Large source operations may be able to alter the pass/fail statistics quite a bit. You might need to limit how often a source's products can be sent in.
Please elaborate on this.
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
Large source operations may be able to alter the pass/fail statistics quite a bit. You might need to limit how often a source's products can be sent in.
It will probably be a moving target. The shill is a determined individual. They will keep finding loopholes and we will keep closing them.I've considered the shill problem. Do you think requiring a minimum number of points or likes would solve it for the most part?
Having said that, the likes system is not perfect since @Astro Labs has more likes than most members. There's something wrong with that picture especially when you look at the valuable contributors below his name. Members need to re-evaluate their relationship with sources. No source should be on that list.
I've considered the shill problem. Do you think requiring a minimum number of points or likes would solve it for the most part?
Having said that, the likes system is not perfect since @Astro Labs has more likes than most members. There's something wrong with that picture especially when you look at the valuable contributors below his name. Members need to re-evaluate their relationship with sources. No source should be on that list.
I just realized small labs are not really at a disadvantage here. As long as they can flood the test environment with shill members, they can change the odds on who gets to send samples for that source's products. The shills get the legit gear, and everyone else gets under dosed gear.Please elaborate on this.
It will probably be a moving target. The shill is a determined individual. They will keep finding loopholes and we will keep closing them.
We also have to consider source to source attacks. Submission of bunk gear with another source's label. Once this happens a few times, no one will trust test results for gear sent in by new members. The limiting criteria might not be fully controllable by any set of specific rules.
While submission will be anonymously discussions between us on who sent what and from where will always hold the most value when it's coming from trusted members . While anyone can send samples when someone such as yourself says you sent compound whatever and the results are this, would hold more weight to me then any random testing.
Thanks for the vote of trust. I think the numbers Millard has been tossing around would be a definite game changer in many ways. One would be the influence of the "trusted" vet would be severely limited against a swarm of random test results.
Trusted members (i.e. those with significant influence) aren't necessarily the best individuals to send in samples. I don't think an admin/mod, no matter how trusted or highly regarded, should be the first choice to submit samples. Those are the guys who always get the best products, the best service and preferential treatment. Sources know they have influence and source know to treat them well. This will be especially true if the "trusted members" are the ones responsible for submitting samples for testing. As @flenser said, the likelihood that they will be "selectively not-scammed" is very high.
The only way for these guys to avoid this is to create an additional, unrelated handle and address that is not linked to their better-known pseudonym.
It's not that the trusted members are not trustworthy. It's that sources are eager to exploit the community's trust in those members for their own purposes.
I just realized small labs are not really at a disadvantage here. As long as they can flood the test environment with shill members, they can change the odds on who gets to send samples for that source's products. The shills get the legit gear, and everyone else gets under dosed gear.
No doubt that this will be an issue and exactly why the credibility of who sent in samples would greatly influence any of my decisions.
This may be inevitable. Just like the inevitable shills that participate in the steroid underground. But shills are regularly exposed in the forum and I think they can be similarly exposed in a steroid testing program.
Let's assume that 1000 members submit 1000 samples. Also, let's assume 5% are shills with 50 samples either false negatives or false positives. What steps can be taken to identify potential shills?
Meso already has a lot of members with amazing shill detection skills.
That's what I think will be a moving target. Considering a shill will need to run lots of members to be effective, he won't be able to spend much time with any one of them. Meso already has a lot of members with amazing shill detection skills. They should detect these more easily as long as selection is in the open. If it's an automation that doesn't report who it selects, or reports them too late, I think shills could be a lot more effective.
I know I am new here and have no credibility whatsoever but I have been burnt by a source here and I am willing to help out any way possible. I am deff interested in this.