TN 4 makes the current mechanics a coin flip system much like Ubiquity (Hollowed Earth Expedition, Desolation, All for One), which is not a bad thing. Ubiquity allows you to take the average which then turns into if you think you have a good edge over the other guy you don't roll and always take the average. It also means the GM never has to roll if he prefers not to, which can greatly speed up combat. Personally I would LOVE SR done with Ubiquity, they pretty much cleaned up any core mechanic of SR4 im not fond of (but they dont have hacking or rigging to worry about).

Technically Shadowrun allows for you to buy successes now, but I don't think I've ever seen it allowed to apply in a stressful situation, usually only in things done during downtime, or data searching and the like. It also does so at a below average rate (1 success per 4 dice compared to the average 1 in 3). I'd probably keep it that way, either leaving it at 1 per 4 dice, or moving it up to 1 per 3 (still below the average roll at TN4), and only applicable in nonstressful situations.

AFAIK, it's basically arbitrary. We can get the desired results either way. TN5 allows bigger DPs for smaller Thresholds. I'm fine with TN4 if we're talking about decreasing DPs in general, and increasing Thresholds in general.

I'm considering an idea that bonuses to a skill from gear, magic, and cyber should all be treated like augmentations and be capped at one half the character's skill.

Thus, a top level human (minus qualities) could have:

+6 dice from base attributes
+3 dice from attribute augmentations
+6 dice from skills
+3 dice from gear, magic, and skill augmentations
+2 dice from skill specialization

Total: 20 dice

This character wouldn't face a significant challenge until a threshold of 10 using a TN4.

Qualities, other metatypes, and situational modifiers would increase this some.

There are a number of ways to handle 'external bonus' caps. (And have been discussed around here several times). The O-RAW SR4A one is pretty crappy, IMO, but it all depends. It might be nice to make skill the more important factor (via the method you suggest). Possibly capped at '= whole skill' instead, of the TN remains 5? Either way, attrib still lets you get a big boost from natural talent, but training lets you use the externals. The perennial issue with this is that some externals *should* give a novice a big boost (maybe). These caps tend to be extra punitive on the low end, while barely affecting the hyperspecialists (theoretically a balance goal).

Random aside: how strongly do people feel that Specialization shouldn't be +1 instead, esp. if the TN were 4?

What is absent in the TN4 vs TN5 discussion is glitches.

Well I think how you cap external bonuses should depend on which method you go for with skill resolution. I think there should be a list of things that can be exceptions to the normal limit, these exceptions would be pretty much specialization and situational modifiers. Things from tech, magic, etc, should all be handled under the cap on dicepool bonus.

Under the current system with no other changes, I'd go for a cap equal to skill.

Under a system like I proposed earlier, I'd peg the cap at closer to 1+1/3 of the skill(round up). So a person with a rating 1 skill (barely trained) would have up to 2 bonus dice (enough to successfully use a smartlink or the like), 3 bonus dice at rating 4, 5 at rating 7, etc.


As for your question of specialization: I'd personally keep it at 2, but maybe increase its cost slightly (3 or 4 karma for a spec). I'd also have it grant an increase to max bonus dice. So a mechanic might have a skill rating of 5, and have up to 3 bonus dice. But say he's specialized in fixing up trucks, so when working on a truck, he can gain up to 4 bonus dice from outside sources, in addition to getting the 2 extra specialization dice.

Glitches would occur at the exact same rate, but critical glitches would happen more rarely.

What you cap bonuses at must also be a factor of how big you want dice pools to be on the high end.

I don't want them getting much bigger than 20 at the top end. Even then it's becoming mechanically challenging and time consuming to make rolls.

So, that's my next question. How big should dice pools get?

Out of Curiousity...
Why is a large dice pool mechanically challenging or time consuming?

I prefer them at 20 or less, personally.

Because holding and rolling that many dice can be tricky. You end up dropping dice, or making multiple rolls. Dice bricks often only come with twelve dice.

