I'm currently running the game as GM while following all the events from SR Timeline. I have nearly the full collection of books (but yet not read them all ofc).

So as i'm into the Harlequin/Bug City arc, I play with SR3.

But the recent topic about Edge and Karma made me think that Karma Pool will soon snowball out of control.

Is there any house rule about it that is commonly considered as good? I would have gone for 1st point at 10, 2nd at 20, 3rd at 30 and so on, but it would be too harsh for metahumans if it was 1st at 20, 2nd at 40... and not balanced if it was 1st at 20, 2nd at 30.

Hard capping pools isn't an idea I like neither.


Maybe something like: 1 to 5pts every 10/20 karma
6-10pts: every 20/40 karma
and so on?


Thought?

is that not how it works ?

Humans get a point to start .. one at 10, 20, 30 etc.

Meta humans get one + one at 20, 40, 60

on average Karma is 6 per mission .. so ten missions to get 5 or 3 Karma Pool .

it seems fairly balanced unless you are handing out Karma like candy ?

Official rule=1 every ten
so:
10=>1
20=>2
30=>3

I was meaning
10=>1
30=>2
60=>3
100=>4

(which is clearly too slow).

It's not. It's around 6-7 for team karma. But you'd likely pack 2-3 personnal bonus points with (IMO) a good playin (Roleplaying-Smarts and motivation).

That makes an average of 10 for the good players. So ten missions makes the human karma pool at 11 (100karma=10 + 1 starting point AFAIR).

MY PC averages currently at 5-6 karma pool but I still do not see them as experienced characters.

so the next question is how often do you refresh the pool ?

and also, if you spend a pool to re-roll failures, you can do it again on same roll for double the KP costs so a PC can spend it like candy to survive etc.

just break out the trolls with FA Assault cannons and watch them burn Karma for fun

Limit Max. Karma Pool to 10.

Bye
Thanee

I just don't like that. Run Harelquin + maybe a few little additionnal adventures, and you reached the cap.

The feeling of increasing your power is cut as soon as you reach the cap.

I agree with you, if you cap it .. no one will play humans as that is the main benefit compared to Stats on a Meta.

the only answer without capping is house rule diminishing returns post 100 Karma

or

Make sure the PC's keep recycling KP. you just need to make them work harder for the Karma gains

vs. better geared / better prepared opponents that dictate PC's using KP to be "more" effective.

In the end it's not how much or how little you have, it's what you do with it.

If you do not want to cap it, I would also go with diminishing returns.

i.e. first 5 points as normal.

10/20/30/40/50 or 20/40/60/80/100

then double the intervals, then triple, quadruple, etc.

70/90/110/130/150 or 140/180/220/260/300
180/210/240/270/300 or 360/420/480/540/600
and so on...

If you want higher Karma Pools, you could do this every 10 instead of every 5, of course.

Bye
Thanee

A previous discussion of the topic.

The disparities created by metahumanity and the Bad Karma flaw throw a serious spanner in the works—consider that at 200 earned karma a Human is looking at 21 karma pool, a metahuman or Bad Karma human at 11, and a Bad Karma metahuman at just 6, allowing failure rerolls on a single test six times, four times, and three times respectively, a massive difference—and for this reason we at SR3R have focused on eliminating the karma pool difference in balancing efforts; essentially, Bad Karma (and to a lesser extent for some builds being a metahuman rather than human) constitutes a bet that the campaign will end before large amounts of karma have been awarded. For a game already in progress I'm not sure how best to handle this, but it seems like the single biggest issue you're likely to face.

Once that's addressed, though, my feeling is that large karma pools are essential to the "high-level" game. There are basically three approaches to mitigating risk in SR3. Reliance on GM leniency and an assumption that if the team gets into serious trouble the GM will scale back the opposition is one approach, but one with obvious problems—beyond creating an incoherent world, it also stacks the deck, removing a lot of player accomplishment and value of making good decisions. Another approach is careful planning; this should be encouraged to some extent, but the nature of Shadowruns is such that there are often planning chokepoints—if the players couldn't simply blast their way through the run guns blazing, the implication is that there's some portion of the run during which any serious failure that raises an alarm hoses the entire run. As run difficulty and complexity increases, the amount of planning required to handle this explodes—I've been in groups in which months on end of weekly sessions have been consumed almost entirely by planning with a little legwork here and there to support it.

