Hacker News new | ask | show | jobs
by jaydles 4548 days ago
DON'T TRUST ME BLINDLY: I work for Stack Exchange, so I'm totally biased. On the other hand, I left a lucrative career in finance for a lot less money here because I believe in what we're doing, so there's that.

MOST IMPORTANTLY: I appreciate Michael's feedback, and he worries about a lot of the same things I do. Moreover, we are incredibly grateful for all he's done over the years - my honest belief is that his contributions (even when they were just fish) helped a ton of people finish a project that may have been what made them LOVE programming. And those people did take the time to learn the fishing techniques underlying those fish, so they could do it better next time.

ON REWARDS: Points aren't the point. Let's be honest. We reward people for helping others with points that essentially convey nothing other than the ability to help in new ways (as you unlock new privileges). No one in their right mind is spending time on the site with the empirical goal of getting points.

The real reason people answer questions is that they like helping people. The points are important, but only insofar as they give you actual feedback on how many people appreciate your effort. The points aren't the reward; they're just a way to measure the real reward people care about: knowing how much of a difference you've made.

So when Michael worries about his points going up even after he's stopped posting, that's the system working. It's not about ensuring the right person is "winning" it's about showing how many people got help.

And he's still helping others today. I respect his decision to leave, but truly think he should be proud of what he's done for the programming community to date. In any case, we're grateful.

6 comments

I really think SO is a gamechanger and I agree with you mostly, but I do think the points system is broken. SO offers all of these points and statistics that catapult people to positions of visibility, bestow clout, and unlock abilities on the site, and you're saying that they don't matter? They do. As a new user, it's practically unachievable to enter the top 10%, because all those people are still gaining points at an incredible rate, even if they're not actively contributing. It's a classic "rich get richer" dynamic. If the points really aren't the reward, then maybe they shouldn't be displayed right next to the user's username.
I disagree that it's practically unachievable to enter the top 10%. I did it in my first year (2 years ago) without even knowing that they kept track of that sort of thing. Answer a single question a day (correctly) and you'll get there easily. You'll actually be in the top 3-5% if you do that.
Add a custom search option that can filter people by reputation level, or by question to answer ratio.

I want to answer questions from people who are able to Google, and who are able to answer questions.

Another thing: add a custom search option that can find tumbleweed questions. I asked about it here:

http://meta.stackoverflow.com/questions/14738/tumbleweed-que...

But Jeff thought it was already implemented, when it's not. The advanced search finds questions with a minimum number of views, not a maximum number of views.

You may be able to implement this yourself using the data explorer:

http://data.stackexchange.com/stackoverflow/query/new

You could do that as a direct step towards finding the questions you desire, as as a rhetorical move towards getting it implemented in SO.

> No one in their right mind is spending time on the site with the empirical goal of getting points.

But isn't that exactly what Michael was doing? He was posting help for two languages he loathed — help he probably wouldn't otherwise give — just because it got him points.

In addition, the Java help he was giving was not even his own knowledge. Just a Google search.

I like that you believe points aren't the point. That's good to hear. However, my experience on SO is as Michael describes.

The most helpful answers I've posted (especially the ones that are obscure workarounds for difficult bugs) have gained very few points. Most of the time I've gotten good answers to very difficult questions is when I either: answer them myself later, or offer a bounty. People strongly react to the points.

That's not to say SO isn't helpful. It helps me five to ten times a day — I find code snippets that achieve the effect I'm looking for. I could spend more time and read the documentation, figure out the right APIs, but that's a waste of time when I develop the same understanding by reading someone else's implementation.

It would be great if the point system scaled by the complexity of the question and answer. Though that seems impossible to determine.

Yup, the real wedge was wrapping a free service in a really slick interface. Points are just icing.

Another poster made a very good point about voting ring vandals and spammers, but I'm concerned about "expiry dates". An increasing number of the top SO results in google are answers that were correct but are now wrong. Things will trundle along for another few years, but at some point the correct results will become un-googlable.

Can you provide some examples of out of date posts?
Take a look at Android questions especially. The API has changed so much recently that many answers are obsolete.
Should they be deleted? Edited? I'm not sure I want to wade into the SO quagmire to fix them.
Thank you for all of your hard work, and the work of the SO team. Helping us share all these fishes around.
One may like helping the people generally, but real-time monitoring of unanswered questions for posting the FIRST answers.... that's something different.