Hotfix date for next week 11/18?

[Crystallus]

I'm just amazed that they didn't realise it was a bug until nearly 3 months after release... don't SE monitor drop rates in game?

Still not adding up to 3 months.

Drop rates being to low were reported the first day someone completed a run RIGHT?

Actually the dev response to the post linked above was on the 14th. So a fix in one week not 3 months.

So what's your point?

[Pellegri]

Still not adding up to 3 months.

Drop rates being to low were reported the first day someone completed a run RIGHT?

Actually the dev response to the post linked above was on the 14th. So a fix in one week not 3 months.

So what's your point?

Except for the fact that turn 1 has been cleared for far more than a single week like your post suggests? The fix is taking a week, the droprates have been fed to Square for over 2 months now, they know when there is an issue and they should not need players to tell them that something is messed up with how drops are handled. Read my post farther up if you still believe that there isn't a "point" to what other people are saying.

[Alcyon_Densetsu]

Coming from a friend who works in a major game studio (not SE): "if you want to make good code, you test. Everything you write, you test. You literally have two machines side by side, one on which you code, the other on which you test, preferably in-engine (once the game's in alpha or more). And it's back and forth and back and forth all the time. And you also proof your code by simulating. To make sure you didn't screw up, you don't rely on re-reading, because that's arrogant; you simulate a thousand or more occurrences very fast to make sure it works. You make your own tools to proof your code, it's not hard to loop something basically. Sometimes I launch a sim before I go to sleep, and check the next morning if out of thousands of occurrences the code did what I want it to do. You can be the best coder in the world, you always make mistakes at some point. And that is how games get screwed up, because a lot of programmers just trust they're good and don't test their stuff all the time. You wouldn't imagine how many games just don't get the maths right. It has to be right, and then only can you fine tune so that it feels right, for the player. And the first player, that's you. If you don't QA your own code, it's bound to be shitty more often that not."

Simply put, for something so trivial as generating a chest from a loot table, you'd run a sim of that piece code, just to see if out of a thousand or more occurrences, you get the expected stats. If we can't trust something so simple, how can we trust all the stats for combat and crafting skills? Is +10% really it, or is it -5% in reality?

And these kinds of things aren't easy to spot afterwards, by a QA team or even thousands of players (PTS wouldn't help much on this), because each and everyone of us only get a fraction of the occurrences, we simply don't have enough data (unless we do QA's work and actually note everything) to conclude if something's wrong. Our vision is biased, for lack of enough samples, unlike the vision of the guy who makes the code in the first place.

Anecdotical: I did a a bit of gathering (mining lv30+). Out of a few thousand ores, I had the distinct perception that I was failing way too often; but I can't tell because I didn't count the nodes. However I could easily compare NQ and HQ and I don't know if I'm that lucky but I can tell you that, mathematically, I got way too many HQ. I usually farm ~600 NQ ores when I do a gathering session, and I never failed to obtain ~30% HQ give or take 5%, when the game says that I have (at most) a base 15% chance, boosted to 25% when I use the skill on ⅓ or ¼ of the nodes, or 20% with the passive stat boost on ⅓ or ¼ of the nodes. Doesn't seem right to me that I end up with 30%.

Just look at this (pertains to this thread's topic), and that. I'd wager we'll find more if we start heavily theorycrafting this game.

[Libby9]

Maybe they did it intentionally to artificially slow down people getting loot!