<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<META content="MSHTML 6.00.2800.1106" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV> </DIV>
<DIV style="FONT: 10pt arial">----- Original Message ----- 
<DIV style="BACKGROUND: #e4e4e4; font-color: black"><B>From:</B> <A 
title=kyle3054@iprimus.com.au href="mailto:kyle3054@iprimus.com.au">Kyle 
Schuant</A> </DIV>
<DIV><B>To:</B> <A title=lev_lafayette@yahoo.com.au 
href="mailto:lev_lafayette@yahoo.com.au">Lev Lafayette</A> </DIV>
<DIV><B>Sent:</B> Monday, October 17, 2005 11:51 AM</DIV>
<DIV><B>Subject:</B> relative strength of randomness and skill</DIV></DIV>
<DIV><BR></DIV>
<DIV 
style="BACKGROUND: #e4e4e4; FONT: 10pt arial; font-color: black"><B>From:</B> <A 
title=lev_lafayette@yahoo.com.au href="mailto:lev_lafayette@yahoo.com.au">Lev 
Lafayette</A> </DIV><FONT face=Arial size=2><STRONG></STRONG>
<DIV style="FONT: 10pt arial"><BR>OK, the idea is pretty simple. The basic 
principle is<BR>the degree of randomness varies according to the<BR>action 
performed.<BR>...<BR>Depending on the type of activity or their 
relative<BR>importance of the incident to the story, different<BR>skills will 
have a different influence of randomness<BR>in determining the Trait Effect. 
</DIV>
<DIV style="FONT: 10pt arial">****</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">OH NO, NARRATIVISM!</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">Okay, so that's, "if it's important to the story, 
randomness will be important, too. If it's not important to the story, not so 
much." </DIV>
<DIV style="FONT: 10pt arial"></FONT>    That's what we miight 
call a "Narrative" approach. Or "storytelling" or whatever. I'd call it a 
"dramatic" approach. Problem with these words is that they all come from 
movies... and in movies, we have the opposite effect. The more important a task 
is, the LESS likely a PC is to do really well or badly. The less important the 
task, the more random the result! So Jackie Chan may fumble making a cup of tea, 
or may make the most delicious dinner ever. But he will almost never fail when 
sliding down a large ribbon through the middle of a multi-storey mall, and 
landing on a foe and knocking him out. </DIV>
<DIV style="FONT: 10pt arial">    In rpg design, we keep aiming 
for the opposite of movies, whereas players keep hoping for action movies... I 
think this is perhaps a reason for trouble in many game groups!</DIV>
<DIV style="FONT: 10pt arial">    In rpg design, we shie away 
from the idea of making ridiculous tasks easier than simple tasks, because we 
feel that players would do ridiculous tasks all the time, then, and the game 
would become... ridiculous. Many systems solve this by having a basic system of, 
the more difficult the task, the greater the degree or need for randomness (ie, 
"you need a good roll to succeed"), and balance it for the "movie" feel with 
some system of Hero Points, allowing you to buy successes or rerolls. </DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">HELP, EVEN WORSE, GAMISM!</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">    Then there's the approach of 
simulating reality. My view is that as you become more skilled in an area, 
random chance plays a smaller part. So for example there's the 100m sprint. I 
have no skill in it, I may run it in 14.50 sec today and 12.50 sec tomorrow. My 
performance will vary by +/-1 second, or +/-1 7%, roughly. But your Carl Lewis 
will run it at 10.01 sec today, and 9.92 sec tomorrow. His performance varies by 
0.09 seconds, or +/-0.05%. His higher skill doesn't just make him faster, it 
makes his performances more consistent, too. </DIV>
<DIV style="FONT: 10pt arial">    In most games, however, the 
randomness of the dice rolls is very great compared to the skill/attributes 
involved, so that in say 10% of contests, I could outrun Carl Lewis. </DIV>
<DIV style="FONT: 10pt arial">    So that's why the dice number 
range compared to the attribute/skill number range is very important. For 
example, in Rolemaster, the number range for dice is 100 (roll d100 and add to 
to skill+attribute bonus), with a 5% chance of going below that, and a 5% chance 
of going above that. It takes up to about level 10 for most characters to have 
skills of around +100. So there's several months of play (of weekly sessions) 
before the skill number will match the random number in importance. A skill +40 
guy has a decent chance against a skill +70 guy, because the random roll is on 
the order of 100.</DIV>
<DIV style="FONT: 10pt arial">    Whereas in Runequest 2, which 
was a roll-under system (attacker roll to hit, defender roll to parry; attack 
and parry were separate skills but let's set that aside for the 
moment) a skill of 70% was FAR superior to one of 40%. The higher 
skilled guy would roll to hit (70% chance) and the lower-skilled guy roll to 
parry (40% chance), for an overall (70% chance to hit x 60% chance of parry 
failing) 42% chance of the higher-skilled guy striking for a wound; whereas the 
lower-skilled guy has just as (40% chance to hit x 30% chance of parry failing) 
12% chance of striking for a wound. </DIV>
<DIV style="FONT: 10pt arial">    In Masterbook (skill+attribute, 
adjusted by d10-d10, compared to difficulty number, highest wins), a 
skill+attribute of 7 is quite superior to one of 4, because the d10-d10 roll 
will make the performance cluster around the skill+attribute level. If you had a 
roll-under system (d10 roll, vs 0 to 10 skill+attribute), then the attacker with 
skill 7 vs the defender with skill 4, it'd be just like RQ2 as above. But with 
the skill+attribute being altered by d10-d10, you get a different result. The 
results are their chances are, -3 (7%), -2 (8%),  -1 (9%), 0 (10%), +1 (9%) 
and so on. So the chance of the 4 beating the 7 is, success if they get +3 or 
more. The chance of +3 or more on d10-d10 is 28%. How to illustrate it, 
um....</DIV>
<DIV style="FONT: 10pt arial">    Let's look at it as a ladder. 
