This isn't really a conversation I can stay in, without winding up banned. See ya.

Stripping all the ranting and sacrasm and flaming out of Critias's last post-- most people who feel the Shadowrun system is too complex generally doesn't fully understand the rules, or is getting bogged down in the broken parts of the game. SR3 isn't perfect, but it's not nearly as bad as the SR4 FAQs are trying to make out.

Insulting my intelligence? Amazing.

I'm not too stupid to run Shadowrun and none of my players are too stupid to play.

We just have better things to do.

It saddens me to see a group of obviously intelligent people reduced to childishness because they're favroite toy is being taken away and they're throwing a tantrum.

Its clear that intelligent conversation concerning this is not possible and I definitely won't be continuing to post here.

To add on to Cain's post...

I'm a geek, through and through. I've calculated powers of 2 for fun repeatedly. I love math. Sometimes when I'm with my wife, I imagine she's a number (I like i best, since it's irrational. j/k! Don't beat me, honey!)

When I play shadowrun, figuring out the best distribution of my pool is enjoyable for me. Even though it is very ooc, it's an aspect I do really enjoy (just like tweaking my cyber list to get that extra .01 of essence). I suspect critia also gets a kick out of doing the math part of the game. And frankly, D20 and what we've seen of SR4, really doesn't support that sort of mental number crunching.

There are plenty of people though (like art majors) who really don't enjoy this beautiful side of SR. It isn't that they CAN'T, it's that they don't want to. I can understand wanting a mechanic which is largely invisible so the game can just move, and frankly, SR's current mechanics are anything but invisible.

They must not have used improved invisiblity.

I likes the numbers...

Actually that example really doesn't make much sense. Do you mean that having rules that characters try to act within to achieve optimum effects is bad? Then I would need to ask what the alternative is.

Do we simply tell the characters how many opponents they are faced with and sit back while they describe the vicious and -dramatic- ways in which they triumph between stuffing handfuls of cheetos in their voracious maws and swilling enough soda to cause insulin shock in less greasy humans?

I suppose that is how some groups "roll" but I see no stigma attached to considering the statistics of your character's action before you make it.

OK, the violence in my answer to Critias was due to his insulting post (we shadowrun srIII player are smarter, those who don't like don't understand it blabla...).
Linking that to the evidence of his non existent skill in statistics I couldn't have make the point ....

By the way I am pretty sure to be the one to like the less sr3 in my group.
By the way I am pretty sure as well I am the one that understand the more of the rules and stats ...

The worst rules are dumb AND complex one.

By the way Raskolnikov, I prefer as a GM to have player trying to find which action is the best in the situation - not in the flawed rules. Because it is funnier, and always different (the flaws in rules stand still). And as a GM I always make rules logical, so if an action seems to hard to reach in a system, I don't follow the "rules".
And my point was more about a stupid rule that say firing 10% less bullets you can be twice as accurate ... I wouldn't call that "strategy" man. This is min/maxing capabilities

Wow 50% sounds like a big number, even though the total probabilities vary by a very small amount.

Trying to attack three people at a time is harder than trying to attack two people at once! Does this mean a character is min-maxing if they elect to only strike two of the three combatants they are faced with? You are making invalid leaps of logic based on an example that is incorrect to begin with.

Let us assume you had made a proper example for what you stated you wished to convey however. A poor rule. The rule does not model reality to the extent you would like it. You maintain this is a perfect time to simply fudge the action. Those of use advocating a hard approach are not saying the rule should be used in a flawed state, rather that any change should be codified and applied consistently from then on.

If something is broken badly enough it -needs- to be changed, then by all means change it, but change it to a rule.

The example is a pretty good one as far as poor rules go, Rask: it's not sensible to hit half as often when shooting 10 bullets as when shooting 9, even if the probabilities to hit are 5% and 10% respectively. It might seem like 5% doesn't make a big difference, until you have to run past 15 people shooting at you with full-auto. Then the difference between a 5% chance of success and a 10% chance may make the difference between life and, um, not quite so much life.

However, Taki, Rask's right that in such cases it is preferable to actually come up with a rule to fix these things instead of just ignoring the rules, or making something up on the spot. It's more consistent if you make up a rule, and if you spend a little while thinking about it, you'll probably come up with something better than if you just fudge it.

