Full Version: Shadowrun 3rd Revised
| QUOTE (nezumi) |
| is there a need for an 'accidental shooting' mechanic? |
I seem to recall a subtle line in the rules that you can do as much damage to yourself as you want with a single use of any weapon. There is also the part in sacrificing about being able to do a controlled amount of damage (regardless of dikote on the knife) to an opponent that is unable or unwilling to resist.
This leads to a fairly simple concept. Any weapon can deal any amount of damage under proper circumstances. For simplicity those circumstances can be listed as: When it doesn't involve a PC, When the victim does not resist, When the victim cannot resist, and more that I think I am missing.
On the accidental shootings issue, all the ones I can remember hearing about had to do with kids playing with a loaded gun much like they would play with a NERF gun (pointing it at the head/torso of the other) or Darwin Award candidates who shot themselves in the head due to some sort of complete ignorance.
In both of those cases, the attack can easily be described as a called shot with aiming against an unarmored opponent that isn't a troll. This would put a light pistol at M or S even if the children weren't defaulting from their NERF Pistol specializartion and were going straight from Quickness.
As a last question, how often did the children involved in accidental shootings immediately call for medical attention?
| QUOTE (Herald of Verjigorm) | ||
I seem to recall a subtle line in the rules that you can do as much damage to yourself as you want with a single use of any weapon. There is also the part in sacrificing about being able to do a controlled amount of damage (regardless of dikote on the knife) to an opponent that is unable or unwilling to resist. This leads to a fairly simple concept. Any weapon can deal any amount of damage under proper circumstances. For simplicity those circumstances can be listed as: When it doesn't involve a PC, When the victim does not resist, When the victim cannot resist, and more that I think I am missing. |
Eh, it still rings hollow to me. Not so much the accidental shooting "rule", that seems to work just fine, but the idea that it's impossible for someone to get more successes than their total dice thrown. It would mean, for example, that someone with 3 or less Int would be incapable of positively identifying anything, even in full view in a brightly lit room at point blank range (4+ successes required on Perception).
It's one of the biggest faults I have with the SR4 rules too, that there are things that are by the rules impossible, even when logic, historical example and even common sense prove otherwise.
| QUOTE (tisoz) | ||
Just so long as it maintains the connotation that Shadowrun is for people who like to sling a lot of dice. I like the handful/fistful as it intuits the dice you are rolling for that specific test no matter their origin. |
Maybe we can rob the d20 guys and call it "hit dice".
Perhaps an easy mechanic would be, for every number the Target Number falls below two, that counts as an automatic success? So a person at point blank in a brightly lit room would have say 2 or 3 automatic successes. Related, a gunshot at point blank against a non-moving target would perhaps have 2 automatic successes.
This avoids failures at ridiculously easy tasks (such as tying your shoes), and increases the threat of accidents and mooks, and allows for smart runners to better take advantage of one-shot kills (you have a rifle trained on a stationary target at moderate range? He SHOULD be dead.)
This avoids failures at ridiculously easy tasks (such as tying your shoes), and increases the threat of accidents and mooks, and allows for smart runners to better take advantage of one-shot kills (you have a rifle trained on a stationary target at moderate range? He SHOULD be dead.)
Then we'd have to completely rework VCRs, because my TN doesn't hit 2 except on a very, very bad day.
~J
~J
I was under the assumption we were going to rework VCRs anyhow. I'm not especially pleased with any piece of cyber that means you never have a TN above 3 for an entire class of actions.
Heh, true. Maybe reworking Smartlinks should be in the works as well; SR3 games pretty much have to suffer from film noir syndrome just to keep sammies from TN-2 ing everything in the game. Maybe they should they just add dice equal to their rating to all driving tests instead, along with giving access to Control Pool?
My idea was that TNs could just go below 2, even into the negative levels. So if the TN should ever get to, say, -5, even rolling a 3 would count as two successes. Note that 1s are still always failures, and the Rule of 1 would still apply, so you'd have to roll anyway, just to see if you botch.
My idea was that TNs could just go below 2, even into the negative levels. So if the TN should ever get to, say, -5, even rolling a 3 would count as two successes. Note that 1s are still always failures, and the Rule of 1 would still apply, so you'd have to roll anyway, just to see if you botch.
By that logic, you'd have to rework more than just Smartlinks - keeping things dark doesn't work either, if the Sam was smart about his eye augmentations.
Between thermal, low-light, eyelights, flare-comp (and ultrasonic, maybe, if you have Essence and cash to spare) there are very few situations in which one will get any vision-related penalties - and most of those are not something you can reasonably use all the time.
