Comment by jackgavigan | Hacker News Reader

Comment by jackgavigan | Hacker News Reader

For context: I'm sat on a trading floor in the City of London right now.

I had a similar idea (inspired by Slashdot) about ten years ago but, like so many of my great ideas (e.g. time travel, teleportation), I never got around to implementing it.

A site like BoredBanker.com will almost certainly be targeted by wannabes, trolls and banker-haters. If you don't keep them off the site, real bankers will be turned off by the low SNR and abuse.

If it were my site, I would put in place a mechanism to filter them out at registration. People registering with the site would be required to supply an email address. Anyone who could provide a Bloomberg account (i.e. an email address @bloomberg.net) or an email address at a domain on a whitelist of known financial institutions (e.g. morganstanley.com, citi.com, gs.com, etc.) would automatically be accepted (and potentially given downvoting rights straight away) and so would anyone they invited/recommended/vouched for.

Anyone else would either (a) go into a queue for approval or (b) be given "on probation" access. If (a), someone would need to review the applications (presumably by checking the domain name of the applicant's email address (and, if appropriate, adding that domain to the whitelist) and/or reviewing their LinkedIn profile. If (b), there could be a feedback mechanism that would automatically hellban the user if they received enough downvotes (or were flagged by a moderator), or promote them to "approved" if they received enough upvotes.

You could also allow "approved" users to hide comments from "on probation" users (similar to how Slashdot used to - may still do, for all I know - allow you to set a minimum karma limit).

That way, you make it really easy for legitimate finance professionals to join, while putting hurdles in place to make life difficult for the trolls and Occupy-ists. And you also generate a bit of scarcity-desirability through the invite mechanism.

Of course, as soon as it becomes successful/popular, you'll probably find that the site gets flagged for filtering out by the web proxy filters the banks use. :-/

I think the best example of recruiting good members is the SumZero model. SumZero is a buyside investor site, where buysiders can exchange and rank investment ideas with each other. It was started by Divya Narendra of Facebook fame.

When I first signed up for SumZero, the application process was pretty extensive. It involved a phone call with Divya or one of the other co-founders to check you out and make sure you weren't just some wannabe, that you actually worked at an investment fund.

I actually think was critical early on to make sure the membership base was high quality and acted as a self-reinforcing mechanism to continue to attract good members.

Is SumZero a community? It looks like a broadcast medium for research reports.

Related to this, take a look at how lobste.rs handles new members: you either need an invitation from an existing one (and your behaviour is associated with that person in the public tree of membership referrals) or you go into a member-visible queue until someone who knows you and is a member can invite you personally.

That's very clique-y.

And This appears to be the lesson from the life cycle of High Signal to Noise ratio sites.

Initially there is the bleeding edge, and then there is the long noisy end.

Net result: nobody uses lobste.rs.

10 points to Gryffindor!

It sounds snarky, but I don't mean it that way. I like the guy who runs lobste.rs and I'm happy he published the code, and I'd really like HN to have some real competition. But, sorry, nobody uses lobste.rs. It's a ghost town.

For what it's worth, I think the answer to site quality isn't pre-filtering users; it's having a simple process to nuke bad users. Lobste.rs put a big bet on pre-filtering; it hasn't worked out. I'm not surprised; one thing I look for on great HN threads is the random first-party (owner of company, primary author of package, &c) commenter who joins just to comment on that story, which is something that can't happen in the pre-filter model.

Since bankers know who other bankers are, a system like the mod_virgule that powers Advogato [1,2] would solve this problem, though the site would need some tweaks [3] .

The attraction of the system is that, if you keep an eye out for clever attacks against the model (i.e., something like security fixes), and if certification activity by "good" users is essentially sane, then such a site needs no ongoing moderation by the site owners.

[1]: http://svn.dprg.org/viewvc/mod_virgule/trunk/

[2]: http://www.advogato.org/

[3]: The codebase hasn't been touched, and there's a form of attack that, while it is known how to solve it, has not been fixed on the main site; also the UI has some cruft and doesn't support an HN-like leaderboard view - the codebase is fairly simple, though, and lots of people have hacked on it over the years.

