Oh man, that takes me back to when I had unshakable confidence that I was always right, even when I knew nothing about anything. So nostalgic...

So, I get where your idea is coming from, and that it's coming from a good place, but I feel like an idea like this will be counter to several of the foundational ideals and practical realities behind WCG. I wouldn't hold WCG to action on this because:

1) Deadlines: WCG already has a method of making sure that results are returned in the timelines that best work for the scientists. It's called the deadline. The deadline varies by project. FAAH had a 24 hour deadline because anything over 1.5-2 days just slowed down the project too much. ARP has a longer deadline because the files are larger (many hosts can't even finish crunching them within 24 hours), and because the scientists and WCG's storage would be overwhelmed if they got slammed with results back too quickly. So, there are already measures in place to control how slow or quickly results are sent in, and 24 hours will not be a neat one-size-fits-all.

2) Moral hazard: WCG is a volunteer effort. If you ask for volunteers and then put high requirements on donating time/effort/whatever, you are going to get way less involvement. It comes across as being a choosing beggar. WCG will err on the side of encouraging anyone to contribute on their own terms. It's uncertain whether a "reliability" badge will encourage more people than it discourages. People who already contribute very reliably are the ones who probably don't need the extra motivation from a badge. This is why I would think a low-level badge for the first completion, or one day or something, would work better. It will encourage the people who are testing it out and maybe they'll contribute more once they get that first validation.

3) Discrimination or Demoralization: A reliability badge will discriminate against certain classes of people, who are good and valuable contributors, but cannot contribute daily through no fault of their own. People who have very small data caps or who are on satellite services may only be able to cache enough work to crunch constantly but only return work every couple of days. People who are in countries or areas with frequent outages (power, internet, phone, etc) may not be able to return work even if they are very reliable, all other things considered. People who crunch on work computers (after all, why not when word processing will not take up 4+ cores) may not get the badge because they don't work on weekends. People belonging to those groups may be demoralized, because they are contributing everything they've got, but it's not good enough to be specially recognized.

4) You bring up a point about badges being exclusionary. That's one way to look at it. But if you consider the graphic design choices, you might see that they're designed to be more inclusive and encouraging than exclusive. Why do you think that all of the lower badges are so colorful and rapidly progress along the valuable metals/gems? It's because the colors are fun, and progress is quick, and the precious metals/gems make the contribution appear valuable. Then look at the 5+ year badges. They're all the same, just with different numbers. And they stop at 100 years. At the point where someone has crunched for 5 years on a single project, the reasonable assumption might be that they don't need much external motivation to keep crunching, except for seeing the numbers get higher. And once they've crunched 100, they may not need external motivation at all to keep crunching. (I personally would like to see 250, 500, and 1k badges, because I'm on my way to them, but I'm not holding my breath. I'll still be crunching without them regardless.)

5) Reliable contributors already get some preferential treatment: If a host is considered to be "reliable" they are first in the queue to get resent work. This can lead to a disproportionately high amount of hard-to-obtain work units for some projects. This is justifiable because any other system would likely irreparably hurt the science projects' progress, but it does also provide a benefit that motivates to be reliable.

6) WCG back-end priorities: WCG would have to come up with brand new statistics to track regarding reliability, and make choices about how it is measured. Is it on a per-project basis, or is it cumulative? Is it measured by number of tasks reported in a day, or by millions of per-workunit calculations to determine average time per workunit, or something else? How long does it take to be reliable - is it a fresh calculation every day, or over a week, or month, or year? What happens when a host blows up and a bunch of units are lost or all come up as errors? Will any or all of this work even be appreciated by the community, or will it just cause people to complain? How much time should be redirected from implementing OPN GPU edition, bandwidth allocation with ARP, server maintenance, bug checks, error code mediation, etc? After all, WCG only has so many people and so much money to keep WCG afloat. And that's because...

7) Budget: WCG is a charitable effort from IBM. IBM is going to funnel enough resources to WCG to accomplish its purpose: generating PR and goodwill for IBM, and whatever tax breaks that IBM can write off because it's kinda charitable work. Unless something is vital to the projects, it is going to be lower priority than keeping the servers, websites, and PR machine going. This is probably why we haven't seen more badges for first results or 100 years and beyond: the people running WCG are probably just spread too thin to get around to it. Their priorities are therefore on the projects.

I say all of this not to shoot down your idea, but rather to readjust your expectations, and broaden the scope of considerations when thinking about these type of suggestions. I don't think the reliability badge is a good idea, per se, but even good ideas may not see action because they are unworkable with WCG's priorities or practical realities.

This sort of exists--would just need to have it appear on the forums and on the My Contribution page once you had one result successfully returned: https://www.worldcommunitygrid.org/stat/viewProjects.do. I call them the blank badges.


Also, let's remember that for nearly all projects on WCG (at least in the five years I've been here), the WCG techs don't send the results directly to the scientists when you finish them. They are stored at WCG until the batch is complete and then the batch is sent to the scientists. So you returning WUs a couple hours (or days) sooner does absolutely nothing to move the science along more quickly (unless you are such a large contributor that you are completing large percentages of batches on your own). It is you being here and returning valid WUs in any number that move the science forward.

It was Treasure of the Sierra Madre with Humphrey Bogart & John Huston.
https://www.youtube.com/watch?v=4OcM23Hbs5U

Did the Muppets do a badges show? That reference went over my head.

I like badges. When GPU projects start we're going to need them denominated in millennia.

OK, how about 'pyrite' ?

CJSL
Crunching away on the thin ice of a new day...

I guess you guys have been here too long to see this forum with fresh eyes. I come here and I see a near ghost town. This place is stagnant. It's the same group of people posting the same kinds of content over and over. It's a pretty small group too. Sometimes new people come in, but very often they leave soon after.

An example: I joined the Linux Team this week. The leader of my team hasn't posted anything here in 4 years. He still crunches, you can see it on boincstats.com, but he isn't part a community and therefore he is a team leader in name only.

If you guys don't embrace change soon you will fade away.