My take on how rep systems work

Remember that tech in the EP universe is quite advanced; thanks to facial recognition, Mesh ID permissions and other long-distance biometrics, it'll be quite easy (and relatively automatic) to read up on someone's reputation. In fact, this is exactly how it was intended; reputation systems are designed to be sousveillance systems in which the public keeps track of itself. The concept of privacy (and refusing positive or negative rep) is anathema to the very concept of a reputation system. We aren't just talking about a merit-based economy... we are talking about a system that also replaces police, public record, and other modern faculties. Especially in the outer system, reputation is a means of keeping track of criminals as well as the legitimate. Every heinous act you pull in an anarchist habitat nets you negative @-rep, and other habitats will know the kind of person you are very quickly.

While the majority of reputation accounts will be tied to IDs, not all of them will. For instance, the Eye likely keeps track of you by your code name, and other information that Firewall keeps. Much of your data will be hidden, because Firewall does not like it when sentinels try to find out the real identity of other sentinels. In another example, people who refuse to register with a rep network will likely still get an account the first time they do something notable, good or bad. Of course their ID won't be in there (unless somebody knows what it is), but I guarantee that people will upload photos of their face, voice samples and other info into the system so that they can be identified later, whether by name or not. You can tie your Mesh ID to your reputation account (or accounts, if you successfully pull off multiple identities), but you are not obligated to do so. Someone else may very well do it for you at some point anyways.

In essence, reputation networks are only barely like Myspace and Twitter. A better example would be eBay (but even then it's not very accurate). The core concept behind a reputation network is that your score is a peer-edited critique on your actions. You cannot control your score beyond simply doing things in hopes of raising it. Others control your score, and other information that you'll find in your profile. It is essentially the futuristic replacement of our modern public record system, which also functions in a merit-based society as a form of economy.

Yeah, I would agree that modern means of faking IDs would be completely pointless. Simple facial recognition would spot most people immediately if they tried something so simple as buying a fake ID to score some drinks while underage. However, the transhumans of the world of Eclipse Phase have much greater tools at their disposal. Buying an extra morph (with credit so anonymity can be maintained) will go a long way to establishing a new identity, as will changing your mannerisms (in order to fool kinesics experts). Hell, if you're a wiz in the mesh, or know someone who is, it may even be possible to alter or erase data in your network profile, clearing up your reputation or spoofing a new face in it, so you can escape means of capture. It'd probably be out of the means of just any hacker, but a hacker with a dedicated team of allies... or the resources of Firewall's Vectors... will likely be capable of changing your reputation as you wish, or potentially how you don't wish. Be warned, however... if someone notices, everything will likely be back, and you will further be recorded as the dick who altered your own profile for personal gain.

As for networks being private, it's somewhat like that. I wouldn't call it "private" so much as I'd call it "invitation only". The Eye is certainly inaccessible to those who are not Firewall agents, and it would be downright stupid for Guanxi to be accessible by the Jovian police. Certain actions will get you in, and certain actions can also get you kicked out (which is essentially what the Blacklisted trait represents). Some are more open... Circle-A, CivicNet, EcoWave, Fame and likely even Research Network Associates are probably open to anyone who shows enough interest to register to the network. The secretive ones like Guanxi and the Eye, however, are likely member-exclusive... and I'm sure there will be others that are similarly restricted.

I'm glad I could keep you on your... um... brain toes....

In the EP universe, to a "normal" person under "normal" circumstances (which means the vast majority of the population in the solar system), there is basically no point in faking or hiding your rep netoworks. First, without rep you wont get what you need, especially in the outer system. Second, without rep you will be viewed with suspicion - a person with blank rep scores would be akin to a person (in today's terms) walking in to shops and to the office in a ski mask.

Ok, I've lurked enough. I have a hard time with some facets of the rep system, and thus the following questions come to mind.

Question: Who or what keeps track of these rep scores? Is there a system of storage banks throughout the system that constantly transmit rep changes and creations between each other so pinging a rep of someone brings back the correct, or near correct, score for the person in question? If there is a system or set of systems that keeps track of the scores (and reasons for ups and downs), it would lead to the question, can they be hacked?

Question: How does the process work in the background for increasing someones rep? Say, if I wanted to give Jack Graham some rep in the f-rep area for creating a game for me, where would I turn to give the information that I want to do that? And do I need to give a reason for it? And how much can I give at one time? What decides if a rep increase is too large?

Reputation networks are decentralized redundant server systems distributed throughout any region where the network is recognized. Any changes to reputation are recorded and cross-referenced to previously existing reputation data, in order to ensure that cheating doesn't occur. People can try to hack a reputation network, but it would involve a great degree of effort (you'd have to hack in and fool every single server on the network, because if a single server realizes the error caused by your hack, all servers will likely be corrected).

Chances are that people don't actually grant a score per se. Rather, they notify the network of what exactly the person did to deserve a reputation change, and people throughout the network decide how much of a reputation increase or decrease it warrants as a group. In fact, it may be that any given person can only cause a single-point rep increase or decrease, and larger alterations to the score are caused by groups of people reporting about the person's exploits, on account of the fact that larger favors likely help out bigger groups of people.

