So very much agreed. That current system also makes it ridiculously hard to distinguish between bugs and features (because it takes a fairly large sample size to distinguish chance variation from a true effect). I imagine customer service gets a fair number of reports from unlucky players, and so has a hard time tracking trends that are true bugs. So they send back canned responses, which is more like salt in a wound than a helpful reply. Badness all around.

You collect some pretty amazing data on player behavior. How about opening that up to the community and letting us play with it? (An API? Armory?) Is there someone we can talk with about looking at data?

TorStatus tracks the status of each server (light, standard,... full). They track EU servers separately, and you can even check population by time. http://www.torstatus.net/shards/eu (Paige's most-excellent math is translating status numbers into estimated population, which I helped a little on.) That said... sometimes I think its better to look for a guild that is a good fit, and then just check that its on a solid server. A few resources to find guilds across servers: http://www.oldrepublic.net/forum/#guild-recruitment http://www.reddit.com/r/SWTORGuilds/ http://www.forcejunkies.com/forums/forumdisplay.php?10-Guild-Recruitment

+5 scav efficiency / + 5 cybertech crit +5 arch efficiency / +5 synthweaving crit +10 inv efficiency / +2 armstech crit +10 biochem efficiency / +2 diplomacy crit +10 artifice efficiency / +2 treasure crit +10 armor efficiency / +2 underworld crit

Ooooh, just got the chance to check the forums. I think the analysis is brilliant and clear - and a big hit on the general forum! (Thanks for the link!) I had a slightly different approach in mind, where I was going to model the population via the latent values from an ordered probit, using some boosting-based methods to validate those estimates. .... but that probably sounds a bit like Greek to most folks. Oh, and totally emailing you now. I love the Sanctum estimates, btw.

I don't know with certainty, but I've got a reasonable approximation. For a while now, I've been collecting data on server status. When a server changes status, I'll log onto a republic and then an empire character on that server, and /who until I've tallied the whole population. So far, I've collected data on just over 100 transitions. (I'd do more, but this stuff takes ages.) Approximate cutoffs: Light -> Standard: 500 Standard -> Heavy: 1500 Heavy -> Very Heavy: [2250-2500]? Very Heavy -> Full: ~ 3000 Since there are several potential sources of error with this process (how close to the transition I can be, internal rounding errors I don't observe, bugs with the /who system), those numbers are really approximations. I'm also uncertain whether they've changed over time. For a handy graph, etc, see: http://inquisitive-myths.blogspot.com/2012/05/population-numbers-behind-server-status.html

As noted by Heriana, this isn't unlikely. You'd have a 5% chance to crit 1 out of 21 objects. To put it another way, suppose that 100k of us went out to craft a 21 items. Of that group, ~1k wouldn't see a crit, and around 5k would only see one crit. If you assume that folks post bad results to the forums more often than good news, this makes for a fairly large number of "X is broken" posts on the crew skills forums. Unfortunately, accurately capturing the influence of small effects with statistics isn't that easy. Assuming you're comparing a baseline 15% chance to crit to a 20% chance with the +5 crit bonus, after crafting 200 items with the droid, you'd have a ~25% chance of seeing a bonus that was statistically significant. Even worse: you'd have a 7% chance of seeing a crit rate worse than that without the bonus. (It'd take about 1500 items to get a decent sample size... which even with grade 1 items is about 25 hours of continuous crafting. I did that once, and promised myself never again.)

If you take a look at TorStatus, they track the status of a server (light, standard,...,full), which is pretty close to the population overall. You can take a look at servers by region, type, etc. http://www.torstatus.net/ If you're not a fan of TorStatus, two months ago Dulfy did a survey of US PVP and PVE server population: http://dulfy.net/2012/02/24/some-numerical-population-data-for-swtor-servers/ http://dulfy.net/2012/03/01/numerical-swtor-population-data-for-pvprp-pvp-servers/ Regardless of where you look, for US servers, the most populated server is definitely The Fatman.

Since 1.2, I've run 400ish of the abundant grade 6 sliced tech parts missions, and crit on around 18% of them using companions at 10k affection, so well within the reported 20% chance to crit (15% base on yellow +5% chance for 10k affection). http://www.swtor.com/blog/community-qa-april-27th-2012 Within that sample, I've had streaks of bad luck. I've had runs of 20 and 21 missions in a row without getting a crit. It is just luck. Eventually everything reverts to the mean. (Personally, I feel better if I go ritually slaughter some jedi to improve my chances, but I'm fairly certain there isn't a game mechanic to back that up. Alas.)

Yup! The 340 mission discoveries sell best. You can get those from running the Grade 6 slicing lockbox missions. Yes. The current bug does not change the returns from Moderate missions. They were profitable pre-1.2, and remain profitable now. Abundant missions are still profitable, but their returns are reduced on crits due to the bug (though... if it crits, you have a mission discovery...). Both are safe bets on average. I posted a graph of the returns here: http://inquisitive-myths.blogspot.com/2012/04/slicing-lockboxes-post-12.html That said, because slicing is guaranteed long-run income, the returns on lockboxes are small. You'd likely earn more credits crafting, if your server's GTN is active.

Its because abundant missions return green lockboxes on crits, and white otherwise. Rich missions return blue lockboxes on crits, but green otherwise. Because of this, a crit on an abundant mission is a slight loss, while a crit on a rich mission is a huge gain.

Just to end this 'affection doesn't increase critical crafting' myth: It has been confirmed by the devs: http://www.swtor.com/community/showthread.php?p=1896308#edit1896308 Datamined from the game files: http://mmo-mechanics.com/swtor/forums/Thread-Crew-Skills-Compendium?pid=10146#pid10146 And tested extensively in game: http://inquisitive-myths.blogspot.com/2012/03/critical-missions-and-companion.html What Grade of slicing mission was that? I've been testing stuff out and hadn't observed a crit without a schematic yet, but the Grade 6s take ages so I haven't run many yet.

Lockbox missions do not give augment schematics: at least, I didn't observe one after running 400 grade 5/6 lockbox missions post-1.2. Maybe the earlier poster was referring to the cybertech speeder / ship schematics? You'll get 1-2 augment schematics on crits of sliced tech parts missions. (So, more affection or a +slicing crit bonus will yield a better chance of schematics). So far, I've been fairly lucky, and have found schematics on 20 out of 100 Grade 6 Sliced Tech Parts missions run (with companions with 10k affection / 400 slicing). Slicing mission discoveries (like A Test of Intelligence) have a chance to return 1-2 augment schematics.

Just bad luck, I'm afraid. I've crit-crafted one of those items for a guildmate. Downside: I keep humming the Jaws theme when I see him approach.

You have about a 3.5% chance to not research an item with a 20% reverse chance after 15 REs. Ouch. That is some icky luck. (But still possible.) You can also use MrRobot's "super evil frustration calculator" to check this stuff out (my name, not theirs) http://blog.askmrrobot.com/2012/04/swtor-re-luck-caculator/