Next time I'm on I'll pay attention to the thread. I've been using mts suite, and the finder, after I discovered that it was a bit help, but I'm still confused by the colors and what's going on. I'll click on a hit that just showed up, but then it's already taken. Thats been one of my biggest frustrations.
None of the hit finders that I've used can reliably pull data from the review sites (turkopticon, or turkerview) but the idea is that a good requester with good ratings will show up as green, and requesters with a bad reputation would show up as red, with yellow being an average requester.
The ratings system is good information for the most part, but until you become more familiar with the caveats it's a good idea to click on ratings and reviews yourself to start to become more familiar with them.
When a requester hasn't been rated by anyone or the scraper fails to pull the ratings data from other sources, it shows up as gray.
You will get used to hits showing up for a second and quickly disappearing. For the most part, there are more workers than good hits, so anything good as a lot of competition.
That's where the advantage of having a second program comes in, or using the built in feature in mts to automatically grab hits.
I use PandaCrazy partnered with panda crazy helper.
When you find a hit that you would like to do, you can click on a button to add it to panda crazy. Panda crazy will search for the hit on a predetermined schedule (like once a second) and attempt to grab it without you doing anything.
Some surveys ask screener questions, and will require you to return the hit, sometimes people let hits expire, and requesters may republish the hit after a short while.
It's rare to catch most good hits either "raw" through the mturk search engine, or on your first attempt, unless it's a batch.
You can also use pandacrazy to build and manage your queue. When a good batch drops, you can let panda crazy continue to accept the hit, while you are working on one just like it or finishing up a different task.
The quick difference between the two is how you tell pandacrazy to start collecting the hit. For surveys you will just want to collect it once, where with a batch you can tell it to collect it as many times as possible.