It's not like you can run session after session in complete darkness filled with thermal smoke. On top of that, anything you do only ends up hurting everyone else who doesn't have the same augmentations more than it does the augmented guy.
However, I don't think the Smartlink situation and the VCR one are remotely comparable. -2 TN does not compare to -6 TN and making most your rolls vs. the handling of your vehicle, which you can lower through customization, for an effective -8 or -10 TN.
In addition, if you run your gun battles using all the tools available to you as the GM, you'll find plenty of ways to raise the shooter's TN above 2. Cover, target running, shooter walking, shoother running - those should come up consistently (assuming your gunfights don't happen with everyone strolling about in the open) and they'll drive the TNs up very quickly.
Between thermal, low-light, eyelights, flare-comp (and ultrasonic, maybe, if you have Essence and cash to spare) there are very few situations in which one will get any vision-related penalties - and most of those are not something you can reasonably use all the time.
It's not like you can run session after session in complete darkness filled with thermal smoke. On top of that, anything you do only ends up hurting everyone else who doesn't have the same augmentations more than it does the augmented guy.
However, I don't think the Smartlink situation and the VCR one are remotely comparable. -2 TN does not compare to -6 TN and making most your rolls vs. the handling of your vehicle, which you can lower through customization, for an effective -8 or -10 TN.
In addition, if you run your gun battles using all the tools available to you as the GM, you'll find plenty of ways to raise the shooter's TN above 2. Cover, target running, shooter walking, shoother running - those should come up consistently (assuming your gunfights don't happen with everyone strolling about in the open) and they'll drive the TNs up very quickly.
I've never had significant difficulty with the smartlink. Shadowrunners, being characters who generally work in the shadows, are expected to intentionally look for dark and mist (and while eye augs reduce those penalties, generally only natural vision completely eliminates them). Like mmu said, add in cover, movement, range, etc. and on average I find shooting TNs are around 5 (7 for shooters without SL).
Eyeless blond, I don't understand your explanation. If the TN is -5, by your explanation, I'd assume you're guaranteed no successes, but likely to get 3 (with 3 dice). If 1 is still a failure, than you've changed absolutely nothing, you still have to roll a 2 or above on each die for it to be worth anything. My idea would mean that a -5 TN would give you 6 automatic successes (or perhaps we should just convert the negative number directly; a little less adding, make the rule a little less powerful).
Eyeless blond, I don't understand your explanation. If the TN is -5, by your explanation, I'd assume you're guaranteed no successes, but likely to get 3 (with 3 dice). If 1 is still a failure, than you've changed absolutely nothing, you still have to roll a 2 or above on each die for it to be worth anything. My idea would mean that a -5 TN would give you 6 automatic successes (or perhaps we should just convert the negative number directly; a little less adding, make the rule a little less powerful).
-2 for smartlinks just really make sense.... TN 2 is only for targets that don't have cover, have no lighting mods. all you are doing is trying to hit the side of a proverbial barn.... should be easy for seasoned professionals. This also assumes that you are resting, not moving over any kind of ground, etc.
maybe the -1 for stationary should be removed, and assume that everything is stationary,.... and if they move, you suffer a penaltly. +1 if you are moving. +2 for running. Most people forget the attacker walking/running modifiers.
the -6 from a VCR kind of makes sense, but I think it should be changed a little bit.
difficult maneuver TN's can get modded up really quickly. maybe the mod should be just level instead of level *2. with a VCR you actually become the vehicle, you feel, everything and when you twitch the vehicle twitches so it makes sense for all driving tests.
2. I think the augmentation that a vcr provides for firing weaponry should come from a targeting computer either in the vehicle or on your vcr rig. Wired reflexes don't lower your tn's why should a vehicle rig. Your weapons are part of you and you fire them much like you would a cyberweapon, so by that comparison, a targeting computer could take the place of a smartlink.
maybe the -1 for stationary should be removed, and assume that everything is stationary,.... and if they move, you suffer a penaltly. +1 if you are moving. +2 for running. Most people forget the attacker walking/running modifiers.
the -6 from a VCR kind of makes sense, but I think it should be changed a little bit.
difficult maneuver TN's can get modded up really quickly. maybe the mod should be just level instead of level *2. with a VCR you actually become the vehicle, you feel, everything and when you twitch the vehicle twitches so it makes sense for all driving tests.
2. I think the augmentation that a vcr provides for firing weaponry should come from a targeting computer either in the vehicle or on your vcr rig. Wired reflexes don't lower your tn's why should a vehicle rig. Your weapons are part of you and you fire them much like you would a cyberweapon, so by that comparison, a targeting computer could take the place of a smartlink.
What augmentation does a VCR provide for firing weaponry?
