Oh, aye, social contracts are always the most efficient ways to prevent rules abuse. Unfortanately, most societies have taboos against bringing violence to bear on strangers who are annoying you. The Shadowrun Missions campaign system appears to be designed for being picked up and played by people with existing characters who did not know each other prior to the game. This is where abuse becomes a more significant problem.

At the very least, discussing ways to break the big exploits allows GMs who get into the situation to throw down the appropriate countermeasure in game to explain why these tricks don't dominate the game world; it's partly a verisimilitude thing.

I´m afraid of hitting our SWAT member with a book. I could not run from him...

Now where did I put my sword???



But in earnest. Running a single-digit number of agents is not abusing the rules, very logical ingame, very efficient metagaming, and results in a very annoying number of rolls. If I put a limit to agent efficiency, I can freely grant success against low-rated systems, and leave the better ones for hackers. Arbitrary numbers of agent run against the hit cap without achieving anything, unless used in a botnet (deployed against massive numbers of targets).

Minor tweak, massive effect. And agents can still do any data searches / live-editing of video feeds / DDOS attacks.

I'm pretty sure that it is actually on one of the drones in unwired? I'm not with my books atm.

PS why the fsk is it that stupid? We already accept that agents are capable of all sorts of on the fly matrix interaction.

Agents == pilots in terms of capability. Heck, pilot programs can take matrix actions and have all of an agents capability.

If you are concerned with the efficiency of autonomous programs, giving them the ability to create code is not sensible:

First, you´ll need yourself a smallish nexus. Second, you have a few agents (start with one) have a replication orgy. Then you give your army of programmers the order to create a suite of ergonomic, optimisation, viral resistant software. Sell it about once and you break even. Not RAW!


A pilot does not have the same capabilies as an agent. A pilot is a drone-specific operation system. And no, no Software-drones in Unwired. The basis for this idea is taking the "there are other types of autosofts, whatever a speciality drone needs" rule out of the intended context (which is the physical world for drones).

(You are right, there have been some suggestions to rule this way because of the SOTA discussion. I must now cite a researcher from my university, who was at that time looking at a technical drawing project of mine: "Everyone wants to create something revolutionary. But you mostly only find out why the old ways are different.")

The medicine autosoft gives a drone the full capabilities of a trauma surgeon. Why wouldn't the software skillsoft give the drone the full capabilities of a programmer? Surgery is a significantly more complex field, as can be easily seen in the time required to graduate. Remember, a graduate programmer has skill pool 8, and our pilot is looking at 6, so its not actually very good.

You know we are already have some and are working on more automatic code generation right? This stuff isn't even particularly far fetched.

Being a surgeon is much about knowledge and motoric control. Both are well handled by SR4s tech. Programming is creative work.

Removing that boundary carries severe consequences, and I´d rather use Unwired with a few powerful tweaks, halve of them given by RAW.

Programming isn't creative work. You get your specification and implement it, and yes I trained and graduated as a programmer. Changing the access ID has to be a well documented and repeatable procedure because corporations do it enmass to sell agents.

Wizarding up a new function etc might be creative, but thats not what we are discussing here at all.

My apologies, I felt that the context was at least somewhat implied. I do agree that drones with the aeronautical engineering skillsofts may not be in a position to design a F22 jetfighter, however, all services - particularly common ones are explictly described as being readily performed.

To be clear, why does a software autosoft not give the drone the equivalent range of capabilities of a programmer when implementing new Access IDs into agents, in a similar way that an autosoft gives a medical drone such surgery capability?

Please note that the diagnostic ability implied by a medicine autosoft granted skillset does indicate that drones have significant reasoning capability.

What I want from the Matrix rules is very simple. How does it look when everything is going down and players are throwing dice. How do I describe the action to the players, how does the player's actions and die rolls describe the scene?

The RAW rules are confusingly formatted and Frank's rules make me feel like I need to know how real world hacking electronic warfare and encryption work, basically I feel stupid when I read it.

I'm not a programer, I have no interest in it. All I want is the illusion that my Hacker is a talented code slinger.

Here is part of a description of a program "Find Mind" "Disturbances in the packet flow can reveal the locations of certain kinds of interference." Stuff like this makes my head ache.