Plus, the larger the dice pool becomes, the longer it takes to pick up the exact number you need and to count hits. Thus, smaller is better at the table.

Just for what it's worth I think TN 4 would be a bad move, I have little else to offer to the discussion other then a gut feeling based on previous discussions, non trivial things should not be coin flips, and trivial things you should not be rolling.

No disrespect intended, but the probability curve of a dice pool, whether the TN is 4 or 5, in no way resembles a coin flip. That would be like saying that the current system comes down to rolling a d3 to determine success.

The only way it'd be a coinflip is if you had a Threshold requirement equal to your dicepool. Otherwise, the probability curve shifts away from 50% chance of success.

Trillinon: None taken, I just have a gut feeling as I said that 4 and up makes success a little too probable. A good way to test this under "live conditions" might be playing around with the prime runner rule from Street Legends, which for all the poopooing it seems to get on here I kind of like.

One thought that occurs to me is keep TN's fixed at 4, but certain things like specialization for example can lower them one step. That could also finally have DNI do something meaningful without dramitcly changing the mechanics of the game or becoming an easy to surmount dicepool modifier by having real true datajack or technomancer lower the TN to do actions by 1. I think I would only allow a minimum TN of 4 or be very careful of situations where multiple TN lowering could stack up.

While we're on the subject a list of things I want to see go away.

1) Penalties for mundane healing for cyber characters.
2) Prejudices (and the accompanying ubiquitous weapons scanners) for cyber characters. By this point ware should be pretty accepted if not cool.
3) The spending edge to uncap hits rule, as this really only benefits mages in the current configuration and it's just a blatant opportunity to cheese the system.
4) Spirits and critters receiving edge.
or
5) The grunt system, if everyone in the system has edge then everyone needs to have edge, if the PC's and Prime runners are supposed to be special because of their edge then it needs to be uniform. It makes no sense to have a group of humanis toughs who all at a minimum have two edge and then give them two edge as a group.
6) Clustering

Things I want to see added:
A form of grounding for foci, as they need a bit more of drawback.

Things I want to see changed:
I want to see a clear delineation in processing power between a man portable device and a server node.

There are others but this is what comes to mind.

Lurker: Well, it's not a change made alone. If the Target Number was 4, then the threshold rules would also be altered, and modifiers reconsidered. It's about changing the scale to be more granular with greater range. So, sure, it would mean more hits, but tough things would require more hits than the current system.

I don't know if I'd want to do away with your entire list, but I'd be interested in hearing your arguments.

Though it does seem like you're generally dissatisfied with Edge. I can't say I blame you. I've always felt there were too many options. Edge should be about extra dice, and really nothing else. I'd even consider going back to a karma pool of dice you can just spend on any roll.

Personally I prefer edge for the ability to reroll as opposed to getting it for extra dice.

That's a problem of the fixed TN system: it has no granularity in low probabilities. If you can roll at all, then you have at least a 50% chance of scoring a hit at TN4, 33% at TN5. To get around that, you need to put the negative modifiers on the Threshold (as in: number of hits required), not the dice pool. Then I don't see the advantage over the variable TN system: there are still 2 different varying quantities to account for.

How about something like that:

TN4
All negative modifiers subtract hits from your roll
In opposed tests, the defender subtracts his hits from the attacker's
All positive modifiers add to the dice pool, with 2 exceptions:
A smartlink makes firearm tests TN3
A control rig makes vehicle tests TN3

Why these exceptions? Because the mundane need character-defining options too. I think that more than power level, it is the ubiquity of mundane solutions that hurt so hard in SR4. If everyone can be be X, then players don't want to play X because they want to be unique. There might be room for something to make Matrix tests TN 3 too, but I don't have a clear idea enough of an SR5 Matrix to propose what.

Everyone prefers to use that option because it's almost always statistically superior.

But dice in Shadowrun means something. Each die represents a piece of talent, skill, or gear that is helping you accomplish a task. You're throwing the combined weight of all of that when you throw your dice. I see edge/karma as the same thing. It's another part of yourself that you can throw at things when they really matter.