First : the flaw in rules will be changed. It is called fixed TN.
Second : the flaw on the TN SR3 is generalized, but it is effectively very caricatural here.

As some people do not very well understand statistics apparently - lets make a calculation.

(Rask: Wow 50% sounds like a big number, even though the total probabilities vary by a very small amount.)

OK
9 bullets : TN11, good sammy sr3 7 dices :
Proba to have at least one success = proba not to have 0 success
proba to have 0 success = proba not to reach TN11 (34/36) multiplied by itself 7 times = 67%
SO 9 bullets : TN11, good sammy sr3 7 dices : 33% to have at least one hit (one chance over three to get a deadly heavy fire)
With 10 bullets the result is : 18%

... The total probability is changed by 15% on total chances, (18% represent only 54% of success of the best choice firing only 9 bullets), which is not a small amount, and I really think one more bullet for that result IS a flaw (especially when 11 bullet will be the same - if you would find a proper weapon to fire 11 bullets).

The flaw is just existing because
1 a difference of 1 on a TN is a too big difference on a dice of 6
2 a TN of 6/7 is the same

hurrah to fixed TN !
Death to SR3 yahaaaaaaa!
(how much enemies ???)

Edit : replace "which represent" by "18% represent" for clarity

We played and go on playing SR3 because even if the the rules can be a lot better and simpler, there is really some very good things in the background, world and so on.
Because it was possible as well to display characters with a lot of option (different ammo, cyber, magic, lots of weapons), and because those option were easily represented - just pick up in a list.

I think and hope rule will be better in sr4 - and since that I won't put my sr3 book in the litter anyway
Rules just can't handle everything. I prefer rules simple than rules trying and failing to simulate reality.

You can be my enemy, until you are (edit:)NOT a hacker I don't fear a lot.

What do you suppose would happen to your example, Taki, if you applied the Rule of 6 to fives as well?

Here's my question for you, Taki: What makes you think fixed TNs will be any less complex? We're discussing gains and losses to the dice rolled, plus probable changes in threshold values. Calculating those odds becomes a fair amount more complex, and could produce any number of annoying artifacts as well.

Most notably things like, "Well, I've already got a 1/3 chance to hit and I can't get any lower, so I'll go for a headshot as well." or going from 9 bullets to 10 makes the difference between hitting 1/3 of the time and hitting never!

Ellery:
My solution for "impossible" shoot:
let say 7 dices with a 8 dices penalty (pretty hard, should be achievable but with a real small probabilities).

I state : 1 dice is the minimum for a pool.
So in the example I only can substract 6 of the penalty, I have a 2 dices penalty left.
The trick: I now ADD the 2 penalty dices left with my 1 die remaining, throw them all, and keep the SMALLER RESULT. Wich means I have to score a minimum of 5 on all dices to have a hit. probability : one chance over (3*3) (seems correct, don't you think ?). EDIT : the right result here is 1/27.

This is a good simple way to handles very hard task, because it doesn't change in any way the normal situation as would explosive dice.
Still it isn't perfect has it doesn't deal with tests which need more than one hit to succeed.

To Cain (Calculating those odds becomes a fair amount more complex):
I disagree. The complexity here is just : counting the penalty.
It doesn't change the complexity at all if you apply the penalty on either dices or TN.

When I think the SR3 system is too complex, I it is not because of TN, even if this design has inducced a lot of complexity in rules. What was good in the SR3 system was the overall balance - it was working, this depends on a lot of details ... And for instant we haven't any of those details to judge SR4 who has better ground design TMO. If SR4 works, it will surely be simpler, and better (because some complexity in SR3 wasn't needed TMO).

Counting the penalty on TN is too bad with a 6 sided die (or you have to stick with TN 2 3 4 5 and 6).

I prefere players to keep in mind a strategy on situation, than a strategy in using rules. Strategic vs Dramatic ?

Taki--your idea has merit. But let's think about it a little more. Firstly, this adds an extra complexity to the base mechanic--you have both a "roll and count successes" and a "roll and keep the lowest" mechanic. But, okay, maybe we can deal with that.