If everything you do in VR is translated into some sort of visual queue that happens to your persona or the environment around the persona, then cant I describe what programs or hacking actions that I'm using behind the scenes by describing what my persona is doing?. Tell me then what my persona can do; can it run, can it hide, can it climb walls? Can it put on armor, or draw a sword?

I understand that the Matrix is an abstraction of the exchange of electrons between programs. Does this mean then that there is no in between places? How does the GridLink node look from the outside? Can you only describe places (nodes) from the inside? If my refrigerator's node looks like a control panel floating in an infinite white space and all of a sudden I want to hack the SoyWagon grocery delivery service, what steps do I take, what rolls do I make, and what does it look like?

Monk raises a good point. Especially for general rules.

Clueless as Gibson was on this stuff. It's pretty understandable if your icon rides a freaking shark made out of military code into the heart of some geometric shaped node and makes something happen.

Horrid as they were about sucking up time on the table the old old rules were actually pretty easy to follow that way. Players who didn't know how to use a computer or what was involved had a fair idea of it.

I'm guilty of sort of losing track of the icon in SR4, probably most of us are.

I think that's pretty much the issue...
...For some bizarre reason, SR has become Computer Weekly where discussions about coding and more complicated rules is the order of the day - Realism is fine, but Netrunning/Decking is about getting stuff done.

Number crunching seems to have become the aim, in its own right [which is always a bad sign] and not a means to an end.

Bizarrely, SR has modern day computing/networks/software in 60 year time...
...Completely ignoring the speed that computing changes, and the immense changes that have changed in the SR universe.

Do people honestly believe that in 60 years time, people will actually be initiating programs and manually Netrunning ??
That whole concept is little changed from the "D&D in cyberspace" of CP2020 and early SR.

Yah, Netrunning is a cornerstone of Cyberpunk...
...But it was never the focus because it's BORING to read about and do - I presume that's why people wanting to play Deckers are so hung up on individual software titles, etc - there's not much else ["yah, I made that radical 20d6 dice roll to open the door - i'm so fucking cool" isn't much unless you wrap it in fancy names like "BlackHammer 5"].

SR has simplified Netrunning dramatically in the last 20 years, which should say something about how cumbersome it was and how little fun it was and how time intensive it was...
...And how desperate the SR devs are to make it part of the game [like they did with making Riggers an offshoot of Deckers, because almost no one-played a Rigger - which was a complex series of number crunching exercises [sound familiar ??]].

Software will most likely be written by other computers, because people just won't be upto the job - it's the way software gets designed now...
...Fewer people code individual lines because it get more and more difficult and time consuming to maintain that code, they use "simple" interfaces that write the code for them.

We can't imagine what computers will be like in 60 years [Traveller had computers with less than 1MB running starships and plotting hyperspace jumps, 30 years ago - Mr Gates was convinced that no-one would ever need more than xx KB in a computer, 15 years ago there was no internet as we understand it, now most people don't know about computers but they do know about applications and interfaces because that's what a computer is to them - the software, not the code]...
...But SR seems stuck in a rut as far as computing and its future is concerned [Agents/Avatars to do all the computer work is the most likely way people will do things in the future - like they do now,m but more so].

In my view, having separate programs and their actions is at least part of the problem...
...Resolution should be much more generic [tho that won't please the "my individually named program +1 is so much better than yours" crowd] with software suites being used like attributes - Offense/Defense/Stealth/General].

But then SR's view of a wireless world is so laughable, it's unworkable...
...Systems so unprotectable that any kid with a crappy RadioShack can bust it and abuse it, banking systems that aren't secure, SINs that are so easy to have or break that they could never be usable - not a consistent workable world, but a "Characters must be able to succeed at everything" setting.

Who would build a computer network or design security protocols, that was not secure and proof against spoofing/etc - after a second crash ??

Looking at all my SR material, from SR1 to SR4 [I was so hopeful I bought a limited edition, a hardback and a PDF] and it's sad to see what SR has come to...
...Retconning, rules,setting, fundamental consistency.

But that's just my view...
...And I'm entitled to it [which is why I stopped playing SR].

To be fair I think they're trying to "get stuff done" with the SR4 rules. I'll give them credit for chopping time down a lot. They just didn't playtest this stuff. Especially the crazy technomancers. That's really the problem.

And I do think hacking as a whole is important to a good cyberpunk setting, it definitly adds. And I'm having a lot of fun with my players and SR4 hacking.

