[Ar-list] Lifting

[iko at chem.ucla.edu]

"Carter T. Butts" <buttsc at uci.edu> said:

> iko at chem.ucla.edu wrote:
> As you are undoubtedly aware (given your email address) the UC
> quarter is now underway....

Actually I’m not at UCLA anymore.  I just use the account… Even after 
graduate school I’m a cheap bastard. 

Are you still in school?

> > [BTW what is the maximum stress and fatigue you can 
> > add at any one time?]

> Something like this would make a good MRS, if someone wanted to write 
> it up.

Hmmm… maybe later.  I have some ideas right now for other things.

> > Also how would one do the test?  Would it simply be a test against > > 
some weight?  Or would it be a STR test where the MS will determine > > the 
weight lifted?

> It should be a contest of STR vs. a function of the weight being 
> lifted, with a positive success margin indicating that the character > is 
able to lift the object to at least a certain extent.  Exactly how > well 
this is performed is determined by the magnitude of the SM...a 
> margin of 1% would suggest that the character is barely able to move > the 
object, while a margin of 100% implies nearly effortless action.

I disagree.  The fact one puts in effort means that he has lifted an object 
effortlessly.  In the action of lifting, all that should really matter is 
the strength/skill of the person and effort put in.  How much he is lifting 
at that time should then be weighed against the weight of the object. 

I guess what I’m trying to say is that the SM is should directly increase or 
decrease the weight lifted.

BTW isn’t a SM of 100% is impossible.

A better analog would be long jump: How far can a person jump?

> > However, infinite skill does not mean one can lift infinite weight.  > > 
Should there be a limit on how high these skills can be?

> Implicitly, there already is -- you'd have to train them, and at a 
> certain point your advancement would no longer keep pace with decay. 
> (Exactly what that point is depends upon your prerequisite levels -- 
> which you would want to train as well -- and upon how intensively you 
> train.)

I talked with a few semi-professional/retired weight lifters.  The general 
consensus is that perfect skill can add (on average) 25% more weight and 
(more importantly) prevent injury. 

In game terms a person with 10 STR and 100 Dead Lift (skill) is not at all 
equal to 100 STR and 10 Dead Lift (skill).

> > Let’s assume that at the attempt the lifter has a 75% success rate > > 
(a rating of 30), that he is putting maximum effort, both his 
> > lifting skill and 
> > strength are near the human maximum (500 each).  So 500 + 500 + 
> > effort – r*weight = 30; where r is difficult per mass.  Would this > > 
work?
>
> Well, that general approach is one way to go about it.  I note that 
> you are treating this contest as having two attributes (STR and the 
> relevant skill) versus one...this is unusual for a skill test (where > 
only the skill is generally involved) but perhaps reasonable for this > 
case.  

I was hoping that someone might jump in with a better way of doing tests 
like these.

After looking at the information at hand, I suggest that the following is a 
better way to handle tests like this:

Roll = percentile roll
Str  = strength
Sk   = skill (could general lifting or jumping or even specific bench press)
Eff  = effort
AMod = penalties for the action (weight if lifting; distance if jumping)
Mod  = other modifiers

((Str + Eff) + (Str + Eff)*(Drf(Sk) – 0.5) + mod) – invDrf(Roll) = AMod

For lifting:

Weight = 50kg * 10^(AMod/200)

Long Jump = 3m * 3^(AMod/200)

High Jump = 1m * 2^(AMod/200)

One of the problems I see is that the distribution of the rolls makes this 
flat, not Gaussian as found in the real world.  To fix this I would suggest 
rolling 11D10 – 10 instead of percentiles.

Thoughts?

> It would be essentially impossible to actually get one's STR rating 
> to 500, however, by normal means.  

So there is a functional human max.  

Not having done the simulation for training/learning, I naively thought when 
one drf’s a score and gets a percentile result that would equate to one’s 
position in the population (200 = 95% = 95th percentile in the population).  
When I really think about I guess that really shouldn’t work that way.