Rerolling doesn't quite fit that. That's saying that your extra something gives your smartlink an extra chance to be useful. There's a conceptual disconnect there.

It's not critical, of course, and it's all about feel and concept, so I don't have much else to argue the point with.

Technically, you can put modifiers on either the threshold or the dice pool. Increasing the threshold by 1 or decreasing the dice pool by 2 will have similar results.

But the advantage in a fixed target number isn't in how many places end up being modified. It's in the ease with which you can tally hits and how easy it is to follow in your head. Every die is worth the same value to your character. It always means exactly the same thing. How you roll never changes. What you're looking for when scanning your dice never changes. These are distinct advantages.

No they won't. With dice pools modifiers, you go straight from 1 die (p=50%) to 0 die (p=0%), but this is only equivalent to a threshold modifier of dice pool/2. Then you can keep adding threshold modifiers until the threshold reaches the dice pool. What you gained is the granular descent from 50% to 0%.

Take another look at the table you wrote earlier. With a dice pool modifier system, it is impossible to come up with a roll that has 20% or 40% of scoring a hit. With a Threshold modifier system, it is possible.

See, I figure edge is more like luck. That lucky guy normally would have missed that shot, but made it this one time because of his reroll. It makes edge more valuable, a few extra dice a couple times a session really doesn't have that same feeling.



I don't really like the drop TN to 3 for some tests. Especially something like firearms, as that creates just another advantage for firearms that melee has to compensate for elsewhere (because realistic or not, the martial artist adept or troll combat brute are archetypes in Shadowrun that you can't just ignore). There's also the fact that as soon as you introduce a TN modifier, the temptation comes to introduce more. Even if you don't do it, someone else will, either a different author in the core book, or waiting until a splat or two down the line, but you will see more TN modifiers, and eventually you end up with outrageous requirements. It's better to just leave that as forbidden ground.




I think it'd be better to have:
TN4 (or 5. While I think TN4 is smoother and lets you have a lot more granularity, which I feel is better for the game overall, if there is a really strong sentiment in favor of TN5, I really don't care about it intensely, I just feel it's something that would be good)
-All actions have a threshold requirement. Easy actions may have a threshold of 1, but they all have a threshold that is defined.
-Rather than opposed tests, when something can oppose you, it rolls appropriate dice and adds hits to your threshold. This has the same net effect as an opposed test, but allows it to be layered onto a threshold higher than 1. For example a tough con might have a base threshold of 6, plus the target gets to roll their dice to add to that.
-Negative modifiers can increase the threshold for you to succeed. For example, shooting somebody may be a threshold 1 test. Shooting someone behind cover may be a threshold 3. Range may increase that to a threshold 4-6.
-Positive bonuses can be either dice pool mods or threshold reducers. Both of these should be restricted by skill in some way, with threshold mods being harder than dice pool mods. (So as an example if you have positive mods capped at skill, and 4 skill, you could have say 4 bonus dice, and 2 threshold reduction)

Like I said before, I feel strongly that a system combining totally fixed TN, variable Threshold, and *some* DP mods (mostly bonuses) would be a solid choice.

Variable TN is a source of significant confusion, has that classic '6=7' problem, and makes it hard to judge the value of DPs (messy synergy effects, -1 TN and +3 DP is very different from +4 DP). The smartlink and control rig effects suggested can be just as satisfyingly modeled with Threshold mods, I think (to a minimum of 1, which is fine; they help more on tricky stuff).

The proposed benefits of relying more on Threshold were mentioned a bit above. It all works with TN4 or 5 equally, so that's just a taste thing, I guess.

Well I guess it really depends. Are we just tossing out ideas here as a general sounding board, or are we actually trying to reach some sort of internal consensus to actually bother rewriting huge chunks of the rules?