The third way of mitigating risk is through karma pool. It's essentially the only expendable resource that doesn't have to be spent in advance that allows a team to recover from serious failure. As an example, take a run in which a door protected by a keypad with anti-tamper system must be successfully broken by the B/E specialist to gain access to a facility—with Electronics 6 and Electronics (B/E) 6, facing effective TNs of 6 for both tests, the B/E specialist is looking at ~66.5% success rate for each part, or a ~44.2% chance of success on both. The problem is, both halves of the test trigger an alarm on failure—even if the odds were better, a 20% chance of hosing the run could well be unacceptable, and with those TNs it would take 12 dice on both halves of the test to reach even ~78.8% success rate. Without karma pool this plan is a non-starter, pure and simple.

This is where karma pool comes in. Rather than needing a plan in which each and every critical test has an overwhelming chance of success, the team can instead find a plan in which the critical tests are likely to force expenditure of karma pool at a sufficiently slow rate that it doesn't run out (possibly plus a safety margin). Planning still has value insofar as it can reduce the expected karma pool expenditure, but the team no longer needs to create nearly-perfect conditions throughout the entire critical portion of the run. In the keypad example above, the ability to reroll failures means that the B/E specialist effectively has 12 dice on each test—and if they roll a success in the first 6, they don't need to spend karma pool. The stakes on each test individually are lowered by the karma pool safety net, while success or failure on the run as a whole remains well within the power of the group to affect via quality of planning.

So yeah. My advice is to focus on the issues of karma pools of widely differing magnitudes, and don't sweat the absolute size of those pools.

Edit: that said, refresh policy can make a big difference. If you refresh karma pool every session, I could easily see large karma pools being spent very aggressively in (e.g.) combat, as the rest of the pool only needs to last until the end of the session and thus has fairly little conservation value; on the flip side, if you have large amounts of action between refreshes, the risk of not having enough later essentially overwhelms all but the most urgent uses in the moment. My advice is, broadly, to refresh whenever the group has clear downtime—if you don't use the training time rules, refreshing whenever you'd ordinarily allow them to spend karma seems like a good guideline.

~J

The use of a Team Karma Pool helps balance the meta's out quite a bit if the humans are willing to chip in a few points.

If high Karma pools are causing the adventures to be too easy, then slow down the refresh rate. When my team went through it, the karma pool didn't refresh until the quest was over and that made it really questionable on when you wanted to pull that point or not. If they are too hard, increase the refresh rate.

If I were going to make a list of when to refresh it'd look like this:

1) Do something 'good' or selfless, at risk: 1 point
2) Do something 'evil' or selfish, to the detriment of another: 1 point (Good Karma is something else!)
3) Survive a combat encounter, with risk: 1 point (not sneaking up and knocking the guard out, but an actual exchange of attacks with deadly intent)
4) Get a good night's (day's) sleep: 1 point
5) Do something unexpectedly impressive, risky, or cinematic without the expenditure of Karma Pool: 1 point (this could be virtually anything, a know-it-when-you-see-it qualification -- dropping the troll gang boss with a single shot, leaping from the roof of the van onto the go-gang's beat up Chevy at highway speeds, etc)
6) Complete a job: refresh pool
**7) Doing something in-character even when the player out-of-character knows it would be better not to: ? points

That would mean that Karma Pool could be a general in-play reward for being smart, or ballsy, or knowing when to lie low for awhile, or getting the job done. Or just being a good roleplayer.

House rules for my table are:

double the cost each time
so karma pool is at 10/20/40/80/160/320 karma
Meta-humans starting at 20

I also allow my teams to do group karma.

and should they wish to reroll a failure a second time it increases an additional point to the karma pool cost
first re-roll 1 point
second 2(3 total expended)
third 3(6 total expended)

I allow a refresh at the end of a campaign or one off but I'll have to consider some of the other suggestions provided

Isn't that the normal rule?

Bye
Thanee

It is, yes.

Just to throw in, we always did a diminishing return on KP as well. Can't recall exactly the break points though.

Been awhile since I've gone into that section of the book.

In my last SR3 campaign I decided to cap Karma Pool at 10. Later I also considered changing it to an attribute that would have to be increased by spending Karma (pretty much backporting Edge to SR3, except that you wouldn't start out higher than 1 for metas or 2 for norms), but by then the campaign had mostly fizzled.