~J
~J
| QUOTE (nezumi @ Mar 23 2007, 04:45 AM) |
| Eyeless blond, I don't understand your explanation. If the TN is -5, by your explanation, I'd assume you're guaranteed no successes, but likely to get 3 (with 3 dice). If 1 is still a failure, than you've changed absolutely nothing, you still have to roll a 2 or above on each die for it to be worth anything. My idea would mean that a -5 TN would give you 6 automatic successes (or perhaps we should just convert the negative number directly; a little less adding, make the rule a little less powerful). |
Well I was just extrapolating from the rule a few dozen pages ago, that if a single die's result beats the Tn by X--let's say 9--you get another success. So, say the TN gets bumped down to -5. You roll 2 dice, and get 1 and 5. The 1 is an auto-fail, but the 7 not only beat the TN, but did so by 10 points, so it counts as 2 successes.
Note that rolling two 1s would still fail.
And sure, Smartlinks are fine I guess.
...not sure where best to post this question, but will there be a related thread on Bioware, Chemtech, and Genetech?
Been trying to play catch up on all the SR3R topics. Still have a long way to go. Somewhere along the line it would be handy to have a compilation of what changes/additions have already been agreed upon.
Been trying to play catch up on all the SR3R topics. Still have a long way to go. Somewhere along the line it would be handy to have a compilation of what changes/additions have already been agreed upon.
I imagine there will eventually be threads on just about every topic in the Shadowrun system. Questions, comments, and proposals on not-yet-threaded topics should go here.
I'm trying to keep decided changes at the top of the relevant threads, but at some point (probably over the summer) I'll see what I can do about making a more centralized repository of all changes.
~J
I'm trying to keep decided changes at the top of the relevant threads, but at some point (probably over the summer) I'll see what I can do about making a more centralized repository of all changes.
~J
...thanks for the clarification. Like I mentioned I am doing my best to get up to date so as not to reiterate concerns that have already been addressed or settled on.
I am tooling up to run a revised version of an SR3 campaign I did a couple of years ago and would like to bring it in line with SR3R (and eventually SOTA2065) as much as possible.
I am tooling up to run a revised version of an SR3 campaign I did a couple of years ago and would like to bring it in line with SR3R (and eventually SOTA2065) as much as possible.
| QUOTE (Kyoto Kid) |
| ...not sure where best to post this question, but will there be a related thread on Bioware, Chemtech, and Genetech? Been trying to play catch up on all the SR3R topics. Still have a long way to go. Somewhere along the line it would be handy to have a compilation of what changes/additions have already been agreed upon. |
I do try to take notes on everything recommended, agreed upon and tabled, which I post regularly. If you go back a few pages, you may find the most recent one from me. It's generally huge, so hard to miss. Unfortunately, the past few weeks I've been falling behind due to my wife having her wisdom teeth removed and a bunch of other weird issues, but I hope to catch up again soon.
So, open questions on this thread:
- Defining and clarifying game terminology: Open. There seems to be lukewarm support for this one, though debate continues over what to call the number of dice used in a test.
- Naming thresholds as part of the core mechanic: Hotly debated. Some people see Thresholds as being everywhere, and that they need to be labeled as such. Others don't see Thresholds as being all that prevalent in SR3, so they don't deserve their own general rules.
- Giving thresholds the same honored spot in the SR3 core rules as TNs by giving them extensibility guidelines as TNs have, and specifically spelling out where TN and Threshold apply: Closed(REJECTED). This one surprised me, as I currently see thresholds as already heavily integrated into the rules, just not in a way that is really spelled out, but vehement opposition to further integrating Thresholds is forcing me to back down.
- Very high rolls --> extra successes. Open. Kagetenshi and Herald have both expressed dislike for this idea, though so far no concrete arguments have come up against it. On the for side there is disagreement on how to implement it: how much over the TN counts as an extra success? Should TNs be allowed to go into negative numbers, so the difference between roll and TN can be large enough that someone can, say, earn two successes from one die roll of 4 against a TN of -6?
- Solutions for 6==7? Open. No proposals yet.
- Getting rid of Open Tests, decking's pseudo-Extended Tests, etc: Open. Open tests in particular seem to be a holy cow for Kagetenshi, while for me they're more of a proud nail, so I expect more debate here.
For E, one person recommended, when rerolling, add 5, not 6. So rolling 6 then 4 gets you a 9, not a 10.
A. I like handful or fistful even if they sound a bit silly. They are easy to remember and give a direct reference to what they represent - the dice in your hand you are getting ready to roll for a given test.