Either one is fine, I enjoy these sorts of discussions for their own sakes. But if the intent was to create some proto-5e design for playtesting, some decision would need to be made one way or the other on several things that have been wishy washy and debated over with no real consensus throughout the thread.

Well, you can't do real work in a linear discussion thread. I thought we were brainstorming.

Actually i'm very happy with the edge system, they just need to be consistent about who gets it and how much they get and certain build types shouldn't get extra uses on top of the rest of their reality altering powers (uncapping hits).

If edge as an individual pool is supposed to be only for PC's, Prime Runners, and Free sprite and spirit equivalent that's fine and makes sense to me. Edge, even a little bit of it can be what sets you apart from the crowd. If it's something every spirit

I actually think trau might be on to something there with the use of VCR, i'm not sure if smart link alone should lower the threshhold though, I think smartlink is one of those things that works good in it's current implementation.

I also don't think melee should be balanced against firearms, just that it shouldn't automaticly be nerfed agaisnt them other then it's existing problems. The biggest change if I were to do a houserule would make melee combat a simple action instead of a complex.

As to my list i'd almost need tos tart a seperate thread to explain my reasoning.

Ditto: Edge is the 'hero factor', as far as I'm concerned. I agree that the various Edge tricks should be considered and dealt with; I see no great reason Edge needs to uncap hits at all, for example.

I think we might also want to review the 'exponential' nature of the Edge +dice (Edge 6 gets 6 uses of +6 = 36, while Edge 1 gets 1 use of +1 = 1). Perhaps frequency and power should be decoupled: either all points of Edge have the same effect (like EP's Moxie), or everyone gets one hit per session of +Edge, etc.

The smartlink is a setting/design choice, not a right/wrong answer. It used to be that smartlinks were a big deal, a transformative element of ranged combat. In SR4, it's a measly 2 dice, in pools that are often well above 12. A -1 Threshold would be 1 free auto-hit, which is obviously *roughly* 3 dice, without the cheese potential, and without reducing glitches. To me, this is a good balance between the weak SR4A smartlinks and the crazy-strong SR3 smartlinks.

Right now, we're just brainstorming. Any proto-SR5 would require someone to take the helm and make final decisions, since consensus never happens.

Another way to look at it: We're finding the problems with SR4, and looking for solutions without fear of breaking things because, conceptually, we can alter them as well.

It's kind of important to let all the little things bounce around together, because they all affect each other. Thresholds, TN, mod capping, skill rank uncapping, changing cyber/DNI, Edge tweaks/overhaul; all hinges together.

Well I think design goals should come first: How do you WANT the game to look, feel and play. Then you try to find (for simplicity's sake) a unified mechanic, and then tweak that for the specific cases involved.

For example: SR4 makes a total mess of how competency in the world translates into numbers. A total genius gets an average 1.33 hits more than an average joe, if only skill dice are counted. That's disappointing. (Well, White Wolf is probably even worse ).

What I would start with, in order to find the balancing points, the TN, and the system of threshold vs dice pool bonuses:

Define what constitutes a test, and define difficulties.

Find probability curves to make those tests with differnet DPs and TNs, and from there, pick a TN and corresponding DP range that fits the design goal.

Tune the ingame and chargen costs of aquiring those DPs so that you arrive at DPs that scale according to the design goals.

For instance:

An incompetent person should have an approximate 20% chance of managing a simple task (a threshold 1 test).

A semi-competent person,....
...

The god of cookery should have a 99% chance of making a simple test, and a 10% chance of making an obscene test.



So now pick a TN (4 or 5) and look at how probabilities scale with DPs for those tasks. Finally look at how to aquire those DPs in game. And I would be quite conservative/careful, there, because you absolutely KNOW that players will find ways to get bigger DPs than you think they should have. Again this is a design goal decision: Do you let bonuses stack by default, and disallow stacking in specific cases, or do it the other way around? Do you try to make everything cost the same, to discourage min-maxing and optimisation, or do you build in deliberate cost differences, to make some things clearly better than others, if only to keep building characters interesting.