I refresh my PC pool each scenario. And I do not lighten the difficulty for my PC. Tbh, SR is my "elite" game. Very gritty. For exemple, I had a mage of the group that stupidely went invisible to reach an NPC shooting behind a cover. He wanted to shoot him from short distance but that led him to stand into 3 other shooters LOS with no cover at all. And he dropped invisibility to maximize damage.

I did shoot him till down. Hand of god is my safety-option for a character not to die.

Maybe I want it too gritty? I tend to find a 10 karma pool beeing a bit overpowered but I might also see that in a wrong way.

Kagetenshi, what do you think is "High-level" in SR?

Overpowered in what way? Is there something in particular that characters are doing with those 10 points of KP that you find overpowered?


"High-level" comes from two directions: high-level runs and high-level PCs. The runs are easier to define, and I would do so in terms of exposure to risk. A team has to be able to either reliably succeed at the things they're going to need to do or have the ability to try again (either directly or by avoiding catastrophic failure and trying a different approach); a team with little Karma Pool lacks the ability to recover from failure by simply rerolling, so they need to run against adversaries with big holes in their security. If the B&E specialist might take a few tries to crack the door, the decker needs to be able to shut down the alarm; if the alarm can't be shut down or the decker isn't highly likely to succeed at it, the team needs to be able to shoot its way through security long enough to complete the objective and get out.

It basically comes down to combinatorics. In general the team needs to have a good chance to complete a run—setting the players up to fail requires great skill to make fun. Run failure should in general be traceable to one of two things: a clear bad decision (or combination of mediocre decisions) by the players, or truly exceptionally bad luck. If the players can make good decisions, get kinda bad rolls but not awful ones, and still end up hosed, it's very hard to feel a sense of agency—that the characters are more than driftwood tossed on a sea of chance.

That means that if the run includes a sequence of things that all need to go off right, those things all need to have an excellent chance of going off right. If a run involves, say, five tests, each of which individually has a 90% chance of success, and each of which has to go right to not fail the run, you're still only looking at a ~59% chance of the run succeeding. That's way too random to feel fun IMO, at least unless there's some relatively clear opportunity to improve their odds that they passed up. If they feel like they exhausted their opportunities for legwork and this was still the best they could do, that's just not fun.

Now, the straightforward way to address this is to let that sequence end up having excellent odds. The issue here, though, is that unless that involves a lot of player cleverness in planning it can end up feeling pretty unfulfilling; getting odds of per-task success up to ~98% gives a ~90% chance of overall success, but that means that there has to be a lot of excitement in the process of getting the odds there because there isn't going to be much in actually making those tests.

As I mentioned, the other approach is alternate paths to victory. Say you've got the same run above, with five tests at 90% chance of success each. If instead of the team failing once any step goes wrong they instead can try to shoot their way to victory, and they have a realistic chance of doing that (say, ~70%—not nearly good enough for a primary approach but still not bad), all of a sudden you're up to ~87.7% success rate. So on and so forth.

IMO, what makes a run "high-level" is where, barring truly exceptional planning by the players, karma pool becomes an integral part of this risk management. You can start reducing the number of paths to victory or increasing the difficulty of the paths, because the players now always have access to a limited number of alternate paths to victory—they can either succeed at a test, or they can fail, spend karma, and then succeed. If I'm doing the math right, with a point of karma pool available on each test in the above five-test run, you're suddenly looking at ~95% success rate. Unlike the easier 98%-individual-success version, though, this is no longer free—each failure costs a finite resource, KP. You can pair this with a difficult alternate path, say shooting your way through with a 40% chance of success, for a total ~97% success rate, but despite making things easier this feels like raising the stakes—if they don't make it through they face either run failure or a battle they could easily lose.

Now, of course, few runs are amenable to such precise calculation, and the numbers for the combat option especially are bogus—but overall it's pretty effective to precisely calculate a few things on the "critical path" and then throw in other options that are just generally "easy", "hard", or "almost impossible".