But now let's see if we've really improved the situation over the variable TN case.

Supposing that an extra bullet will subtract one die, let's consider a case where we shoot 8 bullets and have 2 dice, 9 and have 1 die, and 10 and have the lesser of 2 dice. In the first case, our probability of at least one success is 55%. In the second case, it's 33%. In the third case, it's 11%.

Notice that 33% to 11% is a worse change than the 33% to 18% change that you thought was so problematic in SR3. 55% to 33% is a pretty big jump too. (It's 42% to 33% in SR3.)

You could come up with other changes instead (e.g. if the number of dice goes below 1, the opponent gets that many dice to oppose the test, even if the test isn't normally opposed--this would cut it down from 33% to 22%, which is marginally better than the 33% to 18% change you disliked).

Combine that with exploding dice, and you've got a system that works a little better than the existing variable TN system as long as you don't roll too many dice--but it isn't any simpler, either, which makes one wonder what the point of the whole exercise was.

Especially since you could achieve superior effects by making any number of minor changes to the variable TN system. For example, if you have a rule of 5 that says, "Reroll your 5s, and those that come up 5 or 6 count as 6 instead (but don't roll them again)", then your probabilities go from 33% at TN11 to 23% at TN12 (and then 18% at TN13). If you have a rule of 5-or-6 that says "add 4 and roll again", then the progression of probabilities also goes from 33% at TN 11 to 23% at TN12 (but this time goes to 20% at TN13). And so on. Edit--oops, these values are for TN12, 13, and 14 since the irregularity occurs at different places in the rule-of-5-or-6 system I described.

There's all kinds of ways to fix the variable TN system, just like there's all kinds of ways to fix the fixed TN system. The advantage of fixing the variable TN system is that you don't have to recalibrate every single rule in the entire game, on the basis of a system that you have almost no experience with; and you circumvent the largely-intractable problems of fixed TN systems where people with large numbers of dice do not care about penalties to the numbers of dice, while people with small numbers of dice cannot tolerate penalties that increase the number of hits required for success.

You are quite right in a way (especially when you see my EDIT in last post) :
It seems pretty odd to just add one die and have so many changes in the probability to have hits. And the system I propose was just a patch (which advantage was not to overrules the design, but to occur in a range were the design doesn't apply).
But I said : the good thing in SR3 was details that make the system balanced.
Here you just showed that the penalty of 1 die by bullet doesn't work after a certain time in those conditions.

By the way my example was, as I said caricatural. you can take any system and show it doesn't work at its extremities. The example was here to show the flow of TN, which occurs more often from 5 to 6 which represent a penalty of 1 or 2.
That is way I think the fixed TN is better : the range of use will be often relevant.

Your Highness, I was just taking fun - ok that was stupid and just fun to me - I am sorry for that
I respect your opinion even if TMO SR4 showed improvement in the base design.

My disrespect go only to those who think and say that not following their opinion is a mark of stupidity, which is clearly not your case (they will recognize themself anyway).

... Now let's pray ...
Please Dev give us a good system, make the rule clear, simple and consistent

Has someone got a idea when we could say "alea jacta est" - fate/dices are thrown, and when no change and turning back will be possible before ... errata ?

That's OK Taki.

I think the 'alea jacta est' moment has gone. If SR4 is to be ready by GenCon, I don't think there is time to change things now. We will have to just wait and see.

I still doesn't agree with variable ... Maybe because I am for soft rules (?), I must admit it isn't a easy way to get along with that kind of rpg way, but I had a lot of fun in it.
I do understand that a lot of people preferre hard rules, but I reckon SR4 will be more hard than soft anyway !

Thinking SR4 will be bad will probably help me being pleased by the result

I just don't think we've seen enough statistical breakdown from the developers...have any of them analyzed the effect of a fixed TN like the people on this forum have?

In the thread I started about probability and statistics, the answer from one playtester/freelancer was yes, they are being considered. Of course we got nothing else, but we didn't expect it either.

I would have liked to have seen an explanation on the order of, "We looked at the probability of success for X and Y in SR3 and compared it to X an Y in SR4. We found Z."

So all we have is a Yes and some hope and faith.