Fixed TN - what's the big deal?, Trying to understand certain objections Options

[El Ojitos]

Every time dice are rolled in any RPG it is to determine the outcome of a situation that isn't completely predictable.
What rules try to create is a "realistic" probability model of a given situation. And you can do that in many ways.

The classic D20 systems have only one number - the TN - to measure all applicable factors (ability, difficulty, circumstances etc). Everything is added or subtracted to create one total. That has its advatages. A GM knows that by imposing a TN modifier of 1 he changes the probability of success by exactla 5%. It's easy to judge situations that way. The clear disadvantage in my view ist, that you can't satisfactorily see how well a char has performed. There's only success and failure.

In SR3 there are three numbers that determine the probability of the outcome: number of dice, TN and sucesses needed. All of these numbers can be influenced by a variety of factors. It's hard to keep track of whether a speciel set of circumstances will give you extra dice or reduce the TN. What is more, it's very hard to get a feel for what it does to the probability of success if a GM grants a player an extra die or if he require him to have one more success - even without the unpleasant phenomenon that TN 6 = TN 7.
The basic philosophy behind all that seems clear: the number of dice indicates a char's basic ability, the TN indicates how difficult the task at hand is, and the number of successes tells me how well he has performed in it. Unfortunately that basic idea seems to have gotten a little muddled. And sometimes it's hard to decide, which area is affected. Does my totem give me extra ability at performing certrain spells or does it help me directly whenever I perform one of those spells? In other words: Should it affect the skill or the difficulty?

In SR4 we are led to believe the TN will be fixed at 5. As a simulation of an unpredictable situation that will certainly do just as well as all of the others. The rules and the GM can affect the probabilty of the outcome by influencing the two remaining numbers. Maybe it will be easier to get a feel for probabilties then - it certainly won't be harder.

The critics of SR4 have attacked many (practically all) aspects of the new rules we know of so far. I'm not saying I disagree with them on all aspects. The TN however seems to be of no big importance to me.

[Little Bill]

I believe the question is granularity.
Adding or subtracting an entire die or entire success makes things much more or less difficult than adding or subtrating a point of TN. SR4 will restrict the ability to lightly adjust a difficulty up or down by these small amounts.

[blakkie]

[Ellery]

The illusion of d20 is that you actually know what is happening when the probability of success changes by 5%. When something changes from 5% probability to 0% probability, you've just changed a "slim chance" to "none at all". This can lead to undesirable effects in, say, combat, because someone has now become invulnerable. Alternatively, if you change from 5% to 5% (i.e. you don't change), then bonuses that used to make a huge difference (+3, from 20% to 5%!) now don't matter at all.

So although it's easy to understand that the probability change is 5%, it's hard to know what the impact will be on your game, because that impact varies dramatically depending on where in the 1-20 range you already are. A +3 bonus might be a huge win (in the 15-18 range), largely insignificant (in the 3-6 range), or completely irrelevant (at 20).

It's harder to compute the probabilities for TN5 with variable numbers of dice, and a little harder still with variable TN and variable numbers of dice. The problem, though, with TN5 is that it has the same flaw as d20, just disguised. A static penalty of 4 dice is a huge deal if you're rolling 6 dice, since you'll go from an average of 2 successes to an average of 2/3 of a success--and your failure rate will go way up. However, a static penalty of 4 dice is not so important if you're rolling 18 dice; generally, if 6 successes will get the job done, 4 2/3 will usually get it done as well. And we also have the problem of penalties leaving you with one die (1/3 chance!) or zero dice (no chance!) which is a huge difference; or with penalties always leaving you with one die, giving you a 1/3 chance of accomplishing the most ridiculously difficult tasks. The problems are less severe if you only give dice bonuses, and account for penalties by increasing the number of hits you need to get a success, but the same problems remain.

The advantage of the variable TN system is that a +1 penalty to the TN equates, on average, to about 25% fewer successes. The disadvantage is that it actually ranges from 50% (TN 5 to 6) to 0% (6 to 7). There are a number of ways to partially compensate for this unevenness. The advantage of this unevenness, though, is that it's not systematic--you don't always have penalties becoming more or less important as the TN rises. But it will tend to average out to 25%, since other modifiers will move the TN around. (For some things, like resisting spells, it does cause a strange phenomenon where force 6 spells are the thing to have--again, this could be fixed.)