However regular hackers at least can't just do anything. There is offline storage, allowances for requiring special chips to interact with certain systems (it showed up in SR missions at least), and it seems the sky is the limit for firewall. And RF paint keeps them off of a lot of system.

There is already things that can happen on a forfeit with hacking that will set a shadowrunning team back. I recall one SRM where the team needs to get a file out of a computer. The computer is not wireless enabled, it is the size of a desktop computer (this was before unwired defined nexi). This computer was inside a facility that would be a pain in the ass to infiltrate.

Of course the computer is running ice. The big problem is this, if the computer discovers your presence not only will it shut itself down, it'll scramble it's data storage. Now it can't boot up because it can't read the OS. It's as large as a desktop computer so you aren't slipping out it. Of course you have now alerted the security of the place with your blunder. So now to get your data you are going to have to do it the hard way with a lot of lugging and shooting and yelling.

To make matters worse, the Johnson will dock your pay if you present her with a scrambled computer because now she has to hire someone to unscramble it.

I actually know someone IRL who uses a similar security setup on his desktop computer. His hard drive is encrypted. The decryption is built into the BIOS. The bios requires a password, failure to guess the password in 3-5 tries, I forget how many exactly, results in his bios deleting itself, leaving you with an encrypted hard drive and no key. So, this is not too far fetched. I've seen the real deal.

Paydata on sites you can access via the matrix at large is kiddie stuff. That's not what they hire shadowrunners to get. That's not what's gonna pay for the big guns. Paydata like that is the stuff the hacker grabs during the legwork phase so you have the building blueprints, it's a trivial side-job. It's the internal networks that have high security data in corporations and whatnot that is the juicy stuff. That's where you separate the ameture enthusiasts from the mercenary code-slingers. That's a situation where a failed hack job will sound security alarms, with your team in the building. That's the situation where a botched hack attempt makes everyone's life that much more difficult.

That's where I'd go in the meat, and not an agent, and here's why. Hacking is not about the cybercombat. If you are exchanging blows with ice then you screwed up, it's a bad day. If you play your cards right, security will not show up, you won't have to roll fistfulls of dice. Your brain won't be on the line. This is where edge makes or breaks you. No matter how good your dice you are going to roll some goose-eggs and glitches and critical glitches. You will bomb rolls, you will get spotted. Edge is a band aid for such situations. When an agent gets a sucky stealth check vs patrolling ice, you're hosed. If you are running a skillwire and bomb a stealth check, you're hosed. If you are doing neither you at least can declare edge and hope it sorts itself out. True, you could drop out, but like I said, hacking a matrix site is kiddie stuff, you're inside an ARES compound and you just triggered an intruder alert this is where they make or break a shadowrunner hacker.

And as for the problems of overly hard systems to hack forcing baser tactics. That's up to the GM what you will have to hack. I mean come on, the gm could throw a firewall 9 system at you, true, but he could also throw 3 cyberzombies and a dragon at you too. And in such instances, unless there's a very good reason for the gm doing so, you can roll up a newspaper, thwack him across the nose and say "bad gm".

Got pointed from Frank's original post to here, so I'll just to throw in my two cents...


Sometimes the best solution is the simplest one and going with the saying "a tool is only as good as its user" I'd suggest the following house rule instead: cap how much of the Agent's rating can be used in the dice pool with the hacker's skill. The reasoning behind the house rule is based on the the description "Agents have a Pilot attribute just like drones (see Pilot Programs, p. 213) that determines just how “smart� the agent is. Pilot acts as the agent’s brain, interpreting orders." A skilled hacker can give a complex order to the Agent, which it will execute to the best of its abilities, but a script-kiddy who doesn't have a clue what he's talking about will wind up hindering the Agent as he confuses terms, forgets specifics of orders (there's a limit to how much an Agent can interpret, claiming an Agent understands the script-kiddy's intentions based on a skimpy order is horrible metagaming), and constantly second-guesses himself.

Not that I'd use the house rule myself; it was pointed out that our team's Face is basically a Script Kiddy but despite having both a good Agent and programs, it would have been disastrous if we had him do the hacking rolls instead of the resident technomancer.


If a hacker isn't actively monitoring his commlink or Agent, that pretty much guarantees a spider is going to track him and have a security team come over for an impromptu face lift with their boots. He also isn't going to be around when the Agent rolls poorly and screws up a sensitive mission - not that being around would help, since he can't use his Edge to reroll something the Agent did.