Now we get to what "high level" means for characters. This one's more difficult, because one of the strengths of SR3 is how much agency it gives the players. Obviously having a store of karma pool sufficient to plan on spending it without eliminating the emergency reserve is critical, but players who take big risks in combat can need essentially arbitrary amounts of emergency karma pool to not die; similarly, depending on how well-built the character is, two characters with the same earned cash and karma can have very different chances of success (as one example, consider two B&E characters, one of whom is designed by a more casual player who doesn't take Microscopic Vision and thus is looking at TNs 2 higher for many important tasks). Basically, it's a well-designed character played by a skilled player that also has a sizable karma pool.


This points to a concern that I had not been thinking in terms of. A player who is not given to planning or who takes impulsive risks may well accumulate a sizable karma pool without necessarily being able to handle a "high-level" run; unless the player can be guided towards better play, either by the GM or by fellow players, it at first glance appears that it may be difficult to set up runs that this player could accomplish that involve management of karma pool as a resource. If the player level of planning forces runs that look like "low-level runs" even for pool-rich characters, the karma pool does indeed sort of hang out there and… I'm not convinced, but at least possibly end up mostly getting used to do ridiculous things and still succeed.

I don't know, I'm not actually certain that there's a problem (again, please clarify on what you've been feeling is overpowered about 10-KP characters), but I'm going to have to give this case more thought.

~J

I've done a few things to level out Karma Pool (either intentionally or not);

1) Karma pool refreshes at the end of the *adventure* (since I run online) rather than the gaming session. That makes it more valuable.
2) Karma pool can be spent as normal karma. At higher levels, the humans' karma pool advancement may become more of a liability. He's got 5% less spendable karma than the metahuman. This makes his karma pool more useful to him. However, I've been fortunate in not having any players who just use karma pool to blow through every challenge.
3) Throw more challenges that require burning of karma pool. I generally permit more than one Hand of God uses if the character has scads of KP the second time around.

If your issue is one character who has tons of KP and is spending it on every test to give him a serious advantage, consider comparing #3 and #2. Say when characters permanently burn KP, they get that back as normal karma, so it's sort of incentivizing them to do that. Then give plenty of opportunities to do that.

To answer Katengenshi, I ll first describe how I run the game:

A- If the PC runs a good plan, and no roll derails, it works, even if I could have added some cinematic action. I understand the value of both approaches. But I've picked up the first.

Exemple: PC has to steal in Sylvan's corp the hand written original book from Ehran.

They came up with a good plan. They asked me who did clean up the building. I thought it was a rather small compagny so they payed a compagny for the clean up. So the player decided to take their places, intercepted them and faked tkeir access card. I don't remember how, but they fooled sylvan's security (either they fake-mailed them to inform them form the change of employee, or they used a mask spell to take their appearence). Nonetheless, they did all well and I didn't do any deus-ex-machina in order to add a little action. It just succeded.

B-I nearly never lighten the difficulty. I sometime think: well let's see if the PC will do well.

Exemple: Mercurial's ending: due to job, the physical adept couldn't play and 1 character had already Deadly damage. So just 1 mage 1 decker left to save Maria versus:
-3 teams of 4 Aztech's guard + 1 paranormal critter with each group
-Kyle Morgan + Perianwyr
-A killer
-A techie who'se wife has been already killed by aztech without telling him. An orders were given that at the end of the mission, he should die.


No way they can enter like the scenario describes it (sneak+gunshot). I do tell myself: let's see how they'll do.

a few weeks passes in between, the physical adept is back. They have all the elements of Aztech dooings but no proof. They decide to try to negociate with Aztech but change their minds. They found the info on the techie and make of him an ally. All camera's flow are derived on the decker's deck but the decker decide to already stream them on public network.

He pays a contact from Aztech to get Seattle security director's number, and calls him. Negociation starts. The PC tells him they know all what Aztech did and that they capture Maria. it's already on stream.

Then I make a decision. The security director know that if he negociate now, he acknowledge what the characters says. If they picked someone else, I would have rolled Intelligence for the NPC to see if he's tought of that. So he doesn't acknowledge and says he don't know who are those terrorists and cutt the call.

He calls Morgan to inform him that they're streamed and so that he contacts the PC to negociate. The PC can't know what the security director said.

Morgans goes to the techie. He looks at him. (remember the PC sees the scene through the cameras) and ****

I roll openly for the techie and Morgan. Techie loose. In his scared eyes, Morgan sees a traitor, and kills it. Horrified face from the PC. From angerness, they make the full dossier public. (while a negociation was still possible).