Some of us play characters of widely different power levels, at which point it's nice to have the system behave consistently at all levels. One of my characters is a nine year old street urchin named Jane; if she picks up a gun and aims at someone, she will, on average, get 33% more successes (133% = 1/75%) than if she doesn't aim. I also have a character who is part of an elite anti-shadowrunner and counterintrusion group; if she picks up a gun and aims at someone she too will get on average 33% more successes. The system scales.

So, to me, that's the big deal about a fixed TN of 5. It just doesn't scale with the same grace as variable TN.

Incidentally, you already explained the three aspects of determining success: the difficulty of the action (maps to TN), the ability of the character (maps to # of dice), and how well the character performed (maps to # of successes). If the game designers religiously followed that breakdown, it'd be quite a bit easier to keep track of the different numbers. It's not too far from that now--generally, the player knows their ability, and keeps track of the # of dice; the GM knows the difficulty and states TN modifiers (although the player may have augmentations that also reduce difficulty); and the # of successes determines the outcome according to a simple table or scaling or opposed contest. Fixing this up to be more consistent shouldn't have been that hard.

[sanctusmortis]

My biggest problem is simple: with one dice and a TN of 16, you can still theoretically succeed. With one dice, set TN 5 but 3 hits needed, you can't... suddenly the "not likely" becomes "not possible", and that's not really the spirit of things.

[blakkie]

[blakkie]

Ellery:

Among other things, you seemed to have forgotten with variable TN that in fact a penalty/benefit can mean little or in fact absolutely nothing. Variable TN often hits it's hard cap, TN 2. As well just a plus TN +/-1 or 2 can mean a huge swing in probability for for multiple successes.

There is also the benefit of being able to treat success totals of rolls from different sources equally. This allows you to do things like combine multiple rolls into one (such as Dodge and Soak rolls) where previously the penalties/benefits for each roll was different so they had to be kept separate. It also aids in comparing to each other rolls made at two different times that you didn't know in advance were going to be compared. Such as opposing a spell cast several hours past. If the hits become the Force of the spell you can do this.

Now the later can be done in variable TN by using the same base TN throughout the magic system and having it all work on comparing successes of opposed rolls. But opposed rolls are fixed TNs strength, not variable TNs.

P.S. You need to be more careful about talking just averages hits of a roll of dice for a fixed TN, not how the full curves of how many hits there are. There is a difference there.

[Ellery]

Having a limit of TN2 is somewhat annoying, but a lot less annoying than most other limits since you are very nearly performing at the limit of your ability with TN2. You'll note that I didn't really complain about this limit for D&D either (I didn't bemoan the difference between 1 and 1-3, for instance). Generally, it's more important for games to have rare success than rare failure.

Also, there's no reason why you can't combine the totals of rolls with variable TN, if it's a sensible thing to do. Likewise with remembering the number of hits and using it in an opposed test. These are very often not sensible things to do (e.g. just adding successes from centering to the successes from a main test makes centering as important as skill, which is probably not desired), but they're still options if you want them.

Finally, the difference in coefficient of variation between two dice and six dice is also more than the difference in c.v. between fourteen and eighteen. The same point applies. I didn't mention it this time because it's complicated enough to be distracting and doesn't change any conclusions.

[Charon]

[Vaevictis]

I know it's kind of off-topic, but I would submit that there is an easy way to solve the 6/7 problem. Enforce the 1 is always a failure idea. If you have a TN of 7, roll the die again. On anything except a one, you succeed. To balance this out (because it is a disadvantage for the roller), spending a karma pool lets you reroll the failures from the last "six" multiple they had.

Example: Target number is thirteen, and you roll two sixes, one turns up 4, one turns up six, then you roll a one, you fail. But, you can spend a karma pool and reroll one die from 6, and one from 12 -- the second die will generate a success on anything but a one, and the other will have to roll a six and then anything but a one.

I think that gets rid of the 5-6-7 thing, and in a fair and reasonable manner, myself.