Essentially this is just a botnet running a DDoS attack, albeit with the intention of breaking into a system; while it's certainly possible, this creates a huge problem for the hacker: he can only pay attention to feedback from his primary commlink (maybe three at a time if he's an adept with the Multitasking ability, but that drops back to one the moment he's pulled into cybercombat) while the backpack botnet does not provide feedback to avoid overwhelming the hacker, which allows a single Track attempt on any of the unmonitored personas to go unchallenged and, eventually, invade your system.


Pretty much the same pros and cons as the Hackastack, although the added benefit of each Agent being able to act semi-autonomously makes it a slightly better offensive botnet; defense, however, is still going to be woefully inadequate to protect you against counterhacking attempts.

The four "exploits" in the hacking rules may be useful but they also come with some pretty serious consequences - namely discovery; a hacker can use them but he'd be the troll in milspec armor of the hacking world: he'll get one, maybe two, shots at security before they turtle up and everybody and the janitor will know he's been there. A player can't blame the rules when they piss off the rest of the group by tripping every alarm and sending security to a higher plane of existence through sheer paranoia; when playing a game don't become obsessed over one tiny aspect of the character (namely the "hackers do whatever because they can") at the expense of everyone's fun - which ultimately affects your own enjoyment of the game.

Right, and because a lot of VR Matrix hacking are just trivial side jobs then why should you devote a lot of game time to it? So now the techie feel of the Matrix is taken out of the GM's hands and now resides in the rules themselves.

Now when your character does stuff in the Matrix it feels like you are hacking because you have to decipher rules with techie jargon, and since the rules themselves are so obfuscated you actually feel like you are decrypting code.

So when the confusion you felt when your programer friend tried to explain to you what is really going on behind your on-line banking and the feeling you get after your character hacks LoneStar is the same, you must be using the rules just fine.

The RAW almost works. The problem is that it gives you the feeling that if you already knew how everything works then the way it is written and it's layout would make perfect sense.

I have been running a campaign with four core players and two intermittent players for about ten months, we've played about twenty games. It is the most complex and detailed story I have ever gamed. It is among the best role playing experiences we as a group have ever had. I have recently wrapped up the storyline and given the GM reigns to another for awhile (he's not running Shadowrun) because I don't want to get burned out. At the end of it I confessed to the player running a technomancer that I really can't get my head around the "Wireless World" rules. He said "Oh good, then I guess I'm not such an idiot after all."

At that point I realized that I really needed to learn these rules, and I needed to know them well enough where I could get all of the other players to know them as well, since the Matrix is something every character must be aware of. I've come to realize that I need to make vast assumption about the mechanics of the rules, which is fine but to keep it consistent I need to solidify my mental picture of what the character's actions look like.

What does it look like when my matrix user jumps into VR then sends her drones a command with one action, and jumps into one with another then fires it's weapons then sends a team member an AR message then attempts to find and hack a node. How do I describe the action? How do I make the rules flow with the action?

The most simple, elegant houserule to "fix" the script-kiddy syndrome is switching to logic+skill, and taking away metahuman performance from autonomous programs. Unwired does offer the first; the second is a given for true hackers. Mooks are a solution due to their possible numbers only. (Logic 8 + Hacking 5 (Spoof +2) + Hot SIM + Encephalon 1 = 16/18 dice already). The only concern here is limiting a group of agents.


Drop Out has NEVER been an issue. It is - at most - a valid and limitating strategy.


Hackastack is only another name for Agent Smith, as far as problems are concerned. There is (outside agent smith) no relevant processor limit, there are umpten ways given in Unwired FOR avoiding that. Go figure.


My conclusion: The only issue is agent smith. I´m considering to limit agent hits because it takes away the statistical approaches to agent usage, without impacting anything else. It also gives the non-hackers with lowish matrix pools, but edge, a way of being better than their mooks. Nobody uses massive IC in combat because "true hackers dodge it like neo". Run as many agents as you want, the opposition has hackers, too.

If you want to focus more on the description of the Matrix action and don't want to think too much about the computer side of things, just play things as if it was some kind of physical break-in. The hacker is an intruder, the stealth program is the disguise skill, the spoof program is the con skill, the exploit program is the lockpicking/overall technical skill and so on. The only difference with a physical break-in is that things can look any way they want and there could be situations that aren't possible in the physical world.