B. They are akin to Open Tests - something that comes up in certain circumstances. Define them similarly.
C. Glad to see it is closed, as I see no reason to try to homogenize Thresholds.
D. In play, I usually observe the GM making an allowance for a very high roll in his description, although it does not mean he allocated more successes to the roll. This always seemed fine with me, and I've never noticed anyone complain. Usually it brings a grin to faces around the table.
E. Total non-issue to me. It does not bother me in the least. Actually, most of the solutions for this aggravate the hell out of me.
F1. Open Tests are a non-issue for me.
F2. Decking's Extended Tests, I'll have to look further.
B. They are akin to Open Tests - something that comes up in certain circumstances. Define them similarly.
C. Glad to see it is closed, as I see no reason to try to homogenize Thresholds.
D. In play, I usually observe the GM making an allowance for a very high roll in his description, although it does not mean he allocated more successes to the roll. This always seemed fine with me, and I've never noticed anyone complain. Usually it brings a grin to faces around the table.
E. Total non-issue to me. It does not bother me in the least. Actually, most of the solutions for this aggravate the hell out of me.
F1. Open Tests are a non-issue for me.
F2. Decking's Extended Tests, I'll have to look further.
A. Bucket
B. They tend to be used to indicate special cases, a common mechanic to handle uncommon difficulties. Leave them roughly as is, maybe toss out the term threshold a little more often, but any core definition should indicate that they are generally used to make certain tasks non-trivial.
C. There is no debate C.
D. If groups like rewarding really high rolls, they will do so regardless of there being no rule in the book telling them to. If groups don't like rewarding really high rolls, they will be irritated if there is a rule for it. Leave the topic unstated and let each group decide.
E. Unlike some, I like the uneven probability change of different TNs. It leads to cases where the low light additional penalty is insignificant as well as cases where that additional penalty will cut the likelyhood of success past the point where you risk action. Reality isn't statistically proper at any given time, so I like that the same penalty does not always mean the same thing.
F. I've had fun with open tests. I saw a dwarf with 2 in stealth out-hide a stealth adept, and I've had a know-it-all mage adept roll one 5 on 10 dice (all the rest lower). As before, certainty removes the need for a roll, and open tests are very uncertain.
FF. I don't deck much.
B. They tend to be used to indicate special cases, a common mechanic to handle uncommon difficulties. Leave them roughly as is, maybe toss out the term threshold a little more often, but any core definition should indicate that they are generally used to make certain tasks non-trivial.
C. There is no debate C.
D. If groups like rewarding really high rolls, they will do so regardless of there being no rule in the book telling them to. If groups don't like rewarding really high rolls, they will be irritated if there is a rule for it. Leave the topic unstated and let each group decide.
E. Unlike some, I like the uneven probability change of different TNs. It leads to cases where the low light additional penalty is insignificant as well as cases where that additional penalty will cut the likelyhood of success past the point where you risk action. Reality isn't statistically proper at any given time, so I like that the same penalty does not always mean the same thing.
F. I've had fun with open tests. I saw a dwarf with 2 in stealth out-hide a stealth adept, and I've had a know-it-all mage adept roll one 5 on 10 dice (all the rest lower). As before, certainty removes the need for a roll, and open tests are very uncertain.
FF. I don't deck much.
The same penalty isn't going to mean the same thing no matter what we do—a +1 is going to mean more going from 5 to 6 (1/3 -> 1/6) than from 6 to 7 (1/6 -> either 1/6 or 1/6*5/6 = 5/36) no matter how we slice it, unless we redo the entire die system.
~J
~J
| QUOTE (Eyeless Blond) |
So, open questions on this thread:
|
One variant I saw many years ago is to reroll 5's and 6's and add the result to 4. Here is a summary of the % chance of rolling a certain number or higher for each system.
reroll 5&6 reroll 6
Die result add to 4 add to 6
2 83.33 83.33
3 66.67 66.67
4 50.00 50.00
5 33.33 33.33
6 27.78 33.33
7 22.22 16.67
8 16.67 16.67
9 11.11 13.89
10 9.26 11.11
11 7.41 8.33
12 5.56 5.56
13 3.70 2.78
14 3.09 2.78
15 2.47 2.31
Your math is off—you've got tn 6 in the add-to-6 method being .3333 instead of .1667. I'll check the rest later.
~J
~J
Silly question, do the votes actually count for anything? After all, Kage is ultimately judge, jury and executioner. From my experience, I've seen more cases closed because he sees a convincing argument than because everyone said such and such is a good idea.
I'm more likely to continue thinking about something that I'm not convinced about if it's a popular idea, but yeah, a convincing argument is much stronger than a popular but vague approval IMO.