As a player of AD&D and SR (and a bit of oWoD), I cannot overstate how important having a variable TN to indicate degree of difficulty AND multiple dice for degree of success is. Simple pass/fail is a very poor method, imho, and so is not having the ability for someone to succeed by freak chance no matter how sucky that person is at the task in question.

[fionn]

I really liked the fluidity of the old system, it allowed for 2 different ways to establish a success.
1) by varying the TN or
2) varing the number of successes at a tn.

[blakkie]

[Cain]

Here's where I'm having an issue. Fewer variables means the modifiers get crammed further into fewer categories, potentially meaning more math needs to be done to figure out what needs to be rolled. What's more, possibly more of the onus will land on the players.

For example, under SR3, only the GM needs to come up with the TN. The players only need to keep track of their own dice rolled. Under d20, the players and GMs both need to keep track of the TN modifiers. This leads to situations like the one I encountered not long ago, which went something like this: "Okay, I add 6 for my BAB... minus two because he's twenty feet away... plus one because he's less than thrity feet away... minus four because he's in melee... plus three from my dexterity... minus two since I'm rapid-firing... plus one 'cause I'm using a dagger.... etc, etc.

Now, before you say this happens in SR3, let me point out again that the modifiers are calculated by the GM. That speeds things up a lot more, since the GM is the only one who needs to know every last modifer.

Also, don't forget that d20 has its own probability blips. In combat, the odds of hitting or missing can never go below 5%. You could be half-blind, one armed, both legs broken, against an opponent in Ultimatium full-plate armor, and throwing the halberd with your teeth; you keep the 5% chance of hitting (and scoring a crit). Out of combat, that +1 could shift the odds from 5% to impossible, if the new TN exceeds 20+Modifiers; and calculating those modifiers becomes a more complicated, cooperative task.

[blakkie]

[weblife]

While its true that a penalty of 1 die is vastly different if you have 2 or 10 dice total, you seem to be forgetting that SR4 will have Attribute+Skill as base dicepool for each test.

Assuming attribute 3 is still average, and skill 3 is still average, most normal people, doing normal things, will have 6 dice in their test.

Beginning runners will laikely have some attributes in the 6's, and skill to match, making it 12 dice.

Stronger characters are likely to have upwards of 20 dice in their core specialties.

Basically this means that you will rarely, if ever, have someone rolling a pathetic 1 die in a test. They'd have to be defaulting to a low attribute to be in that situation, a situation in which failure is likely to be expected.

Also, the way I read it, even a single success means the test is some kind of success. Extra successes just mean you got the business done even better.

Fx if my character shoots another character and I end up with 1 net success, then the other guy is hurting atleast a little.

Point being, you are correct that the method of removing/adding dice can get skewed at the low end, but upwards it scales pretty well. Although, yes, you will hit a point where you can expectedly deliver 5 successes or more every time you use the skill. However, whether this situation is attainable or not depends wholly on the price in karma to achieve it, and if its even possible to "cap out" within the game rules.

Most tests will likely be within the 6-15 dice in the pool range.

[Nerbert]

[blakkie]

[Cain]

[Ellery]

[El Ojitos]

[blakkie]

[blakkie]

[TheOneRonin]

I keep most of the target numbers hidden in my games. What I do is have players group and call out their die results, starting with the lowest. It would look something like this:

Player 1: I shoot the guard.
GM: Roll your pistols skill.
Player 1: I get one 2, two 4's, one 5, one 7, and three 8s.

I know the TN for this test is 5, so I start counting at 5. That gives him 5 successes. It's not as slow as you might think. Though my players are in the habit of sorting their dice immediately after rolling them, so I don't really have to wait too long for them to do it .

Here's a tangent...if the TN in SR4 is fixed at 5, and all that really changes is the number of dice rolled, how can you have any hidden tests? Won't the players ALWAYS know exactly how many successes they get on a given test? I suppose opposed rolls is one way, and maybe having a hidden success threshold is another, but otherwise the players always know how well they do on every test. I suppose it will speed things up substantially, which is a good thing.

[blakkie]

[El Ojitos]