Now I pause to think a bit further. I *wrongly* said it was Aztech security that was near Morgan. It didn't seemed important at this time (previous session) BUT NOW it makes the hell of a difference. If caught, it directely points at Aztech. So I decide they leave. Kyle and Pery were reluctant in using BTL on Maria according to the scenario so I decide they leave the dirty work to the killer. They leave.

To me at this moment, Maria is done.

The mage and adept tries to reach the building, sneaking their way (if failure, 3 squads shoots at them and I swear, they would have done it, so it's really intense). To the decker, I tell him that Jorge (the killer) is about to use a simsense device with BTL on Maria. I telled him more to emphasys on Maria's downfall but he says: I look at a slave that would shut down electricity. I answer: 'kay, roll a d6. If high it's good (not always can electricity be handeled through matrix but it's not always impossible). Intense again: he rolls high! And then he shut it down, saving Maria.

In the end, they get Jorge and save Maria and deal with Aztech to limit the Megacorp's wrath.

By overpowered, I mean this:

If, in average, saving karma for crucial rolls implies saving something like 5-6 karma, that's 10 more (15 karma seems ok for a semi-experienced character) for fights.

In my game, there's not that much fights. This implies that the major threats can be taken down quick if you maximize your karma use in the primary blows.

Slightely OP to me.

Then about risk management, take in account the fact that I'm usually able to improvise a way for the PC to recoverfrom a failed test/plan if they planed well (see Mercurial's exemple).

I don't want it at high level to be Ocean's Eleven.

Mercurial's exemple was derailing and heading to a failure. Hence each moment went electric. AND they did it nonetheless in the end.


Now reroll anything to a success because you have 40 Karma pool....Any half decent plan can work...




Other thing that scares me. How do you feel if hand of god make you lose 40 karma pool?

Actually the way Karma Pool is designed isn't half bad, because there are diminishing returns. Double or triple reroll a difficult test a couple times and you're out of Karma Pool. Botch an open test and it's 6 Karma Pool to reroll just 3 of the dice.

What I do like is the idea of throwing burned Karma Pool back into the Good Karma. The scroll-hoarding mentality from D&D (keeping expendable items indefinitely instead of using them when appropriate) seeps into any RPG where you have to permanently expend a resource for a temporary benefit. I also like the idea of being able to spend Karma Pool as Good Karma if you really want (though any RPGer worth his salt knows it's better to be lucky than good).

Another thing that would probably be nice would be to assign a particular number of temporarily expended Karma Pool to buying successes or lowering difficulty, rather than forcing the permanent burn. Something like 5 Karma Pool for a success or a -1 difficulty.

I think that any Karma Pool under about 20 or so is probably manageable. Average difficulty (6 and under) tests might get a lot of rerolls, but once you have that much experience these average rolls shouldn't be tripping up the players anyway. If they save them for the truly difficult rolls, diminishing returns kicks in and they'll have to spend a lot of Karma Pool to get a noticeable effect.

Total thread necromancy, but I know you're still around and this is an important topic that I just dropped on the floor:


It isn't so much a matter of adding difficulties if things are going too well, it's about how complex the run is from the beginning. That said, published modules do suffer from a relative lack of scaling—the Sylvan run just isn't that complex as written, so if you want to play them as written I could see that being another case where high KP could legitimately be unbalancing. On the other hand, the same could be said about high attribute or skill levels from high earned karma.


The thing is, the amount of KP saved for crucial rolls is a lever that the GM can push on—by increasing the number of expected crucial rolls, or even the amount of uncertainty as to the number of expected crucial rolls, you increase the amount of KP that a group is likely to save. Granted, though, this is sensitive to your players' risk aversion and tendency to plan carefully (as noted above).

That said:

My solution does involve adding moving parts to the run, which could be at least partly what you mean by being Ocean's Eleven. On the flip side, though, how do you address the attributes, skills, spells, and foci that 390 karma buys? Because of the way difficulty scales with TN, it seems like you need to add moving parts anyway on account of not being able to just crank single-test difficulty.


This, however, is an issue. As I've said above, I find the issue with KP to be not so much absolute size but differences in relative size—and burning KP causes just that, nothing more spectacularly than HoG. I don't really have a solution to that.