~J
~J
| QUOTE (Kagetenshi) |
| Your math is off—you've got tn 6 in the add-to-6 method being .3333 instead of .1667. I'll check the rest later. ~J |
Sorry about that, that was done in 5 minutes in excel before a meeting. I didn't come up with this system, I got it a LONG time ago (pre-internet). I will look around to see if I still have the hardcopy I printed out. The person who came up with this had a fairly detailed mathematical discription of this method.
Well, let's see. If I'm properly awake:
| CODE |
TN 6+rr 5+rr 4+rr on 5+ ?2 5/6 5/6 5/6 3 4/6 4/6 4/6 4 3/6 3/6 3/6 5 2/6 2/6 2/6 6 1/6 1/6 10/36 7 1/6 5/36 8/36 8 5/36 4/36 6/36 9 4/36 3/36 4/36 10 3/36 2/36 20/216 11 2/36 1/36 16/216 12 1/36 5/216 12/216 13 1/36 4/216 8/216 14 5/216 3/216 40/1296 15 4/216 2/216 32/1296 |
Which comes out to approximately:
| CODE |
TN 6+rr 5+rr 4+rr on 5+ ?2 .833 .833 .833 3 .667 .667 .667 4 .5 .5 .5 5 .333 .333 .333 6 .167 .167 .278 7 .167 .139 .222 8 .139 .111 .167 9 .111 .083 .111 10 .083 .056 .093 11 .056 .028 .074 12 .028 .023 .056 13 .028 .019 .037 14 .023 .014 .031 15 .019 .009 .025 |
Unless I've done my math wrong somewhere.
Edit: I also doubt you got the system pre-internet
though there's about a year-long timespan you could have gotten it pre-web. Unless it was intended for something other than Shadowrun, of course.~J
| QUOTE (Kagetenshi) | ||||
Well, let's see. If I'm properly awake:
Which comes out to approximately:
Unless I've done my math wrong somewhere. Edit: I also doubt you got the system pre-internet though there's about a year-long timespan you could have gotten it pre-web. Unless it was intended for something other than Shadowrun, of course.~J |
You're right, not pre-internet, pre-web. It's hard to connect something like the old prodigy or compuserve BB to today's internet.
I can't find the old hardcopies, but I did redo the math and my numbers agree with yours. It looks like in the reroll 6's and add to 5 result in lower probabilities of getting target numbers of 7 and higher, and are significantly lower in the 11+ range.
The reroll 5 & 6 and add to 4 has similar odds to the reroll 6 and add to 6, with slightly more favorable odds of getting the higher target numbers.
If you're going to redo the die rolling system, the one I suggested would eliminate the 5 = 6 issue and give similar odds for getting higher target numbers.
I think you mean the 6=7 issue the other oddity in the dice rules is how big a difference 5 -> 6 makes, though I'm not sure it's a problem.
~J
the other oddity in the dice rules is how big a difference 5 -> 6 makes, though I'm not sure it's a problem.~J
List of Liber Non Grata added to first post.
~J
~J
Update on direction: sometime soon, hopefully over the Summer, I intend to begin advancing the in-game world starting from the last of the SR3R canon works. The SR3R canon contains all books from Shadowrun 1st, 2nd, and 3rd edition, English printing, except as listed in the Liber Non Grata and as altered in the relevant SR3R threads. While this will be under the SR3R umbrella, its development will be much more centralized—however, this should not cause any difficulties to those seeking to use SR3R rules together with pure SR3 (and possibly SR3-adapted SR4) setting developments.
Also, the following is to be considered a key concept in the development of the SR3R rules and the alteration of preexisting in-character material: "when fluff and rules disagree, everyone loses."
The SR3R project is not intended to disparage or lay claim to any trademark owned by WizKids or their affiliates.
~J
Also, the following is to be considered a key concept in the development of the SR3R rules and the alteration of preexisting in-character material: "when fluff and rules disagree, everyone loses."
The SR3R project is not intended to disparage or lay claim to any trademark owned by WizKids or their affiliates.
~J
That reroll 5/6-add 4 thing is very interesting. Remember, though, that the important thing isn't just to eliminate 6=7, but to make the relative difference between hitting two adjacent TNs as small as possible. Not only do we need to eliminate 6==7, but we also want to minimize times where, for instance, 5->6 drops the probability in half, relatively speaking. The above system does that, but there are a few that may be better:
Reroll 5s and add 4; reroll 6s and add 5.
Reroll 5s and add 4; reroll 6s and add 6.
Calculations: (EDIT: FIXED)
Reroll 5s and add 4; reroll 6s and add 5.
Reroll 5s and add 4; reroll 6s and add 6.
