Goblin Squad Member
|
Cheatle brought this to my attention and Nihimon and I did a small test. If someone has evidence to the contrary please share it.
When you get a recipe drop in a group you will not be able to learn it.
Doing a 'prereq check' is not verification, that check appears to happen regardless if the recipe works or not.
So,
Don't form parties, until it is fixed.
|
Goblin Squad Member
|
This has been the case since alpha 7. I mentioned it in General chat channel a few times.
In alpha 7 at least 75% of my recipes, probably more, that dropped in a party were broken. more like 20% of the ones I got solo were broken.
Note that the "in party" broken recipes were quite random. At one stage three of us in a party got Steel Wire +1 and two copies were broken but the third players copy was OK. Other times all were broken.
In alpha 8 I avoided parties for this very reason but the one time I did party up I got two recipes and both were broken.
|
Goblin Squad Member
|
Is it intended that everyone in the party gets a recipe/spell/etc.? Or does one player get a working original and everyone else gets defective copies?
Coppies are not supposed to happen, the current looting system is a placeholder. The end product is you looting the corpse, which is a timed action that leaves you vulnerable.
|
Goblinworks Game Designer
|
| 3 people marked this as a favorite. |
Yeah, programming figured out what was going on with this on Friday and fixes should be working their way through the system and in game soon.
Essentially, the loot system was assuming that everything that creatures would drop would be in multiples that could be easily split between multiple party members. This was pretty much true back when most loot came from granular Spoils. When we moved away from the Spoils system and had everything have individual drop chances, the system was happily giving fractional (rounded down to 0) items to every party member, which looked like the real thing but didn't actually work. As Rob described it when we finally figured out what's going on, it's like the system saw that the scroll for a spell was supposed to drop, saw that you were in a group of three people, and "helpfully" tore that scroll into three useless chunks so you could each have a piece.
The fixed system should instead divide up the drop chance for every item on the creature's loot table. Rather than succeeding on a 75% roll and getting a drop then splitting it into thirds, each of three party members would instead have an individual 25%* chance of getting the item, and if they're very lucky all three of them could get it. Things that have internal randomization (i.e., a chance to roll on a different table or produce a random number of coin) roll for each player and could come out differently if two players both got the same sub-table drop.
Right now, loot is only distributed on killing blow and split between the killer and his or her party members. The plan is to eventually award kill credit based on threat so multiple people not in a party can share a kill and they're not so vulnerable to kill stealers hitting something that's about to die.
* There's actually a 5% bonus per player past the first to encourage grouping, so a full group of six actually gets 25% more loot than they would solo. Additionally, the loot chance scales based on each individual player's applicable Knowledge skill (linearly up to 50% extra loot at 300 skill total).
|
Goblin Squad Member
|
Just to clarify with an example: if I have 150 applicable knowledge skill and an item has X% chance of dropping then I have a (X+25+5n)% chance where n is the number of party members beyond the first? I assume that no item has above a 25% chance or that would allow going beyond 100% in a group of 6.
Goblinworks Executive Founder
|
Is the knowledge skill actually implemented and working now?
And the fixed system will result in each player having a better chance if they are in a group- the odds of a single player would be 25%, not 75%, and the current system is adding everyone's together and then dividing the loot. (The other option would be to divide the 75% chance by three, leaving three 25% rolls in defiance of the laws of probability.)
|
Goblinworks Game Designer
|
| 4 people marked this as a favorite. |
I don't know if it's in now. It'll be in whenever Paul's changes on Friday afternoon get in.
The formula for the chance to drop any item on the loot table is:
([Base Chance]/[Group Size]) * (95% + [Group Size] x 5%) * (1 + [Relevant Skill Total]/600)
So if you have an item with a 60% drop chance, a group of 6, and everyone has maxed out skills, they'd each have a drop chance equal to:
(60%/6) * (95% + 30%) * (1 + 300/600)
10% * 125% * 150%
18.75%
Which is nearly twice as much loot, on average, than if one person with minimal skill had killed the thing. Each individual person only has a third the base drop chance, but overall the amount of loot given out went up dramatically.
We cap the chances at 100% for anything. So if you think something has a very high drop rate and your high knowledge skill is going to waste by pushing it over 100%, bring a friend.
