblet
- Most prestigious is large plot in Shirogane, while least prestigious is small plot in The Goblet
- Even though The Lavender Beds is most popular, surprisingly, it's large plots are on 4th spot of popularity in all categories
- Materia's most popular server seems to be Ravana, followed by Sophia
- Dynamis's most popular server seems to be Seraph, followed by Maduin
- Sagittarius is dead
- If you want house and you're on Balmung or Mateus, good luck
- In entire game, there is 729 000 plots, but only 70 888 free plots. Out of these free plots, over a half (38 020) of them are on the new servers
- New servers and wards carry every stat
- Amount of free plots in EU and JP is very similiar - however JP has most of them in new unrestricted wards 25-30, while EU has most in the FC-only plots.
- If we ignore new servers, EU would have nearly 3x less free plots than JP, and NA 2.5x less than JP.
- There is roughly 20-25% less free plots in NA than EU or JP
- Differences in EU/NA and JP's behaviour are significant. This doesn't have to be entirely because of cultural differences, but also because different server populations - JP servers have average of 12 800 players, EU 16 500 (19 900 excluding new servers), NA 22 800 (excluding Dynamis, since we have no data on it's population). With same amount of wards for each server, variabilities in available plots are really not surprising.
Popularity of districts:
Plot availability in datacenter per population:
Conclusion:
Aside from using system that has very bad limits, housing suffers from SE using same brush for all servers and regions. Average JP server has significantly lower population than EU/NA server, and unsurprisingly, this results in EU/NA having less plots. There is different demand for FC plots, EU/NA servers have quite a bit of free FC plots, while usually having close to none private plots. This can make housing seem not bad if you're on JP, new server or very unpopular old server, but situation is certainly bad on most of old servers of EU/NA.
Suggestions:
Instanced housing, if that is not possible, keep adding wards, if that is not possible, reevaluate distribution of FC/private plots. But it might be too late for that, you can realistically only change distribution on brand new wards, unless you want to change current free plots in FC wards to private plots and create absolute chaos.
Important disclaimer:
Without access to number of bidders on each plot at the end of lottery phase, it's hard to determine real demand, which is main limitation of this analysis. I wanted to make much more regarding comparison of EU/NA and JP regions, however, truth is that it is hard to evaluate, because JP simply has much higher demand for FC houses, while EU/NA has higher demand for private housing, making relevant comparisons hard, and merely comparing total number of plots cannot tell us much.
This is my first data analysis, I'm sure there will be some mistakes. I also don't account for wards 25-30 in new servers. Even though they're technically in the game, they're useless since they're not purchasable.
Additionally, I wanted to get some info about demand. I've made different scrapper, which took every single available plot in the game and number of it's entries. Unfortunately, most of plots do not have number of entries, nevertheless, I've tried to get some conclusion from those servers who had 80% or more of the plots with known entries, all of that is in "Demand" sheet. This data only focuses on small plots, since medium and large plots are always in demand and would be kind of pointless.
In case you want this data, it's here: demand.json
This data is from 16 hours before roulette cycle ended, so it's definitely not super accurate, since a lot of people bids in the last moments of cycle.
Another stuff, I ended up not using this. It's just individual free plots in whole game, without number of bid entries: plot_info.json
Update
New stuff with first wave of EU demolitions:
Google sheet with latest data (6.7.2023, after first wave of EU demolitions)
Google sheet with demolition data
Demolished plots:
1.7.2023
6.7.2023, after first wave of EU demolitions
![]()