Thoughts on the Item Durability Generator

I've been working on the item durability generator for a while now, mainly to get a working Android version up. I've gotten the preliminaries done and had some time to analyze how the generator is working.

One main problem has crept up once I started using the generator, however: items are breaking too soon.

When I developed the idea, the total resistance of an item was supposed to be its average durability in days: how long it should last on average before breaking. I was finding that items with a durability of 500 days were lasting for 6; items with a durability of 200,000 days were lasting for 500. So I investigated the cause.

What I found was that random variables were, as they often are, the cause. Each time the stress is added, there is also a d100 roll to see if the item breaks. Obviously, given the uniform distribution of die rolls, after about 100 of them (and much before too) I had a good chance of getting below the break chance.

I tested this on a bunch of different simulations: 2000 items of various total resistances. My findings were as follows:

Resistance: 500
Average # of Days Lasted: 13.06
Standard Deviation # of Days Lasted: 6.87
Average Break Chance when Item Broke: 12%
Standard Deviation of Break Chance when Item Broke: 5.33%

Resistance: 2,000
Average # of Days Lasted: 27.84
Standard Deviation # of Days Lasted: 14.03
Average Break Chance when Item Broke: 6%
Standard Deviation of Break Chance when Item Broke: 2.91%

Resistance: 5,000
Average # of Days Lasted: 46.22
Standard Deviation # of Days Lasted: 21.84
Average Break Chance when Item Broke: 4%
Standard Deviation of Break Chance when Item Broke: 1.87%

Resistance: 10,000
Average # of Days Lasted: 70.58
Standard Deviation # of Days Lasted: 31.11
Average Break Chance when Item Broke: 3%
Standard Deviation of Break Chance when Item Broke: 1.34%

Resistance: 50,000
Average # of Days Lasted: 192.79
Standard Deviation # of Days Lasted: 62.91
Average Break Chance when Item Broke: 2%
Standard Deviation of Break Chance when Item Broke: 0.55%

Resistance: 200,000
Average # of Days Lasted: 547.05
Standard Deviation # of Days Lasted: 99.83
Average Break Chance when Item Broke: 1%
Standard Deviation of Break Chance when Item Broke: 0.21%

As you can see, the number of days increases while the break chance decreases. I've been contemplating a few solutions to this problem which I will outline in a future post, but they are as follows:

1. Increase all items' resistance to very large values in order to get an average closer to what I expected.
2. Only begin rolling break chance after a certain point.
3. Reconfigure the system.

Time to weigh options!