Calculations: (EDIT: FIXED)
| CODE |
| TN rrol6 +6 rrol6 +5 rrol5/6 +4 5(+4),6(+5) 5(+4),6(+6) 2 0.833 0.833 0.833 0.833 0.833 3 0.667 0.667 0.667 0.667 0.667 4 0.500 0.500 0.500 0.500 0.500 5 0.333 0.333 0.333 0.333 0.333 6 0.167 0.167 0.278 0.306 0.306 7 0.167 0.139 0.222 0.250 0.278 8 0.139 0.111 0.167 0.194 0.222 9 0.111 0.083 0.111 0.139 0.167 10 0.083 0.056 0.093 0.106 0.134 11 0.056 0.028 0.074 0.093 0.102 12 0.028 0.023 0.056 0.074 0.088 13 0.028 0.019 0.037 0.056 0.074 14 0.023 0.014 0.031 0.041 0.059 15 0.019 9.26E-03 0.025 0.033 0.045 16 0.014 4.63E-03 0.019 0.028 0.037 17 9.26E-03 3.86E-03 0.012 0.022 0.029 18 4.63E-03 3.09E-03 0.010 0.016 0.025 19 4.63E-03 2.31E-03 8.23E-03 0.012 0.020 20 3.86E-03 1.54E-03 6.17E-03 0.010 0.016 21 3.09E-03 7.72E-04 4.12E-03 8.23E-03 0.012 22 2.31E-03 6.43E-04 3.43E-03 6.28E-03 0.010 23 1.54E-03 5.14E-04 2.74E-03 4.74E-03 8.19E-03 24 7.72E-04 3.86E-04 2.06E-03 3.75E-03 6.77E-03 25 7.72E-04 2.57E-04 1.37E-03 3.06E-03 5.36E-03 26 6.43E-04 1.29E-04 1.14E-03 2.42E-03 4.39E-03 27 5.14E-04 1.07E-04 9.14E-04 1.84E-03 3.42E-03 28 3.86E-04 8.57E-05 6.86E-04 1.41E-03 2.84E-03 29 2.57E-04 6.43E-05 4.57E-04 1.14E-03 2.26E-03 30 1.29E-04 4.29E-05 3.81E-04 9.14E-04 1.86E-03 31 1.29E-04 2.14E-05 3.05E-04 7.09E-04 1.46E-03 32 1.07E-04 1.79E-05 2.29E-04 5.42E-04 1.21E-03 33 8.57E-05 1.43E-05 1.52E-04 4.25E-04 9.47E-04 34 6.43E-05 1.07E-05 1.27E-04 3.42E-04 7.83E-04 35 4.29E-05 7.14E-06 1.02E-04 2.70E-04 6.20E-04 36 2.14E-05 3.57E-06 7.62E-05 2.08E-04 5.11E-04 relative differences Avg -0.251 -0.294 -0.237 -0.215 -0.194 Std Dev 0.154 0.121 0.062 0.044 0.045 Std Dev (First 12 TNs) 0.152 0.126 0.061 0.072 0.077 |
Note the averages, btw. What those basically mean is that every time the TN is raised by 1, the difficulty increases by that percentage chance. The standard rule of 6, for instance, means that on average each +1 TN makes the test 25.1% tougher, +/- 15.2%.
It's the last lines that are most important, the standard deviations. These show that, of the probabilities shown, "Reroll 5s add 4, reroll 6 add 5" gives the best overall relative difference between different TNs. The other two rerolls work as well, both being far better than either the standard Rule of 6, and the standard "fix" of roll 6 = add 5. The differences between the three later rerolling methods tend to disappear when looking only at the lower TNs; here the "Reroll 5/6 add 4" gives the best results, but it's very close.
IMO the one that gives the best "end behavior", yet still has good performance over the most-used TNs is the "Reroll 5s and add 4; reroll 6s and add 6" rule. This one has an added advantage in expanding the useful range of TNs, that is, the range of TNs that you can reasonably expect to see someone actually hit in a given game. You notice the problem especially in decking, where a Detection Factor of 9-10 is pretty much a requirement to do anything in the Matrix, but by the time you hit 17 or so you're pretty much guaranteed to never be detected.
For those who were wondering, "Reroll 5s and add 5; reroll 6s and add 6." turns out awful; the numbers are already cluttered or I'd show why.
Another thing we should work out is the Rule of Ones. Instead of all 1s, how would you feel about a simple majority of 1s, along with no successes, activating the rule, sorta like Critical Glitches in SR4?
My initial reaction is negative, but the numbers aren't coming out the way I'm expecting them to. I'll have to look at that botch proposal in more detail.
~J
~J
OOPS! I made a huge mistake on those calculations; I forgot to do a relative difference on that last one, the Reroll 5 add4, reroll 6 add 6 one. Fixing now...