Imagine there are possible ranges of results, 1 to 10, skill plus a d10-d10 
roll, and the results are capped at either end. So if you have two guys of 
skill 4, the two ladders look like,</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">10   10    (10%, 
coming from 4% chance of 10 alone, plus 3% for the prohibited score of 11, 2% 
for 12, 1% for 13)</DIV>
<DIV style="FONT: 10pt arial">  9   
9      (5%)</DIV>
<DIV style="FONT: 10pt arial">  8   
8      (6%)</DIV>
<DIV style="FONT: 10pt arial">  7   
7      (7%)</DIV>
<DIV style="FONT: 10pt arial">  6   
6      (8%)</DIV>
<DIV style="FONT: 10pt arial">  5   
5      (9%)</DIV>
<DIV style="FONT: 10pt arial">(4)   (4)    (10%)</DIV>
<DIV style="FONT: 10pt arial">  3   
3      (9%)</DIV>
<DIV style="FONT: 10pt arial">  2   
2      (8%)</DIV>
<DIV style="FONT: 10pt arial">  1   
1      (28%, coming from 7% chance of 1 alone, plus 6% 
chance of the prohibited score of 0, 5% for -1, etc)</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">Each has equal chances of success. The ladders are 
matched in probability, the chances of each result for each guy noted. . Their 
performances will hover around 4, with a 10% chance of its being 4, a 10+9+9 = 
28% chance of its being 3, 4 or 5, a 10+9+9+8+8 = 44% chance of its being 2, 3, 
4, 5, or 6, and so on. Now consider one guy of skill 4 vs another of skill 7. 
The comparative ladders look like,</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">10   10    (10% for 
the score (4) guy, 28% for the score (7) guy, since results are 
capped at 10)</DIV>
<DIV 
style="FONT: 10pt arial"> 9    10    (5% 
for the score (4) guy)</DIV>
<DIV 
style="FONT: 10pt arial"> 8    10    (6% 
for the score (4) guy)</DIV>
<DIV style="FONT: 10pt arial"> 7    
10    (7% for the score (4) guy)</DIV>
<DIV style="FONT: 10pt arial"> 6     
9     (8%)</DIV>
<DIV style="FONT: 10pt arial"> 5     
8     (9%)</DIV>
<DIV style="FONT: 10pt arial">(4)   (7)    (10%)</DIV>
<DIV style="FONT: 10pt arial"> 3     
6     (9%)</DIV>
<DIV style="FONT: 10pt arial"> 2     
5     (8%)</DIV>
<DIV style="FONT: 10pt arial"> 1     
4     (28% for the score (4) guy, since results are capped 
at 1, 7% for the score (7) guy)</DIV>
<DIV style="FONT: 10pt arial"> 1     
3     (6% for the score (7) guy)</DIV>
<DIV style="FONT: 10pt arial"> 1     
2     (5% for the score (7) guy)</DIV>
<DIV style="FONT: 10pt arial"> 1     
1     (10% for the score (7) guy)</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">So this illustrates the way add-subtract dice, 
with a comparative resolution system, works. It also illustrates the problem of 
"range of randomness vs range of abilites." Which is all a very long way of 
saying, do you want the dice to be very important.</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">For example, you could have ability ranges of 
1-10, and use d20 with them; the dice are twice as big as the abilities in 
importance. This encorourages players to get a heap of low-level skills, instead 
of a few highly-focused ones, since skill 7 can easily be beaten by skill 4. Or 
you could have ability ranges of 1-20, and use d6 with them; the dice are 
one-third as big as the abilities in importance. This encourages players to get 
a few highly-focused skills, since skill 7 will only with difficulty be beaten 
by skill 4. </DIV>
<DIV style="FONT: 10pt arial">    And then of course the question 
is of realism, whether you want that. For my part, as I said, I feel that as 
skills improve, the range of results narrows, too; the extremely good or bad 
results are still possible (like, I can beat Carl Lewis in a 100m sprint, in the 
1 in 1,000 times he falls over along the way). So in a realistic system, you 
might have something like,</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">"Skills range from 1 to 20. Performance is 
adjusted by a d6-d6 roll. However, when you have skill level 6-10, you get to 
roll d6-6 twice, and take the result closest to 0 (negative or positive). When 
you have skill level 11-15, you get to roll d6-d6 three times, and take the 
