QUOTE (mfb)
if fixed TNs scale so well, then why does it require GM intervention to prevent unaugmented characters from making 1km sniper shots in complete darkness without aiming?
I'm never going to agree with you that a guy with a nearly superhuman ability with shooting in adverse conditions who knows the exact location of a target should be unable to hit said target with a gun while blind folded. I've actually seen people do that. I will say that Shadowrun gives out range penalties which are somewhat low. Extreme Range actually seems like a harder shot than point blank while blind folded. And of course every edition of Shadowrun has presented per-bullet accuracy which is totally unrealistic and extremely high. But none of that is a condemnation of fixed TNs in general.
QUOTE
scaling with fixed TNs requires increasing threshold and reducing dice pools. if you are at the shallow end of the dice pool, you will quickly reach the point where you don't have--and can't get--enough dice to succeed. if you are at the deep end of the dice pool, you will quickly reach the point where no amount of pool-reducing or threshold-increasing can stop your mighty power. the whole reason SR4 requires hard caps because fixed TNs don't scale well--they had to artificially box the mechanic in.
Uh... we're talking logs and exponents. In variable target numbers you can add 6 to the TN and then the character will need six times as many dice (or an arbitrary TN reducing ability) to have any realistic chance of success. With a fixed TN you can add two to the Threshold and the character needs six more dice or an arbitrary threshold reduction to have any realistic chance of success. Those statements are pretty much the same, save for the fact that rolling six more dice is tractable and rolling six times as many dice is intractable. A 10 die pool in SR4 is a lot like an 8 die pool in SR3, so which is less daunting to you: rolling 16 dice or rolling 48 dice? Seriously now, are you going to tell me that rolling 48 dice looking for exploding sixes is something that you ever want to do in your whole life?
People at the high end of the dice pool in SR3 don't actually get any better by getting more dice. If you are looking for two elevens you basically aren't going to get it if you are rolling 7 dice or 17. You haven't hit a hard cap, you've just come to a place where getting any actual improvement in the tasks you can expect to achieve requires picking up dozens, if not hundreds of additional dice. And what with skills costing 2.5 times rating very early in that process it's simply never ever going to happen. They are, in effect, capped. It's just that the cap they happen to be at is a result of the fact that they live in a world of exponential difficulty and they advance linearly. They rapidly achieve a state where they are realistically not going to fail at the easier tasks and realistically cannot expect to ever get a meaningful chance at the harder tasks. A capped fixed TN system provides the same result, but it doesn't bother to make meaningless distinctions between people in the upper tier of potential - and instead throws down the very real possibility of actually i=using more potent creatures and things without scrapping the system repeatedly to reset into vehicular and naval reference frames.
Is SR4 perfect? No. Is the math always great? No. But the core math pits exponential difficulty against exponential skill - and that allows the game to at least theoretically be mapped onto an essentially infinite scale. Something that Variable TNs from previous editions could never accomplish. Yes, you can nit pick on weird crap like the fact that extreme range is double the per-shot penalty of pitch darkness (which is just bizarre), and if you really want I can crack open an SR3 text and start reading off inane probabilities generated for ordinary actions by glitches in the TN modifiers. But the fact that you couldn't generate the Universal Brotherhood Infiltrators by actually rolling the dice (it took over 4000 attempts with Willpower 6 recruits to create one "Good Merge") is completely tangential to the fact that you could have written a mechanic for SR3 that did generate the proportions called for in the stories.
The ability to distinguish between "extremely good, but has no realistic chance of doing anything more difficult than X" and "even better, but still has no realistic chance of doing anything more difficult than the very same X" is not particularly important to me. However, the ability to distinguish between "Good enough to do X, but not Y" and "Good enough to do X and Y" is critical. And a fixed TN system does the latter across a much larger field. And by much larger field, I mean literally hundreds of millions of times the field range.
-Frank