(Very few things have a drop rate over 66% such that high skill will guarantee capping out the drop chance.)
|
Goblinworks Executive Founder
|
This math means someone solo'ing with little or no supporting knowledge skill would essentially get the base drop change - in this case 60 percent
(60%/1) * (95% + 5%) * (1)
60% * 100% * 100%
I'm sure this can't be right. Any chance we could get an example of the math (or even a comment on the logic) for a low level character who isn't in a party?
|
Goblinworks Executive Founder
|
Hmm. Two 27.5% chances is different from one 50% chance. I'm not sure it's strictly better, either.
But if knowledge skills stack in the group, things get strange. Do knowledge skills stack in group, and is there a cap?
Also, I think that displaying the fraction of a recipe/spell and allowing them to be recombined into a whole one might be a better solution. Perhaps partial recipe drops might find a place in the future.
|
Goblin Squad Member
|
| 1 person marked this as a favorite. |
This issue really needs to be fixed before GW implements anything (like stationary ranged attacks) to strongly discourage solo PVE. If soloing is too hard because you're standing still, and joining a party still means broken loot, then that's just bad for everyone.
|
Goblinworks Game Designer
|
It starts as a higher expected return... But I think there's a point where the odds of getting at least one are lower.
It's pretty much always worse if you're in some weird situation where you only want one of a thing and you only want to kill one guy to get it.
In normal situations where you're going to be killing a lot of things, though, it should average out to be the listed bonus. If you're in a party of three people, after the end of a session you should have 10% more loot than you would have gotten solo (even discounting that being in a group probably also lets you kill faster overall than you could solo).
Because the math's easy for me, assume creatures with just under a 91% drop chance for whatever it is you want (and that looting skill is the same for everyone to generate that 91%). Such that, in a party of three, that 10% bonus takes you up to 100% split three ways, or 33.33% each.
If you look at the raw numbers, that's only a 70.37% chance of getting at least one item between the three of you. That's way less than the 91% chance you'd have solo. But it also doesn't include the 25.93% chance of getting at least two items between the three of you, and the 3.7% chance you'll all get an item.
If you killed 1000 of the creature in a session between you, and you did it individually, you'd get an average of 910 items: 90 of the kills you get nothing, 910 you get one.
But as a group, for 296 of the kills you get nothing, for 444 of the kills you get one item, for 222 of the kills two of you get something, and for 37 of the kills you all get something. Totaled up, as a group you actually got 999 items, the same 10% bonus expected.
Quest rewards are on a different system, and as a game we're not boss-focused, so under most circumstances you're going to be killing multiple creatures and having the law of averages work in your favor, rather than killing a single creature that only has one thing that you want.
Which is why I'm hesitant to create situations where specific creatures are by far the best source of particular loot :) . Any time they don't drop it, you feel like you've lost out.
I like how you are putting thought into this, Stephen. You aren't just running the numbers. You are also running the "emotions"- a little behavioral economics goes a long way :)
|
Goblinworks Game Designer
|
Goblins have different loot than ogres, and, on average, their loot is worse because their difficulty level is lower. But they don't have any inherent loot penalty for being goblins; a level 1 goblin warrior should have pretty similar loot values to a level 1 bandit recruit (and to a level 1 ogre, if the weakest ogre wasn't level 3). And the tables are set up in such a way that if, say, bandits become more desirable to kill because the market has decided the kinds of things they drop are overall more valuable than the kinds of things goblins drop, we can adjust what's on their tables to try to even that out.
The situation I worry about is if we had, for example, a goblin boss that only spawns once per escalation, who had a 50% drop rate for Recipe: Awesome Goblin Hat, and he was basically the only reliable source for that recipe. Now assume your settlement has one crafter that can learn that hat recipe and will make one for anyone who wants one, such that a second copy might be valuable to sell, but not nearly as valuable as one. In that case, as Decius points out, you don't particularly care that in a group the chance for 0 is equal to the chance for everyone to get one, you just care about the fact that the group has lowered your chance to get just one.
We would like to do some things like that, in the sense that there's a cool reward for rare monsters beyond just generic components and a settlement trophy if killing the thing finished the escalation. But we'll probably wind up governing those situations with the quest system (which can ignore the drop chance splitting) or just inflating the drop chance high enough that even in a group everyone gets one.