I want to make botches more common, however the problem with the SR4 rule is that it makes botches among people with 2 or 3 skill just way too common. In SR4 it sort of works because by default, the average person is rolling 6+ dice for a given skill, and not less than 3 even if you don't have the skill. In SR3, where 1-4 dice is the average case, and 6 is basically the ceiling, cutting the number of 1's in half means you're going to have an awful lot of botches.
Well, it's not just cutting the number of 1s in half—Eyeless' proposal gets rid of the godawful "botch while succeeding" possibility. One die, TN 4 has a 1/6 chance of botching now and a 1/6 chance of botching under his proposal, two dice had a 1/36 chance of botching and now have a (1/6)*(3/6) or 3/36 or 1/12 chance of botching—actually, I should have checked this earlier, as this is the first value I've checked that seems unreasonably high. Assuming we're taking the ceiling of half the dice for the 1s requirement, 3 dice is a (1/6)^2*(3/6) chance, or 3/216, or 1/72, which… I dunno, 4 is an average task, and for someone with 3 dice to be botching more than one in a hundred times…
It's not totally broken except maybe for Skill 2, but I'm not sure I like it. The other big problem with it is when you have someone with Biotech or First Aid 3 (proficient, mind you), you end up with about a (1/6)^2*((6/6)-((1/6)*(3/6))) = (1/36)*((6/6)-(3/36)) = (1/36)*(33/36) = (33/1296)… huh. The numbers never quite come out as bad as I expect them to—I guess a 2.5% chance of botching isn't quite enough to keep someone from trying to stabilize someone else, though in the case of a plainclothes cop (from SRComp, Biotech 2) it means a (1/6)*(33/36) or a 33/216 or a 11/72 chance of botching.
I don't know. I'll have to run more numbers, but the only glaring flaw thus far is "don't have Skill 2".
~J
It's not totally broken except maybe for Skill 2, but I'm not sure I like it. The other big problem with it is when you have someone with Biotech or First Aid 3 (proficient, mind you), you end up with about a (1/6)^2*((6/6)-((1/6)*(3/6))) = (1/36)*((6/6)-(3/36)) = (1/36)*(33/36) = (33/1296)… huh. The numbers never quite come out as bad as I expect them to—I guess a 2.5% chance of botching isn't quite enough to keep someone from trying to stabilize someone else, though in the case of a plainclothes cop (from SRComp, Biotech 2) it means a (1/6)*(33/36) or a 33/216 or a 11/72 chance of botching.
I don't know. I'll have to run more numbers, but the only glaring flaw thus far is "don't have Skill 2".
~J
Well, keep in mind that your skill 2 guy--who spent a grand total of 4 karma for the privelege, btw--will fail that "average task" 25% of the time; I don't see anything totally wrong with making a larger mistake 8.33% of the time. At the same time, why should that same guy with a skill of 2 botch a TN2 test just as ofen as he botches a TN 8 test? Shouldn't harder tests (higher TNs) be easier to botch?
If you feel that nervous about it, maybe we can also retcon the rule that you can spend karma pool to negate a Rule-of-1 glitch?
If you feel that nervous about it, maybe we can also retcon the rule that you can spend karma pool to negate a Rule-of-1 glitch?
What seems stupid to me, is that skill 1 man has more chance of botching than skill 0 - attribute 2 man. Because he's defaulting, he should by definition be a lot more likely to screw it up, not less likely.
EB - good suggestion, with karma pool
Angel - yeah, I think you basically got what's bugging me. Botch is based off of skill, not difficulty of the task, time taken, or even whether you're defaulting! EB's suggestion of allowing karma pool helps because it means if you're spending your full attention on it, you're probably not going to botch, but if you've had a really busy couple of days, you're less likely to be able to avoid mistakes. A second suggestion would be to say defaulting acts like a skill of 1 for the purpose of botching.
Angel - yeah, I think you basically got what's bugging me. Botch is based off of skill, not difficulty of the task, time taken, or even whether you're defaulting! EB's suggestion of allowing karma pool helps because it means if you're spending your full attention on it, you're probably not going to botch, but if you've had a really busy couple of days, you're less likely to be able to avoid mistakes. A second suggestion would be to say defaulting acts like a skill of 1 for the purpose of botching.
You could, but then a 10Str Troll is going to botch awfully often when defaulting to Str.
I rule I use is that when defaulting, all failing to reach 4 is a fail for attribute, failing to reach 3 for another skill.
I rule I use is that when defaulting, all failing to reach 4 is a fail for attribute, failing to reach 3 for another skill.