Brainpiercing, yes. However, we're not designing a new-from-scratch, we're 'fixing' SR4A. So it's appropriate to have a general airing of grievances with SR4A, see what issues are common, and see which glom together. Both are important.

I do agree about sorting out the probability curves, and that stat+skill shouldn't be so overshadowed by DP mods. That's why I liked the idea earlier of capping bonuses using just skill; it rewards skill and allows those people to get the highest (I also think the cap should be pre-situation penalties, so that you can't have a bunch of virtual dice that automatically absorb all issues). The alternative to this is rigorously controlling stacking source-by-source, including in every future splat. :/ Hmm.

Personally, I wouldn't mind many uses of edge going away.
I like the concepts of spending edge before the roll to add dice and make 6's explode.
I like reducing the severity of a crit glitch.
I like hand of god.

But the ability to bend the rules of statistics to insure that virtually every roll is "above average" for their dice pool by rerolling dice that miss means that a player's actual dice pool is far and above what his dice pool says it is.
Using Smallroller (http://www.fnordistan.com/smallroller.html) it appears the % chance of getting 4 or more hits with 12 dice in a dice pool is 60.69%. That seems reasonable from a game balance perspective. A TR4 should be a "difficult" task.

With edge, it's not. It's so routine, that all someone is doing is determining if they use edge or not. It becomes "How much can I do before I need edge refreshed?".

If I get 3 hits (21.20%), and burn edge the chances of getting that extra hit is 97.40%. (Chance of failure = .5%)
If I get 2 hits (12.72%), and burn edge the chances of getting two more hits is 92.49% (Chance of failure = .9%)
If I get 1 hit (almost 4.62%), and burn edge the chances of getting three more hits is 76.59% (Chance of failure = 1%)
(chances of me getting 0 hits on 12 dice is about .7%)
Being able to reroll dice that missed with edge means that the chance of failure on a TR4 with 12 dice drops from 40% to under 4%.

(note, a 4% chance of failure on a TN 4 is basically a 21 dice pool)

*Cough* Fixed that for you.

No problemo. Thanks.

Ditto: I don't like or see the need for 'reroll misses'. I'm fine with 'full reroll'.

Well... I would tend to heartily disagree.

Edge is supposed to be just that: An edge. Rerolling is just as valuable to high-edge characters as it is to low edge characters. It is as valuable to that corpsec goon rerolling to perhaps not get blown up this turn, as it is to the PC rerolling for hits on the attack test. If the PC can afford to do that against a mook, good for him. Chances are, he'll miss that on a more important test. (IF the edge refresh mechanic isn't set up to dish out edge at every corner.)

With non-mook enemies, without group edge, and a five minute workday, it is very likely they could reroll quite a bit, possibly extending fights well outside of PC comfort zones.

On the other hand, re-rolling enables you to resist a very large amount of damage, or dodge a lot of enemy hits, and is perfectly in line with what karma pool used to be. I like that extra bit of safety, both as a GM, and as a player.

I would not give out full edge refreshes every session, though.

How often do people refresh Edge anyway? I get the feeling it's very different among groups.

For our group it usually depends on what we're doing. Generally it refreshes at the start of each scenario, which could be 2-3 sessions between refreshes, or it could be refreshed multiple times in the same session, if the session is running long and we're moving through things quickly.

What do people think of my earlier comment that Edge frequency and Edge power shouldn't increase together? Possible results would be that Edge is always non-scalar (that is, 1 Edge = 1 reroll, or 1 Exploding, etc, but never +X, as in Eclipse Phase), in which case luckier people just do it more often; or that Edge refreshes at a fast but *fixed* rate (e.g., everyone can Edge once per session, but Edge 6 gets a +6 for his). Obviously, there are consequences or anything like this that would need to be addressed.

This is another example of something that kind of hinges on other changes. If we're using 'karmagen everywhere', then Edge cost is already increasing, which largely addresses the Mr. Lucky issue in the first place.

I agree that it shouldn't, but at the same time I don't like either solution.