This sounds a little like yall are thinking about implementing a Settlement-level quest system in addition to Player-level quest system- which would be totally badass if you did.
|
Goblin Squad Member
|
Goblins have different loot than ogres, and, on average, their loot is worse because their difficulty level is lower. But they don't have any inherent loot penalty for being goblins; a level 1 goblin warrior should have pretty similar loot values to a level 1 bandit recruit (and to a level 1 ogre, if the weakest ogre wasn't level 3). And the tables are set up in such a way that if, say, bandits become more desirable to kill because the market has decided the kinds of things they drop are overall more valuable than the kinds of things goblins drop, we can adjust what's on their tables to try to even that out.
I know this is asking a lot, but would you mind publishing those loot tables? Just so we'd get an idea how big loot chances are and what to expect from any particular group. And to satisfy my personal curiosity, as well :-).
|
Goblin Squad Member
|
Oh, btw, if someone's out testing that: is it possible to combine the fragments (by trading all of them on one player) and then have a working recipe? I.e. does the game correctly add three thirds of a recipe together to make one whole recipe?
We tried trading recipes to "fix" them with no success but not sure if we tried that exact combo.
|
Goblin Squad Member
|
Yes, I I tried just generally trading them to fix in the last alpha and it very occasionally seemed to work. Since the nature of the bug has since been explained to be the "splitting up" of singular resources, I figured it might be possible to undo this damage by patching the split-up bits back together. Then again, since the issue is already known and will be fixed, there's really not much need to test it any further. Just curious :-).
|
Goblin Squad Member
|
The situation I worry about is if we had, for example, a goblin boss that only spawns once per escalation, who had a 50% drop rate for Recipe: Awesome Goblin Hat, and he was basically the only reliable source for that recipe. Now assume your settlement has one crafter that can learn that hat recipe and will make one for anyone who wants one, such that a second copy might be valuable to sell, but not nearly as valuable as one. In that case, as Decius points out, you don't particularly care that in a group the chance for 0 is equal to the chance for everyone to get one, you just care about the fact that the group has lowered your chance to get just one.We would like to do some things like that, in the sense that there's a cool reward for rare monsters beyond just generic components and a settlement trophy if killing the thing finished the escalation. But we'll probably wind up governing those situations with the quest system (which can ignore the drop chance splitting) or just inflating the drop chance high enough that even in a group everyone gets one.
You've got it right the way you're doing it with the loot tables and the bonus in groups. I really like the system you're describing and it will be a blast just going out and farming in groups. I would stay away from unique bosses that have a high percentage of dropping specific recipes. At least if you're doing this have the uniques also spawning at a random enough location that people aren't just location farming in a single spot waiting for long periods of time.
...a Settlement-level quest system...Those are escalations :-). Ignore them at your Settlement's peril.
Oh is *that* what those are for? :P
|
Goblin Squad Member
|
Its annoying to fight out of party but to be honest their is no huge advantage to being partied up other than shared drops:
- chat is single focused on either general or party so you need voice coms if you want to keep an eye on general anyway.
- being in a party does not help you track other party members unless they happen to be already on your minimap.
- there is no facility to cast buffs or heals "to the party"
So overall it is no great loss to be out of party, the big disadvantage is losing the shared loot drops and higher loot chance.
|
Goblin Squad Member
|
...it is no great loss to be out of party...
Interestingly enough, I just addressed this in another thread. The other downside is if you're not able to achieve your "share" of the killing blows.
Only you can increment your Achievements while solo, but in a group, don't all get Achievement-credit?
I don't know if it's in now. It'll be in whenever Paul's changes on Friday afternoon get in.
The formula for the chance to drop any item on the loot table is:
([Base Chance]/[Group Size]) * (95% + [Group Size] x 5%) * (1 + [Relevant Skill Total]/600)
So if you have an item with a 60% drop chance, a group of 6, and everyone has maxed out skills, they'd each have a drop chance equal to:
(60%/6) * (95% + 30%) * (1 + 300/600)
10% * 125% * 150%
18.75%
Which is nearly twice as much loot, on average, than if one person with minimal skill had killed the thing. Each individual person only has a third the base drop chance, but overall the amount of loot given out went up dramatically.
We cap the chances at 100% for anything. So if you think something has a very high drop rate and your high knowledge skill is going to waste by pushing it over 100%, bring a friend.
(Very few things have a drop rate over 66% such that high skill will guarantee capping out the drop chance.)
Is this still correct, Stephen?