Unless I'm reading you wrong, that means someone defaulting with Attribute 1 is immensely better off than someone defaulting with Attribute 20, who is in turn dramatically better off than someone with Attribute 5000. Or was that to be combined with Eyeless' suggestion (one 1, all others failures)?
Anyway, I dislike high botch chances because it encourages not doing things (failure is often equivalent to having not tried, botching involves actively making things worse). I essentially never use a skill at 1—that 1/6 chance of botching is overwhelming. Assuming I'm not alone, why don't we just eliminate Skill 1 and make it cost 4 points of karma to buy Skill 2? For the record, I don't like that idea, but as it stands Skill 1 is worse than useless. In my opinion, your suggestion also pulls Skill 2 into worse-than-uselessness.
What about the karma botch-negation needs retconning?
~J
Anyway, I dislike high botch chances because it encourages not doing things (failure is often equivalent to having not tried, botching involves actively making things worse). I essentially never use a skill at 1—that 1/6 chance of botching is overwhelming. Assuming I'm not alone, why don't we just eliminate Skill 1 and make it cost 4 points of karma to buy Skill 2? For the record, I don't like that idea, but as it stands Skill 1 is worse than useless. In my opinion, your suggestion also pulls Skill 2 into worse-than-uselessness.
What about the karma botch-negation needs retconning?
~J
I think you're reading me wrong, someone with attribute 1 needs to roll a 4+ to not botch. Basically, they have a 50-50 chance of seriously screwing up. Someone with attribute 2 will have 25% chance of totally screwing up... and so on. So, attribute 20 man would be better off giving it a shot than attribute 1 man, because by definition att. 1 man isn't built for the task.
It basically takes the (default modifier/2) rounded up, and subtracts it from the roll. If what you're then left with is all ones, you botch.
As for karma pool, we just use it to re-roll as opposed to avoid and ooops.
It basically takes the (default modifier/2) rounded up, and subtracts it from the roll. If what you're then left with is all ones, you botch.
As for karma pool, we just use it to re-roll as opposed to avoid and ooops.
I was talking to Nezumi there, you posted while I was writing my reply.
~J
~J
oops
Yeah Kage, good point. My idea was pretty bad 
Hmm... Skills start at 2? Not bad. You're right that basically skills are useless until that point anyway.
Hmm... Skills start at 2? Not bad. You're right that basically skills are useless until that point anyway.
Hehe... the irony here is, for all skill levels other than 1, we're still talking lower or comparable botch rates to those in d20.
Anyway, yeah, I guess I can see how for SR3, with the lower number of dice we're talking, adapting these rules might be a problem. I'd just like to see botching be slightly more probable at higher skill levels; as it stands whenever someone Rule of Ones (the single time I actually saw it anyway) the GM goes out of his way to completely screw over the player, because the rule happens so infrequently. If it happened more often there wouldn't be that tendency so often, I think.
Anyway, yeah, I guess I can see how for SR3, with the lower number of dice we're talking, adapting these rules might be a problem. I'd just like to see botching be slightly more probable at higher skill levels; as it stands whenever someone Rule of Ones (the single time I actually saw it anyway) the GM goes out of his way to completely screw over the player, because the rule happens so infrequently. If it happened more often there wouldn't be that tendency so often, I think.
| QUOTE (Eyeless Blond @ Apr 19 2007, 09:51 PM) |
| If it happened more often there wouldn't be that tendency so often, I think. |
Not likely. D&D related forums are full of arguments over the validity of critical fumble rules. A non-trivial amount argue that 1/400 is still too common to shoot yourself in the foot while more have horror stories of the GMs who had a chart of interesting unintentional suicides to roll on every time you got a 1 on an attack roll.
It's a GM problem, not a probability one.
Gecko Crawl and Levitate have the exact same target numbers and drain codes. Now, Gecko Crawl doesn't increase your TN if you mass more than 100 kilos, but is that really a big enough compensation for losing both the ability to use it to manipulate objects and the ability to leave surfaces?
Thoughts?
~J
Thoughts?
~J
Seems good enough to me.
In addition to fixing just about everything else about the current drug rules, what are the effects of a mage with Focused Concentration ingesting Psyche? No additional bonus? Something else?
Also, Cram's crash effects are expressed in terms of stun damage, but they're implied to last for a fixed duration. Should that be changed? If not, how does that interact with, say, the Stun-reducing properties of the Pain Editor or the Trauma Damper?
~J
Also, Cram's crash effects are expressed in terms of stun damage, but they're implied to last for a fixed duration. Should that be changed? If not, how does that interact with, say, the Stun-reducing properties of the Pain Editor or the Trauma Damper?
~J
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.