~J

I really need to reraed the full thread

It made me think:

buying a dice is usually weaker than re-rolling failure.

Add to that that additionnal dices are bought at *2 Karma points. Suppressing the multiplier would make the option more attractive but does someone have an counter argument?

My initial reaction is to concur. It lets you hit thresholds you couldn't otherwise, and there might be some edge case hiding somewhere where that's important, but spending KP is a pretty high price to pay.

~J

I've doubled the cost after 8 karma pool and it seems to be working fine. But I also 'give' them it for free when they're supposed to get it. Instead of earning 10 karma, 9 going to good and 1 to pool, I just say every 10 karma your pool goes up by 1. The humans use their karma more often, and burn it when they need to, while the meta's use it less often and have yet to burn any of it. the team is all at <100 karma atm as well.

A) Kage, stop being a necromancer. No one likes Necromancers.

b) The small group I have been running on and off for... 12 years? Something like that... has 1 human, an orc, and a dwarf. IIRC, the orc has the largest amount of good karma at 160 something, a karma pool of 9. The human has a pool of 14. All that has changed from a GM stand point is that I can fuck with them more, and not have to worry as much about accidentally PTking them.
It encourages them to take an extra risk once in a while, but at this point they are half way to being Prime runners, and should have the ability to take a risk and come out on top without me fudging dice for them.

Okay, I ran a game for about four years with the same characters. By the end of it, I recall that they were all somewhere around 150-200 karma. So, it was a high-powered game.

Fortunately, I had set a staggered-rate gain for karma pool. I can't recall the exact break points, but it did go up at regular intervals. Because of that, karma pools were manageable. They still were fairly large, but since the challenge level was also up there, it wasn't unbalancing. I also refreshed every session/story break; basically, any significant downtime would refresh karma pool.

What I'm getting at is, there's plenty of reports here (including mine) that say staggered-rate karma pools (aka diminishing returns) works wonderfully. As long as the challenge level increases to match karma pool and other improvements, it's not unbalancing or overpowered.

To some extent that's no longer the question. I mean, I do think it gets tricky to handle high skills/attributes with low karma pools, but since the game works at low/moderate KP levels it will presumably broadly work at those levels even if they take longer to reach. My argument is that staggering is unnecessary—there's already diminishing returns at the point of use, and the returns that don't diminish (freedom of use, total number of uses on tests with reasonable chance of success) enable and are largely counteracted by more complicated runs.

Though I must admit that if I ignore whether or not KP causes problems and just look at how quickly it accumulates relative to the karma cost of high skills or attributes, it is a bit odd. I think it's important to gain KP quickly at the beginning, and the most reasonable apparent fix would be to have a two-tiered formula (rather than a steadily-increasing one)—but this ignores the IMO crucial issue of whether this "fix" does anything useful, not to mention whether it does anything harmful.

(For that matter, the "oddness" is relying on some very dangerous intuition. Most average-strength people would probably have an easier time gaining strength than learning how to use throwing weapons. For that matter, it's also probably a lot easier to learn how to use clubs acceptably than it is to learn Heavy Weapons (never mind that Heavy Weapons includes the disparate categories of machine guns and assault cannons), and both strike me as easier to learn (at least for someone with no preexisting combat skills) than Underwater Combat.)

But yeah. I think we've established that diminishing returns provides acceptable results—the question now is if they provide benefit over canon accumulation (modulo issues about KP disparities).

Edit: one thing that comes to mind—if you really don't want to make your runs elaborate enough that players are managing and conserving their Karma Pool, one solution might be to change the refresh mechanics—players only have access to some portion of their Karma Pool at a time, and there are "sub-refresh" events that might restore some spent karma, but also allow players to replenish their accessible KP from their reserve. Not terribly thought-through, just an idea that popped into my head.

~J

I don't know about that. My experience is, three rerolls on one task is pretty powerful. But even then, it's more that at that point, they could make six different rerolls instead, which is potentially even more powerful. Since karma pools are a measure of experience, player characters will not only have more rerolls, they'll have more dice to reroll, making it a really potent combination.

Larger dice pools aren't a problem by themselves, but it does mean players are less likely to fail. But when combined with a lot of karma pool, they're even less likely to fail, which makes challenging them even harder. Since I don't like escalating things, I personally find it preferable to prevent the problem from cropping up in the first place.