QUOTE (Cheops)
Or go back to SR3 like I am doing. Skill 5 is professional and 9+ is world class. Variable TN makes it possible to actually challenge someone with a big dice pool.
...
A normal task is TN 4 while an impossible one is TN 10+. Instead of just requiring more successes to accomplish hard tasks you needed to beat a higher target number. This meant that a high skilled character had better chances of getting a success against a hard target number while also getting tons more successes than an untrained person would get on normal difficulties.
I am sick and tired of that meme. I was sick and tired of that meme
five years ago. Variable Target Numbers are dead, because they
do not behave as advertised. White Wolf went static target numbers too, and that is not why they collapsed financially and ended up getting purchased in a bankruptcy fire sale by the division of CCP run by Ryan Dancey (the man who had orchestrated their collapsing market share starting in 1998 or so while working for WotC). They collapsed because in the face of a deliberate and coordinated attack on the shelf space and market share by OGL 3e D&D they decided to try to get "new fans" by making a whole new World of Darkness that no one liked.
A TN 10+ isn't impossible. It's not even especially hard. A random civilian with a relevant interest (skill 3) succeeds 23% of the time. The supposedly world class guy succeeds 54% of the time. What this means is that if you have a difficult stunt, the chances are very good that Jackie Chan, even a cybernetically enhanced Jackie Chan, will fail utterly. On the other hand, if grab a bunch of random dudes off the street and ask them to perform that stunt like you were dong MXC, almost one in four of them will succeed. The variable target number generates the wrong results coming ad going because each die has the same chance to succeed as any other die - so the MXC concept of "let's send a dozen random civilians to attempt this task, one after the other" is more likely to generate success than any badass could possibly be capable of. I fact, our supposedly "world class" character is no more likely to succeed at any task than
three random dudes are.
Compare to the improved system, where TNs are fixed at 5, and the number of hits required to perform various feats increases as the difficulty of the task increases. When moving from a Dice Pool of 3 to a Dice Pool of 9, a task requiring 3 hits goes from a 4% to a 62%. The more skilled individual is substantially more valuable than is a group of less skilled individuals for attempting difficult tasks. Even large groups of individuals with less skill. To have the same chance that one person will succeed at a Threshold of 3 out of the group of people with a dicepool of 3, you'd need a group of
twenty three. Which is extremely different than needing a group of just 3 (since Shadowrun doesn't normally deal with groups of more than a dozen people in any case).
And the effects of modifiers are even easier to understand and predict in the fixed target number system! By a lot. You add or remove 3 dice, and you on average add or remove one hit. Do you know what happens to the averages when you increase or reduce TN by 3? Of course you don't! Because the effects are different depending upon what the TN was to begin with. If the TN
was 5, then your chances of success
were 33% per die rolled. If you reduce it by 3, your chances of success are 83% per die rolled. If you increase it by 3, your chances are now 14% per die rolled. On the other hand, if your TN
was 6, then your chances of success
were 17% per die rolled, and a TN decrease of 3 puts it at 67% per die rolled and an increase of 3 puts it at 11% per die rolled. And yeah, remember also that if the TN was 7, your original chances were the exact same 17% per die rolled, and yet with the TN decrease it goes to 50% per die and with the TN increase it goes to 8.3% per die.
It is a nightmare of opacity, and more importantly
it doesn't generate the right results! The goal is to have the task resolution test not only generate a
predictable result (which the SR4 system is superior at doing), but also to produce a result which is appropriate within genre - namely to have the highly skilled ninja be
substantially more likely to succeed at picking a lock, sneaking across a courtyard, or leaping over a razor wire patch than a random contestant on an inscrutable Nipponese game show. And fixed TNs coupled with variant thresholds delivers on that, and variable TNs does not. Furthermore, the fixed TN, variable threshold system actually can produce effects which are
genuinely impossible. And it segues simply and effectively from impossible to plausible to easy as characters get a larger dice pool. A dicepool of 3
cannot get 4 hits. It's impossible. A dicepool of 16 succeeds 83% of the time and can even "buy hits" to succeed automatically in downtime. There is no equivalent in SR3 mechanics, and cannot be - since the dicepool of 16 is
by definition less likely to succeed at any task than is at least one person from a randomly selected group of a half dozen sarariman.
This is not mere opinion. That the SR4 system is faster, more transparent, and generates better results out of its random number generators than the SR3 variable TN system ever did or could is a mathematical
fact. If I was going to design a new edition of any game using a dicepool system, I would use SR4 as a base and not SR3. In fact,
I did.-Frank