As for Agent Smith, I think most of it has been covered by Unwired. If you run your agents in a lot of nodes you've hacked everywhere, just consider it as a botnet. If you run them on your own gear, it's a hackastack situation, which isn't much of a problem to me because the hacker has to buy all of the gear then spend some time changing the accessID on each agent and it won't be very different from a rigger having a pack of drones or a mage with a pack of spirits, two situations that aren't uncommon and that nobody considers wrong.

Alternatively, leverage the fact that the spiders literally control the battlefield. They can call in sec personnel to pull machines offline at will. This is unlike any physical analogy possible. It's more impressive a weapon than the Enola Gay; they can unmake part of the digital battlefield, along with any enemies contained therein (if they are digital in nature). They then purge it of targets and restart when the whole network is clean.

Drop-out enables you to own mook swarms. If you can identify that you are under attack and it is sufficiently annoying, you drop out of the Matrix and clean yourself up. Meanwhile your honeypot system has gone up and starts tracing the bots back to their sources for the spiders to fry. DDOS is a common occurrance in 2070, and the corps are allowed to go out and destroy the OS of anyone who unwittingly hosts a bot on their Commlink. Protip: they're not going to be gentle, not even if you program your Agents to ask in the voice of a little girl.


Whilst the Matrix in 2070 works as a rules system, I have issues with its depth and the particular game design choices made by the developers.

There could be three levels of competence for matrix actions:

Rating 1-3 agent/IC: max. 2 hits =>Cheap, reliable performance
Rating 4-6 agent/IC: max. 4 hits => dangerous for non-hackers / damaged hackers
Hackers/Spiders => mostly battle each other

Something very important with agents, is that it's up to the GM to decide if they can or can't do some complex task (by complex task I mean something that needs more than 1 matrix action and which needs the agent to react to what happens or a task with unknown parameters). This means that the GM can simply tell you that your agent was unable to complete its task, and there's nothing you can do about it. So relying on agents alone is relying on the GM to decide whether you can successfully hack something.

Another thing is the agent's vulnerability. Your agents can be infected during its hacking session, and if you fail to notice it, you can get in trouble easily.

In the SR games I've played the matrix site hacks are rarely a problem. One or two nodes, bring a friend along. Granted the SR games I've run and played in generally have missions go down like this. Johnson gives details, team breaks into frantic information gathering, the face calls his contacts, the hacker tries to data-search for maps or crack a few nodes if the maps aren't public. Generally speaking the general matrix is useful for gathering legwork type information. This is where that slow hack is handy, because why bust your hump and get in trouble over your preliminary legwork. If there's any serious hardcore, cold sweat, black ice breathing down your neck hacking going on, then it's generally something mid-run and it is often time sensitive so there's little room for error. Just to reiterate, script kiddies knock over porn sites, shadowrunners infiltrate R&D facilities and crack secured mainframes on site for the latest wiz weapon blueprints. The only way I could see pre-run hacking critical to the main job is if a hacker is trying to compromise the security system of a compound he can link up to, so he can install a backdoor and figure out what cameras and RFID tag trackers he can use. With the right moves you could patch video feeds from the surveillance cameras to your shadowrunner team, and use the guards RFID badges if present against them, showing real time locations of the baddies. And if they made all that available through the matrix they deserve what they get. That's not to say nobody will make such a mistake however, not every compound is perfect. Even that sort of job is not screw-up tolerant, because if you trigger an alert while hacking that you can bet your programs that they will do a security audit pronto and your backdoor will be discovered and removed and any contingencies you planted in the system will be removed.

As for agents, I honestly don't see agents a problem. I already pointed out the horrible stealth problems with them. Also they are cumbersome, you gotta command them then they gotta acknowledge and perform the command. Now, it doesn't say if you can see through the agents eyes. So a conservative GM approach is to limit agent feedback from out of node agents a list of status messages, like the console window of the IRC bot I use for dice rolling on my irc games, the Funkagroovitalizer. Don't even have the agent make a perception check when it enters a node unless its script says so or the commanding persona sends an order for it to do so.

