Jeweler probability and socket color data

I believe a witch with below 50 hours played can farm 1 player chaos already, over time it's accessible to every casual gamer.

At this point getting enough to afford average amount of jewellers for 6 sockets takes about 3 hours worth of runs. (trading your other orbs drops for jewellers)

Seems about a good balancing of effort required to me.

It's not like you will have tons of endgame items besides the one to repeat this process on anyway; so you can make this a long-term goal.

A little simulator I created, lets you simulate (1000 times, or just once) how many jeweler's Orbs it will take to hit your goal (usually 6, but the option to change is yours).

You can find it here:
http://thehammer.x10.mx/PoEJewelers.htm

This simulator takes into account how many sockets you you have and will not let you roll that same number on the following turn. For example, if you start with one socket, after your first roll you cannot have one socket. Whatever number it lands on next, it cannot be that number of sockets on the following roll, etc., until it hits the maximum number*.

*My program assumes the devs put a mechanic in to where if you hit a 4, 5 or 6 on a 3 socket max item (one-handed weapons), it will still reward you with three rather than forcing a re-roll into a valid number, if that is the case, a 3 socket is the most likely thing to roll as it would have 80 chances for 3 sockets + 30 for 4 + 5 for 5 + 1 for 6, which gives it 116 chances in 306. I'm not horribly concerned if this is not the case as most of us would use it to guage the amount of jeweler's orbs we would need to get a 6 socket item.

My point earlier was that I feel that not all game content should be accessible by all people. People who take the game less seriously, or are not as good don't deserve to get to end game content where you can get 6 link equipment and godly rares.

Keep in mind I am talking about the people that just want to go out and kill stuff and haphazardly spend points in their passive tree. If these people get access to everything, then the people who spend the time to plan a build, fight with strategy, and are efficient with their resources are not rewarded. If Dude A is better than Dude B and gets the shaft because everything is chance that is mess up. But if Dude A gets to the next zone because he is better then Dude B and then has the opportunity to play more content or get better drops, that seems fair to me.

People which do prepare their build(s), are better skilled in playing or even have more time to invest, will always end up with stronger builds, with faster/higher rankings in the leagues, and with a high chance, better equipment too. (by the more amount of high level "grinding")

I see no problem, to get a potential access to every content also to casual players; I would be more afraid of parting the community by such a "feature".

If "Dude B" play such a casual way, not to take care about anything, he will stuck in some place, long before "Dude A" gets into troubles.

"

Mr_Cee wrote:

If "Dude B" play such a casual way, not to take care about anything, he will stuck in some place, long before "Dude A" gets into troubles.

That is my point. With the game as it is, that point will be the Maelstrom of Chaos. Not everyone should be able to farm there.

At least it should "cost" some effort to make + play a viable char, ack.

I've updated my tool, there was an error in the code which gave the "1 socket" option an extra chance (101 vs 100), so I fixed that. I also added a table that shows the percentages of each socket count in the current run.

In a 1000 run simulation, they should very closely mirror the percentages in this spreadsheet which breaks down the likely-hood of each socket possibility and shows the actual percentages of hitting the various socket counts.

The TL;DR version of my tool and spreadsheet:

The average amount of jeweler's orbs it would take to hit 6 sockets would be ~224.

.

The expected number of steps before reaching the final number is actually not too difficult to calculate exactly.

Using the 0.9.6f numbers from Chris, the expected number of jewelers required to get at least X sockets on an item that can get up to 6 is:

3: 1.8
4: 6.0
5: 37.0
6: 223.0

It doesn't matter how many sockets you begin with (1, 2, 3, etc.).* On items with lower numbers of sockets, this will be the same as using TehHammer/Milhausen's assumption that the devs will let you effectively roll 5 sockets on a 4 socket item.

Now, if your item can have at most X sockets, and if we don't use TehHammer/Milhausen's assumption and instead assume you can only roll at most as many sockets are on the item, the expected number of jewelers before getting the max sockets is:

3: 2.2
4: 7.0
5: 44.1
6: 223.0

Again,** it basically doesn't matter how many sockets you begin with.

Keep in mind this is just the expected number and it can take many more or less (the standard deviation is very close to the expectation).


*Trying to get 5 will take on average 37.1 steps if you start at 4 rather than only 37.0. The other cases do not differ in the final digit.
**Again, in the case of 5, this varies from ~44.0 to ~44.2. The other cases do not differ in the final digit.

I wish I had read this before wasting 20 jewelers attempting to get 6 socket.... These probability needs to be way upped...