Talk:Quality

This table is outdated. Game code of 1.0 now clamps quality at 5 (masterwork) unless the pawn crafter has an "inspiration" which adds 2 grades to the crafted item. So crafting legendaries without inspiration isn't possible in 1.0, with inspiration pawn needs to craft excellent or masterwork tier for it to be internally promoted to legendary.—Preceding unsigned comment added by ‎Evzhy (talk • contribs) 21:45, 7 January 2019 (UTC)

I know the table is outdated, but unless someone good at datamining or experimenting manages to grab the values, we don't have the values for the new table so we can't make a new one.

Also, this talk page is broken, and you will need to manually sign by adding ~~~~ after your message; if you consent I will refresh this page by deleting it. PigeonGuru, Strategist ( talk | contribs | assist ) 14:06, 7 January 2019 (UTC)

No, we don't delete comments, learn to use Template:Unsigned.

Yoshida Keiji(Let's talk)

21:56, 7 January 2019 (UTC)

I meant to say that unsigned comments aren't the only reason- the page is not updated and hence needs to be refreshed. But yeah, thanks for telling me that there is that template. PigeonGuru, Strategist ( talk | contribs | assist ) 22:41, 7 January 2019 (UTC)

Ok done [1] —Preceding unsigned comment added by ‎Evzhy (talk • contribs) 01:22, 9 January 2019 (UTC)

I'm new to the game, but how come noone here knew about a debug screen "Quality Generation Data" that shows all chances? —Preceding unsigned comment added by ‎Evzhy (talk • contribs) 13:18, 9 January 2019 (UTC)

So there is such a screen in dev mode. Maybe when I have the time I'll update the table. Alternatively, you can also help update it. PigeonGuru, Strategist ( talk | contribs | assist ) 14:02, 9 January 2019 (UTC)

2021, Version 1.3.3117

A long time passed since this discussion. Are the current values still correct? I tried to check the "Quality generation data" in 1.3.3117 but I couldn't access it. In 1.2.3062 I got the following:

It's not only rounded to two significant figures, it also differs in a lot of places by more than rounding errors. I suppose this in game table doesn't represent the actual chance but rather the percentage of items generated by the game. Can someone verify this? What is the actual formula? Ickputzdirwech (talk) 15:35, 3 October 2021 (UTC)

I think the table that the game generated for you, and also the table on the Quality page right now, were found through simulating the creation of many items and counting the results. Since there's inevitably some variation to doing that, the results will be slightly off unless. I poked around in the source code to find the actual algorithm for determining quality, and tried simulating it myself: https://github.com/Cheldra/rimworld_stuff/blob/master/quality.py. I think it's either really hard, or else mathematically impossible, to find the probabilities analytically. The results with 10,000 simulations per level were about the same:

0 [64.83, 30.07, 4.91, 0.2, 0.0, 0.0]

1 [43.31, 43.67, 12.12, 0.89, 0.02, 0.0]

2 [20.25, 53.23, 23.47, 2.96, 0.09, 0.0]

3 [9.1, 50.71, 33.41, 6.5, 0.27, 0.0]

4 [4.78, 45.17, 39.52, 9.94, 0.58, 0.0]

5 [2.24, 34.93, 47.05, 14.58, 1.18, 0.01]

6 [0.93, 24.25, 52.16, 20.41, 2.23, 0.02]

7 [0.42, 15.45, 53.47, 26.68, 3.9, 0.07]

8 [0.13, 9.11, 50.67, 33.62, 6.32, 0.15]

9 [0.06, 5.64, 46.85, 38.18, 9.02, 0.26]

10 [0.02, 3.27, 40.52, 43.66, 12.05, 0.47]

11 [0.01, 1.8, 32.13, 49.23, 16.08, 0.75]

12 [0.0, 1.0, 24.74, 52.52, 20.54, 1.21]

13 [0.0, 0.62, 19.91, 53.72, 24.18, 1.57]

14 [0.0, 0.4, 15.68, 54.41, 27.31, 2.2]

15 [0.0, 0.21, 12.17, 53.62, 31.24, 2.77]

16 [0.0, 0.12, 9.24, 52.46, 34.43, 3.75]

17 [0.0, 0.08, 6.89, 50.43, 37.93, 4.66]

18 [0.0, 0.03, 5.04, 47.43, 41.3, 6.19]

19 [0.0, 0.02, 3.57, 42.57, 46.24, 7.6]

20 [0.0, 0.01, 2.44, 37.56, 50.4, 9.58]

Cheldra (talk) 23:40, 3 October 2021 (UTC)