[2021] Conclusion after a long break

I was summoned by the linking of a post.

While we aren’t ready to pick out specific languages to translate to just yet, Germany is one of our top countries for player base. Last time I looked it was number 2.

The crafting percentage is accurate. I am absolutely certain of this. If you believe it to be wrong, I am very open to looking at evidence showing it is wrong. Evidence that will convince me looks like 100+ contiguous crafts shown in an unedited recording. If you can show me this, I will rewrite the code.

Having said that, the feeling that it is wrong is enough for us to re-evaluate how it works and is displayed. I’ve talked fairly extensively about this on stream but it boils down to that because the highs aren’t very high and the lows feel like you’ve been cheated, we should probably change how it works. So we are working on a replacement system that alters the system just enough to feel much better but doesn’t betray its identity.

2 crafts is not a statistically relevant number of crafts to prove a problem in the probability.

It is possible for you to fail 100 crafts at 90% success chance in a row. The chances of this happening are astronomically tiny.

If you get 2 fails in a row at 90%, the chances of that happening are 1%. So, while it is very unlikely, it is possible.

If you do 10 crafts on an item all in the 85-95% range, all of those feel like they should be sure things and any one of them failing would feel bad. There is about a 65% chance that one of them will fail. If any of them fail then all the success are pointless so you only remember the fail at high success chance.

Now, if you do that repeatedly, all you end up remembering is dozens of failures at high success chances. This is when we hear people say things like “it always fails at high success chance”. They aren’t wrong, most crafting failures happen at high success chance. I know this seems counterintuitive but it’s mathematically correct.

Possibly, we do have some big plans for crafting that will make this problem go away though.

Yea, I can’t really give full details on it just yet but we are doing a ton of automated testing to make sure that the difficulty to get different average tiers from different starting points stays pretty similar to what it is now.

Some scenarios get very slightly easier and some scenarios get very slightly harder.

The goal in the changes is mostly to remove the feels bad that I was describing above. While still keeping the difficulty at a similar place.

We aren’t planning on changing the core functionality of the crafting system.