Skill Tree [work in progress]

[TriMoon]

Let me know if there's interest in this, and I will continue to work on it until it's finished:

I'm very interested in this, so keep going!
Maybe a DHTML page that people can download and use would be better, or just host it on some free web

[kabuto202]

If you shared the .pdf file with me I can help you out with it.

[Droj]

I've just finished making one myself using Googledocs. Rather than make a separate post I thought I'd post it here for reference/usage. There's all a table of Inspirational and smelting data.

Salem Guide

[Thorke]

I've just finished making one myself using Googledocs. Rather than make a separate post I thought I'd post it here for reference/usage. There's all a table of Inspirational and smelting data.

Salem Guide

I hope this information ends up in the wiki(s) as well.

Aside: I wonder why no one has added a column to the smelting data showing the expected value of the number of bars for a given run. I know it is unintuitive to many to talk about .72 bars of iron for a 8/8 run, but it is useful to think of what maximizes the number of bars over a long series of runs (i.e., 72 expected bars over 100 runs). (unless I am missing something it should be (1+lime)*iron/100).

[MagicManICT]

Aside: I wonder why no one has added a column to the smelting data showing the expected value of the number of bars for a given run. I know it is unintuitive to many to talk about .72 bars of iron for a 8/8 run, but it is useful to think of what maximizes the number of bars over a long series of runs (i.e., 72 expected bars over 100 runs). (unless I am missing something it should be (1+lime)*iron/100).

Statistical analysis? You'd have to form a bell curve for each one of the possible inputs and then create a 3D chart showing what peak outcome is. No, I forget the algebra for it, but it's not simply additive. That's why it's only a bit over 50% at the peak for 1+ bars vs something like 70%. Excel and I think the OpenOffice spreadsheet has the formulas for probability calculations built in.

[Thorke]

Statistical analysis? You'd have to form a bell curve for each one of the possible inputs and then create a 3D chart showing what peak outcome is. No, I forget the algebra for it, but it's not simply additive. That's why it's only a bit over 50% at the peak for 1+ bars vs something like 70%. Excel and I think the OpenOffice spreadsheet has the formulas for probability calculations built in.

I think you are making things much more difficult that they really are. If the code is written so that the chance of creating one bar from one ore is independent of the chance of creating one bar from the each other bar, then yes indeed the expected values are additive (for instance, if you flip a coin once you have a 50% chance of getting heads; if you flip 10 coins, you will on average get heads 5 times). Now, if the trials are not independent then neither this analysis nor the one we are using for 1+ bars is valid.

Also, please note that the 50% and 72% numbers are not inconsistant since portions of those 1+ results are for 2, 3, 4... bars and those add up.

[Scalorn]

I was hoping that the pretty skill tree would get finished..

Here is an massive and ugly skill tree done using graphviz based on the data in the wiki:

WARNING: Image is ~7000 x 4000 in dimensions:
http://i.imgur.com/ZQAfN.jpg

[Thorke]

I was hoping that the pretty skill tree would get finished..

Here is an massive and ugly skill tree done using graphviz based on the data in the wiki:

WARNING: Image is ~7000 x 4000 in dimensions:
http://i.imgur.com/ZQAfN.jpg

There is much to be said for massive and ugly, but up-to-date, easily maintained images. Tools are good.

[darnokpl]

There is much to be said for massive and ugly, but up-to-date, easily maintained images. Tools are good.

Fast moves skill allows you to use 'Jump' and 'Bullrun'.
Fencing skill gives you 'El garde!'? and 'Thrust'.

[NoEvil]

I was hoping that the pretty skill tree would get finished..

Here is an massive and ugly skill tree done using graphviz based on the data in the wiki:

WARNING: Image is ~7000 x 4000 in dimensions:
http://i.imgur.com/ZQAfN.jpg

Massive, sure. Ugly? Hardly. It's neat, simple, and easy to follow, Fucking awesome work, thanks!