Remember that reputation networks rely on its community of members to actually update any one person's reputation score. In order for you to get the rep you gained, the guy you did the favor for needs to post about it. In that sense, the speed you get your rep is also largely reliant on how little your clients and allies procrastinate on reporting what you do.

If I had to put an average on it, I would say that your reputation would likely fully update in all regions of the system in about 2 weeks. This could be sped up to about 1 week if you can get people off their lazy metaphorical asses and tell them to blog your deeds on the network.

Well, the watchers of a network are the people on the network themselves. Everybody acts as a watcher. Think of it in a lot of ways like something akin to Wikipedia. Anyone who feels like it can contribute and help maintain the integrity of the reputation network. You would have to manipulate every single person on the network to ensure that you'd be able to maintain your network gain... and at that point, you've probably earned that rep bonus simply by merit of showing how awesome you are to the entire network solely for the purpose of "hacking" it.

What I'm wondering is, does the listing all the networks you belong to show up when some person do a mesh/AR check on you? (and reciprokely)
that could be problematic in some cases.
for exemple, in a Planetary Consorsium habitat, a team of Mercurian Shield is doing an ID control.
if you got some @-rep, it would raise some flag. but what about the G-rep? would that get you stacked in simulspace cell faster than you can spell "Kovacs"?

it could also be a source of annoyance.
Imagine a young woman who spent years on a scum barge, but wanting out really bad, and having a job in a respectable hypercorp. So she works her arse off, earning enough rep and creds to get an egocast to a PlanCon habitat. She writes resume and letters, and eventually secure an interview. She rocks it, until the rep she earned on the barge shows up.

"Ah, Miss, with such a resume, I think you'll fit right with us at Eden. Now, I only need to do a quick rep check for the accounting and HR department. *does the check* "Aaah, miss, I see you have some rep in the anarchist network. I'm afraid this is against the policy of the company to hire people whom relationships could present a security risk. aah such a shame, but I have no choice. good luck with your job hunting."

I guess you can get your muse to check some rep in particular, either for security measure or recruitment

how does the interface look like?
I like the idea of a 'Mesh facebook'

I could see G-rep would be a closed network, where not even those with reps on the network knowing what their G-rep is. Somebody who gave somebody in Guanxi a ride in their cab/shuttle would probably earn some g-rep for it, but might never know.

People are less likely to end up on more specialized networks, like RNA or The Eye, but if you design an amazing AR suite, you're likely to earn rep on @-rep, c-rep, f-rep and g-rep. A local crime boss who liked your suite is likely to give you a bit of rep for it. It won't be much, but it will make getting a favour a little easier. "You have made the don's life happier. Therefore, we shall help you."

A hypercorp hiring agency is less likely to care about a potential employee's positive @-rep entries (indicative of being a good citizen and competency), and more about any negative @-rep entries (Does the individual not play well with others? Are they incompetent?) and most importantly, negative c-rep entries (which is where you'd find things that really annoy the hypercorps, like sabotage).

By specialized, I mean that it concerns itself with scientific matters, rather than day-to-day affairs. You'd earn r-rep by helping wash test tubes, but not for, say, mowing the Martian scientist's lawn. If you needed an extension on your fab materials quota, you could use your @-rep on almost anyone in the outer system, but your r-rep would only work on a scientist/technologist, and even then maybe only if you're planning on using it to advance scientific/technological ends.

i-rep is similar - you earn it for performing Firewall related activities, and that's where you'd mostly use it, too. An outsider might earn a small amount for helping a Firewall agent, but only if they were on the job at the time. At the same time, you might also give them some @-rep or c-rep (or both!), depending on where you are, and how secret your activities at that time were.

@-rep and c-rep are just much more general-purpose, and which you use depends more on who or where you are, rather than what you do, or are trying to do.

Aah. I worded that badly. Where people with rep on the network don't *necessarily* know what their reps are. You're right that a write-only rep network makes no sense, and anybody who actively participates in the network would have to have access.

The whole thing is probably tied up with crypto, forming a weak chain of trust. If someone trusts you, they'll share the private key with you, and as you get more trusted and/or become a bigger fish, you get access to more details. Low level might be "So-and-so has gotten so many positive ratings", medium level might give safer reasons, and high level gives the details that could get someone in trouble. Different regions might have different sets of keys, so that an outer system criminal might need to get an inner system criminal to vouch for them before they get access to the Martian g-rep servers, assuming that their reputation doesn't precede them.

Yeah, public key would work better. You might also be able to do what you're describing with key splitting. You're given pieces of keys, and once you have enough pieces, you can start decrypting (and encrypting?) certain things. Once you have enough key pieces, you can also learn about the necessary channels to get your own key pieces. Perhaps we can assume Sufficiently Advanced Mathematics, so that the key pieces that you're given are tied to your public key, so you'll need your private key to decrypt the result. This would prevent two or more people from just pooling key pieces, forcing each person to get vouched by enough other people.

The easiest way to do that would be to run a private key through a secret sharing algorithm to divvy it up, encrypt some of the fragments to a public key, and distribute them somehow. Barring someone throwing an EP quantum codebreaker at it, the character could then decrypt them with the character's private key when necessary.