Just spitballing. Any ideas? I generally avoid increased complexity, but you *could* have 2 stats: one for Edge frequency, one for Edge magnitude. The person who only cared about rerolls, for example, would just leave their magnitude at 0 (or 1).

Personally I'd eliminate the extra dice option and just make it a reroll X times per edge. Keeps edge valuable, edge increases in power linearly.

That's the only thing I've come up with. But I don't like it either. :\

Hehe. Well, no one said it'd be easy. Let's recap:

• Edge (controllable luck/hero-factor) is something we want in the game.
• Edge's exponential power is a problem.
• Edge is overcomplicated/overloaded?

Fixes:
• Increasing (if not exponential) cost of Edge; possibly via 'karmagen-everywhere'.
• Non-scalar Edge/reduced Edge options (no +X).
• Rarer Edge (no X uses).
• Pay separately for Edge power, and Edge frequency.
• Other alterations to the refresh mechanic.

How do other RPGs that use a Luck stat handle this? Eclipse Phase's Moxie is a pool of auto-crit/reroll 'doses' (like the non-scalar suggestion). SR3's Karma Pool (which had its own problems) filled at a slow, predictable rate.

It would be interesting to see a standardized edge refresh (one of my biggest problems with shadowrun is how much a lot of it depends on the GM adjucation. See also earlier complaints on karma rewards, and timeframes for doing doing non-combat things), but standardizing it in any meaningful way would be really hard to do.

One off the cuff solution may be any critical success gives you your edge rating in dice that you add to a pool (this pool may or may not have a cap). The dice in this pool may be applied as you wish to any test.



The definition of critical success may need to be redefined. Instead of succeeding by X net hits, perhaps getting more net hits than is average for your dicepool. Say 1/2 of the dice pool for TN5 and 2/3 dice pool for TN4 (just offhand, could be adjusted upward if that's too easy). To avoid trying to roll lower dice pools frequently to get extra edge to use later, perhaps make a minimum dicepool and/or test threshold required to qualify for a critical success. (Say you need a dicepool of at least 9 dice to qualify for a crit success, and be attempting a test of at least threshold 3-4 depending on the TN)

I do think that with edge, either the number of dice you get each time or the number of times you can use edge should be a fixed number.

Fixed number of dice: When you spend a point of edge, you add 3 dice to your pool. You may spend a number of points per adventure equal to your Edge.

This makes edge always useful for adding dice, regardless of how high or low the character's edge stat is. It makes edge represent how much luck or hutzpa a character has over time, as opposed to how much they can bring to bear at one time.

Fixed number of uses: Every character starts the adventure with three edge points. When they spend edge to add dice, they add a number of dice equal to their edge stat.

Under this system, everyone can try to "give it their all" as much as anyone else, but what that means isn't equal. Some people, when they bring their will to bear, become unto gods. Others don't.

Another advantage of this method is that a GM can adjust how much edge is available per adventure, and can even give out edge points as rewards during the game, much as Savage Worlds does with bennies.

Karma Pool: Each character gets a number dice equal to their Edge x 3. They may use each die once, adding them to their rolls as they see fit.

This option is more old school and flexible.

What about the +Dice vs. Reroll (misc.) issue? Edge has many uses, some crappier than others, some more dependent on Edge power *or* frequency. (This is aimed at the Karma Pool variant, mostly).

I'd keep reroll if we're keeping close to the current edge system, and just get rid of the bonus dice option. If we're changing up the system to something more clearly defined, I'd remove the reroll option and make it a flat +dice option.

Well, at least for the nWoD you only need one success to achieve the task you want, if it is a harder task you roll with fewer dices if it's easier, you roll with more dices. 1 to success is just like any ordinary succes. When you get 5+ success than it is an extraordinary success and something extra awesome happens.

Well actually I don't think either two or three are accurate. The edge system does work right now. There are aberrations, but anything pushed to the max will have aberrations. In game design I think it's a poor strategy to take player choices away.