If you really want to mess with a script kiddie, take the power out of his hands, because that's what he's doing in real life. And you can do this without any serious rule modifications. I don't recall seeing anywhere it saying that agents get +2 dice for hot sim, so don't allow a player to try to talk you into giving it. You roll for his bot, and you tell him the report status of his bot if it isn't linked to his persona and off-node. Make him define the agent's script. If something comes up that isn't covered in the script then have the agent sit there stupidly. And remember that agents don't think outside the box or follow commands that require a degree of free thinking, and if you're trying to finaggle an unfamiliar node. Giving that degree of seperation from the bot will separate the player from the experience and make it as tedious for him as it is for you. Even as is, with personal experience as a GM, a shadowrunner running an agent is cumbersome, hands off, and jury rigged. It just isn't as elegant as one might think, and when you are in that Ares mainframe lifting those wiz blueprints, you want that extra layer of security edge can give you against bombed stealth rolls, because if you trip an alarm, the whole facility will go on alert, and since you are on site, that's very bad. There's no second chance. If you are using skillwires on the matrix, that's just foolish, the cap out at 4 dice which gives you stunted pool, to make matter's worse, you don't have edge. You can try to mitigate it by running hot, but then if you get hit by say, a black hammer, and you bomb your cybercombat defense roll, you're going to wish you had the opportunity to use edge, a little edge might give you enough dice to weasel out of being hit by the black ice, giving you a more time to drop into AR, log out, or just crash the ice like demolition derby. As far as I know, skillwires can't even get specialized, so you are losing out on a quick way to get a fast +2 on actions you can see your hacker doing.

There is but one actual rule modification I can think of. Since skillwires work by overriding your actual intuition with programmed rote, I'd make codeslinger incompatible with it since it's overriding your innate talent. But where does this all add up? Well, let's suppose your hacker realized that avoiding detection is key in hacking. So she skill specced hacking (Stealth), she's a codeslinger (hiding in a node) (this move is debateable since it really isn't listed as an action, it's more reactive, I would allow it in a SR game I'm running through), and she's running hot. That's +6 dice and we haven't looked at skills and programs. Let's suppose she eventually ends up running a rating 6 stealth program and has a 6 in hacking, that's 18 dice. One encephelon later and she's hurling 20 dice. Now if you are an evil GM you could say the encephelon and skillwires conflict as a houserule, but assuming you don't Mr. Script Kiddie is hurling 15 dice tops without edge. May seem like a lot, but I just rolled that a few times and got 2 2's a 5 and a 3. Those 2's are possible detections, I've seen rating 4 ice swing that easily. And he can't fix that booboo with edge, he's stuck as he triggers an alarm and now he's busy fighting while Ms. Sleek stealths on by. I suppose he could snag the skillwire expert system in augmentation but by then you've already sunk 13 grand in cyberware alone into this, and I think it's a little around your arse to get to your elbow. Not to mention that if someone manages to compromise your pan your skillwires could suddenly crash at exactly the wrong time. And Mr. Agent? he's swinging 12 dice tops, the lowest of the three and stopping occasionally to ask for directions and can't use edge. He's definitely going to trigger something.

For getting a feel of how the wireless matrix works you might want to watch some Ghost in the Shell (like the TV series). I get the feeling SR4 is patterened after it. Except that you need later SR4 supplements for stuff closer to ghosthacking.

As an added bonus you get a feeling for the general Japanese racism (though I don't think they deliberatly put that in there but it leaks through a little).

The trick with the rules is to keep them simple and describe whats going on as you go. Um. It's hard to say better than that. What you could use is logs of someone doing this.

Found some legal stuff on youtube from ghost in the shell.

Some AR examples here
http://www.youtube.com/watch?v=aRTerjZuNxw Note the AROs

Some hacking examples in this though without the origional audio
http://www.youtube.com/watch?v=FWkMKzoip6Q
You however are much freer in how you want to describe the virtual world. But glowing things with gates works well enough.

This gives some idea what it might look like from a drones POV.
http://www.youtube.com/watch?v=QeIcO-gbETw...feature=related

Yeah, I've used the Stand Alone Complex series a lot when I need to visualize something. And lately I've been reeding my Unwired PDF, I really want to wait for the hardcopy because it's easier on the eyes. I've gotten to Matrix Topology and it has helped.

Hi everybody,

I just wanted to talk about some other points i found about matrix rules.

