Shrinking cloaks, belts and maybe other holding equipment

[spicemustflow]

Why do cloaks and belts change their holding capacity every time I assess them or log in? It's annoying, I buy a knife belt to hold my twenty throwing knives (like Machete) and the next time I'm on, I can't fit a toothpick in it. Anybody noticed that?

[Synthesis]

Your assess command is not exact.

I seriously doubt your containers are shrinking...it's simply that the value that 'assess -v' returns is random, within a certain range.

[brytta.leofa]

I wonder if the coders have made the mean value time-dependent, to prevent the dreaded Averaging Twink.

[spicemustflow]

What is the dreaded averaging twink?

[Archbaron]

What is the dreaded averaging twink?

[spicemustflow]

Anyway, could I be so wrong in my assessment that the number of stones I get is completely random every time I look? I wouldn't mind that, but that number reflects the carrying capacity definitely. When I buy a big belt I want it to stay big...

And if by the dreaded averaging twink you mean the person who looks at the store item several times, waiting for the number to be the highest, well...   

[BreakBad]

The actual carrying capactiy of the belt won't change, AFAIK. It's just your character's (inaccurate) perception of the item that is changing.

[brytta.leofa]

And if by the dreaded averaging twink you mean the person who looks at the store item several times, waiting for the number to be the highest, well...   

Not highest.  One would expect the "incorrect" value (what you see) to be derived from some continuous statistical distribution whose mean is the true value.  Given reasonable assumptions that I can't name off the top of my head, the average of multiple measurements is as good a way to estimate that true value as any, and I do mean "as good a way" in a very strict and approving sense.

An easy way to break this estimate would be to give the measurement error a slowly-changing, time-dependent bias, such that you can't average out the error in a short time.

[spicemustflow]

Heh, and it never occurred to me that my estimate of how much can I put in a pocket might be wrong, so I was the dreaded maximizing twink...

[Yam]

Never have tried to stick a beer bottle in a jeans pocket then, have you?

[Angela Christine]

And if by the dreaded averaging twink you mean the person who looks at the store item several times, waiting for the number to be the highest, well...   

Not highest.  One would expect the "incorrect" value (what you see) to be derived from some continuous statistical distribution whose mean is the true value.  Given reasonable assumptions that I can't name off the top of my head, the average of multiple measurements is as good a way to estimate that true value as any, and I do mean "as good a way" in a very strict and approving sense.

An easy way to break this estimate would be to give the measurement error a slowly-changing, time-dependent bias, such that you can't average out the error in a short time.

That isn't a twink, that is realism.  If you examine the item for a moment, then your perception is likely to be less accurate than if you spend several moments examining it.  Using Assess three times and taking the average is simply spending more of your precious time to get a more accurate estimate.  Perfectly logical.

At least I hope so, because I always assess 3 times.   

[spicemustflow]

Never have tried to stick a beer bottle in a jeans pocket then, have you?

Never, that's what asses is for. At least I thought so.

[Yam]

I really hope you're not saying that asses are for sticking beer bottles into.

[brytta.leofa]

An easy way to break this estimate would be to give the measurement error a slowly-changing, time-dependent bias, such that you can't average out the error in a short time.

That isn't a twink, that is realism.  If you examine the item for a moment, then your perception is likely to be less accurate than if you spend several moments examining it.  Using Assess three times and taking the average is simply spending more of your precious time to get a more accurate estimate.  Perfectly logical.

Well, I was picturing more like 50 times...at that point you can probably plot 'em and see the shape of the distribution.

Blast, now I'm getting an odd urge to actually do this.

[Synthesis]

An easy way to break this estimate would be to give the measurement error a slowly-changing, time-dependent bias, such that you can't average out the error in a short time.

That isn't a twink, that is realism.  If you examine the item for a moment, then your perception is likely to be less accurate than if you spend several moments examining it.  Using Assess three times and taking the average is simply spending more of your precious time to get a more accurate estimate.  Perfectly logical.

Well, I was picturing more like 50 times...at that point you can probably plot 'em and see the shape of the distribution.

Blast, now I'm getting an odd urge to actually do this.

Done it.  Not just with assess.  I had reams of data, until my frickin' hard drive crashed.

[Angela Christine]

I'm not sure if that is twinkery, but it is extreme nerdity. 

[spicemustflow]

I really hope you're not saying that asses are for sticking beer bottles into.

No. An embarrassing typo.


edit: but the while thing with assess not being accurate is just plain annoying. Ok, if the error margin were 1 or 2 stones, but 7 or more? I will send a log and demand to get the tailor subguild skills after I'm done equipping myself.