closest-to-zero result. At skill 16-20, roll four times, and take the 
closest-to-zero result." </DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">This would give the effect of the randomness being 
one-third as important as the skill level, overall, but that as skill level 
improves, the randomness becomes less important, not only in absolute terms 
(d6-d6, -5 to +5, is relatively larger compared to 1-5 than to 16-20) but also 
in practical terms (rolling d6-d6 four times and taking the closest-to-zero 
result gives a much greater chance of a close-to-zero result than a single 
roll). </DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">Of course, such a system would be unwieldy and a 
pain in the arse to use, but it's just an illustration I made up just this 
moment:) Consider it the "average results for a character" version of the WoD's 
system. In that, you need to roll 6 or above (or whatever it is now) on a d10 
for a "success", but as you increase in skill, you get to roll more d10s. So you 
have greater chances of scoring successes. In this system I just made up, you'd 
have greater chance of success, but also greater chance of a consistent result. 
</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">In d4-d4, I wanted this effect, of consistent 
results, and also of the higher-skilled guy feeling he could quite reliably deal 
with the lower-skilled guy. So I had the Performance Ladder,</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">Famous</DIV>
<DIV style="FONT: 10pt arial">Olympic</DIV>
<DIV style="FONT: 10pt arial">Outstanding</DIV>
<DIV style="FONT: 10pt arial">Excellent</DIV>
<DIV style="FONT: 10pt arial">Good</DIV>
<DIV style="FONT: 10pt arial">Middling</DIV>
<DIV style="FONT: 10pt arial">Fair</DIV>
<DIV style="FONT: 10pt arial">Ordinary</DIV>
<DIV style="FONT: 10pt arial">Poor</DIV>
<DIV style="FONT: 10pt arial">Terrible</DIV>
<DIV style="FONT: 10pt arial">Crap</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">The normal human range of abilities is Poor to 
Outstanding. The d4-d4 roll adjusts their performance by -3 to +3, giving them 
results of Crap to Famous; results are capped at those. So the random die roll 
has a range 7, the abilities the same range of 7, and the actual performance a 
range of 11. (By comparison, what Lev is proposing is essentially a range of 100 
in abilities, 100 in die rolls, and 4 in results: critical failure, failure, 
success, critical success). But (in d4-d4) the best an Ordinary guy can do is 
Good (uni graduate level), while the worst he can do is Crap (zero knowledge 
level). The uni graduate )Good) guy can at best manage an Olympic performance, 
and at worst will be Ordinary. He can never be Poor, or Crap, or Famous. Not 
without help, that is (ie die roll bonuses). The Ordinary guy will consistently 
be Ordinary, Poor or Fair, and the Good guy consistently Middling, Excellent or 
Good.</DIV>
<DIV style="FONT: 10pt arial">    So it's a bit like the system I 
just made up above, but less unwieldy. The system I made up above gives you the 
narrower range of results as the skill increases; d4-d4, you've got the same 
range of results and probabilities, but because the randomness (-3 to +3) is 
smaller than the range of performances (0 to 10) you get it that it's impossible 
(normally, without bonus/malus) for the Good guy to do Crap, or the Ordinary guy 
to do Famously. </DIV>
<DIV style="FONT: 10pt arial">    It's not entirely realistic, 
but I think it comes close. It gives a higher than realistic chance of really 
bad or really good results in performance, but I like that because really good 
or bad results make for entertaining games. </DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">Sorry I write so much. I took a disadvantage, 
"loquacious", and used the points in "handsome." No, really!:)</DIV>
<DIV style="FONT: 10pt arial"> </DIV>
<DIV style="FONT: 10pt arial">Cheers,<BR>Kyle<BR>Better Mousetrap Games<BR>home 
of d4-d4 and other stuff<BR><A 
href="http://www.rpgnow.com/default.php?manufacturers_id=339">http://www.rpgnow.com/default.php?manufacturers_id=339</A></DIV></BODY></HTML>