Although agents are quite unbalanced (in my games i have reduced agents to be IC to leave some space to the team hacker) i have to say matrix system is still unbalanced. Why? Sure a hacker will have some more options than agents to boost their hacking pool (in the way of specialties and qualities). But a player of mine made me freak out showing me what a adept hacker can do. It can have all the qualities a hacker can have and still may take some nice magical tricks. We are talking about players that can have dice pools around values of 10 (4 from skill, 3 from improved skill, 2 from hot sim, that without taking into account specializations, qualities, cyberware)... From character creation maybe he can't blow zurich orbital node but sure it can give a good fight and win in high security nodes (rating 5-6).

To my point of view, leaving the agents aside, another problem is that there's no difference between commlink and node ratings. Where are the red nodes? why stationary nodes are placed around ratings of 1-6. i'm sure i wouldn't mind if my security node is as big as a wardrobe, as long as it keeps my info safe (so I assume the wardrobe will have some much power than a portable commlink).

Also, after comparing the black IC of 3rd and 4th, we saw that has become now some kind of spam in the security nodes that may annoy you. Reasons? The first and the most important is that the chances of a black IC to kill you are so low that most matrix users doesn't give a crap if it's normal IC and black IC (try to kill a guy who has +2 to matrix tests, has better programs than you, and has between 3 and 5 IP) . Yes, they may destroy your commlink if you aren't in hot sim, but how much is that? 10000 (asuming you have backup programs and you only have to buy the commlink) ?. Please, cheap bioware is at least twice that price.

Solutions we managed for now (while we wait for Frank's House rules)

Nodes:We raised nodes ratings to 3-12. Being 6-7 the average corporate node, and leaving higher ratings for security nodes and other sensitive systems. That leaves more challenging nodes.

Black IC: As the Black IC doesn't gets bonuses from active alerts in the node, we decided to design an special capability of the Black IC. We decided to give the program the ability to track unauthorised users as a free action (of course bonuses of being hit by one of them applies to the extended test). If the Black IC manages to pinpoint the hacker connection, it improves his capabilities to fight the hacker as he can anticipate the actions of the hacker and reroute better the lethal biofeedback. In game terms, Once the user is tracked, the Black IC raises his rating and his atack programs by +4 and overrides cold sim users. We are arguing now if we should give him more IP passes but we think is fine for now. This way is quite more lethal but it gives you time to try to kill it or run away.

I'm open to opinions and suggestions

P.D:Sorry if my writing it's a mess, I still need to improve my English.

Without house rules the sky is the limit on firewall. So in theory you could drop a 25 firewall on them and watch the players squirm.

Hacker Adepts are nice. Eventually rolling 6 more dice than a regular hacker. Though a regular hacker could use the BP to take their edge to magical levels, which rather balences things out.

If you're worried about something being broken it's technomancers. Though as far as I can tell from the fluff they're supposed to be broken. Hacking better than everyone else and making drones do crazy things. The tradeoff being the corps want them and they're extremely tight for BP and karma whereas the hackers can pick up other useful skills and stuff.

The main rule of abusse is that technomancers can surpass the limits of ratings stablished in the core book. I have to say that most of their powers give dice to their programs rather improving their ratings (These balance a little the game when you are targeting the stealth program of the PC for the matrix perception test in order to detect him). But still having programs high enough as rating 8 or more is unfair (to the point where a secure node can't detect the technomancer while is tearing appart the node or it's defenses).

Also, free sprites invites players to abuse them, rolling up to the problems of agents or worse (because of the multiple uses of sprites).

Also, the poor sistem of binding sprites, technomancer foci (widgets), the echos and improvement based in karma-only expenditure that even awakened doesn't have makes me feel this kind of character still needs some serious testing.

I think the main problem is that they differ too much to the hacker but mayor changes for something more fair is quite dificult to design.

Technomancers are not THAT good compared to a dedicated hacker with a sub-zero implant cluster. Cyberware really starts to shine when you want the node, too. But I think TMs can finally be played without abusing the hell out of their advantages (with karma based chargen). Should really help their image.

Adept hackers with high skills, good equipment and maybe even some qualities SHOULD walk over all common nodes. Maybe playing something with less dice is more challenging, no? About 16 dice make you survive a rating 5 system, but only with a bit of sweat. Autonomous programs are only trippwire - they call in the "security consultants/threat assessors".