One alternate mechanism (there could be several) sure to ruffle feathers is to qualify users by their GMAT score (e.g. 700+ only), assuming such information were easily verifiable.

Such a 'qualification club' is inline with my own view of the type of exclusivity sought after by sterotyped or caricature B-school personalities.

I take it you have very little experience with the GMAT. It is impossible to verify that a certain GMAT score belongs to an email. The only easy way to verify a GMAT score is to be a b-school and have applicants select your school for score reporting.

There's such a massive opportunity there for an organisation like GMAT to use PKC for this kind of thing.

For you and me maybe, but not for GMAT. It is my experience that the primary motivation for big organizations is to get their beaks wet. GMAT charges for additional score reports after the initial 6(?) that are included in the test fee.

Well, they could charge for signing your public key...

The above suggestion is an idea. Understood that today GMAT might not allow such functionality, but the organization has the data internally to send score reports.

Correct, not supported today, but feasible with business development.

How is GMAT going to monetize the easy score verification? And is that effort going to be worth the additional revenue? I can't imagine there is a big market for GMAT score reporting outside of b-school apps. I think most employers just accept self reporting or a copy of the score report in pdf.

I wouldn't think it'd generate very much money to directly monetize easy score verification other than as an additional data point for advertisers.

However, I do believe easy score verification could be part of an overall strategy to increase the GMAT business. The thinking is that if society started using indicators of intelligence for various reasons (e.g. entry into an exclusive online forum, government civil servants employment, running for Congress...) the indicator of intelligence that is preferred by society could sell more tests. GMAT could create an advantage in the marketplace by being the easiest score to electronically verify.

I do not think the GMAT is a good indicator for civil service positions. Certainly no better than any of the state civil service exams and at the federal level the civil service exams were deemed to cause problems with affirmative action policies. The military has the ASVAB and I can't imagine anyone saying the GMAT is superior to the ASVAB. The small "d" democrat in me finds exams for public office abhorrent and I would hope that they would run into many legal challenges.

I scored a 720 on the GMAT. I do not think you should make any assumptions about my performance in a government position from that test result. All that the 720 indicates is that I am a skilled test taker.

Interesting! I was skeptical of the GMAT until I studied and took the test. My personal opinion is GMAT is a very difficult test and studying for the test sharpened my own writing, logic, and reasoning skills. I am very impressed with the cognitive ability of the test takers in the very highest percentage.

For testing logic and reasoning I think the LSAT is the best as far as the popular standardized tests go. I think the ASVAB is the best all around test for getting a general idea of an applicants skills.

So, if not tests, then how? (just curious). Personally i'd prefer a system where u need to pass a difficult academic test to get into the govt, rather than have connections etc. (Yes, IMO if no testing criteria is involved, its all up to connections).

Is GMAT an American thing? I'm in Australia and have never heard of it.

Good ideas, but I wouldn't give pre-screened users the keys to the kingdom right away. I think it's an unwritten law that trolls and sockpuppets are going to exist in every industry, and just because someone comes from company X doesn't mean that they're going to be a quality user. You could relax the requirements to attain full membership perhaps, but I sure wouldn't give them the means to easily manipulate the system at-will with nothing more than an email address.

I think you'd have to have a policy to avoid particular items that get Compliance reaching for the BLOCK IT button - eg strict policy against people offering trading recommendations, for example.

I agree, a similar system is put in place at Quibb, it works quite well.

it's not a coincidence of the audience and the topic that the top comment is basically just scheming of ways to make it more elite and exclusive instead of how to create a self-policing meritocracy.

why do you think HN is successful? did they limit registrations to @yahoo.com, @redhat.com, @microsoft.com, @google.com, @venturecapitalisthere.com? is that why the SN ratio is high?

and besides, doesn't bloomberg already have forums and chat for this exclusive bank crowd already? what is the point of recreating that on the web if someone like me can't see it?

Reply on news.ycombinator.com