Can't Find Wonderful Hits? 5/15 Deflated Football Friday!

and to elaborate, adwerx/todd mosier's new instructions: https://www.mturk.com/mturk/preview?groupId=3SFRD5IA5Z4458LR2TVQF3HWRY1BFR

clearly a result of a script frog posted that skipped his entire original first step of visiting the office site FIRST before image searching (frog's script only did the image search and he was encouraging people to do them in <10sec)

so shitty frog scripts = shitty data = unhappy requesters = shittier instructions (ie n/a no longer allowed) and mentions of rejections and shittier pay (because n/a no longer allowed after being flooded by the frog script)

 

Masters, this was painless and quick if you know how to follow instructions.

Title: Usability Study For BlockStudio: Phase 2 Requester: Center for Game Science [A29TXPX00CTRVG] (TO) TO Ratings: ☭☭☭☭☭ 3.67 Communicativity ☭☭☭☭☭ 3.24 Generosity ☭☭☭☭☭ 4.08 Fairness ☭☭☭☭☭ 3.21 Promptness Number of Reviews: 26 (Submit a new TO rating for this requester) Description: Test a new game-authoring app (45 minutes) Time: 60 minutes Hits Available: 2 Reward: $5.00 Qualifications: Masters has been granted, Total approved HITs is not less than 1000, HIT approval rate (%) is not less than 98, Location is US

 

If people could do that then I wouldn't have an RTFM avi.

 

I see the 20 cent CS hits are back.. eh.

 

Look bud I'm trying my best to get my grade 10 can you give me a bit of credjudice instead of these false acriminations I don't have enough people words to make it understand you the way it understands me and the more thinkings I have about it I can tell it doesn't take rocket appliances

 

for to gooder it up, that is for to a man

 

Script for CrowdSourse, pleassseee!

 

They Don’t Think It Be Like It Is, But It Do

 

and to further elaborate...

I've always told people that taking shortcuts (scripting is an example) is optimal only if the data submitted doesn't really stray from what would've been harvested using the provided instructions (otherwise bad stuff happens like my last post). Also if the quality is higher it's no longer optimal ($$-wise because fewer n/a's) but the instructions/pay don't deteriorate. The only way to optimally shortcut hits that would stray datawise (upward) is if founds and n/as take the same amount of time to harvest (with the bonus of data quality being maintained). So you shortcut if you're optimizing for $$ only if it decreases, not increase the n/a's, and only if they take the same time, otherwise if it decreases you're optimizing for quals (since time got added), and if it increases you're optimizing for nothing (adwerx example)

edit: left out the obvious third case, neither increase/decrease in n/a's = optimizing for both quals/$$

 

FFS, what is it with most of the people who rate pay on TO?

I just wasted my time on a requester that had tons of glowing 5/5 reviews on pay (at a quick glance) only to find first-hand the pay was a joke.

So then I start reading the ratings more in depth, and here's an example:

"Really loving this requester. The base pay is 0.05, BUT what makes these tasks worthwhile is the bonus. ... a $0.22 bonus, which took me all of 5 minutes.

Honestly don't let the 0.05 scare you away. Or do, as that will leave more cash in my pocket."

"Worth it" = 27 cents for 5 minutes?

Don't be "scared away" by $3.24 an hour?

Just... Wow.

 

Agreed. It does have pictures, though.

 

This is it but be advised someone reported getting a bunch of rejections on crowdsource keywords in the last couple of days. He suspected it was for going too fast but won't know until they respond.

https://greasyfork.org/en/scripts/6232-jr-crowdsource-keyword-ranking-mturk-com-scripts

 

Depends on how shitty of a mood I'm in but I'll usually leave a comment for those people asking them why they rate it a 5 on pay and do they enjoy working for 18th century wages.

 

I've been really frustrated by TO lately. Most people's 5/5 seems to be more like a 2 or 3 for me. I don't even know what it would take for some of them to give a 1.

I see the same rating inflation everywhere, though. When shopping on Amazon or other sites with user reviews I don't even look at products with 3 (out of 5) or fewer stars. The actual range of useful commentary and opinion is 3-5 stars, so 3 is really equal to 1 and 5 is maybe a true 5. It's annoying but TO seems to be exactly the same way. Until some more objective measures can be incorporated it will continue to be this way, with people all too happy to take the scraps from the employer's plate.

 

And places like IMDB are even worse. The real range there is 7-10. People seem scared to rate anything lower than that. I do not understand it.

 

Title: Donate this $0.05 to the {GIVE A RICH PERSON MY NICKLE FOUNDATION} Requester: Lee [ A1B4KC5KP7FH1Z ] (TO) TO Ratings: ☢☢☢☢☢ 5.0 Fairness ☢☢☢☢☢ 5.0 Promptness ☢☢☢☢☢ 5.0 Generosity ☢☢☢☢☢ 5.0 Communicativity Number of Reviews: 42 (Submit a new TO rating for this requester) Description: Qualification: {Ramen Poverty} Has Been Granted Time: 33 minutes Reward: $0.05 Qualifications: None

 

I think most of the people leaving these good reviews are new or they turk as a hobby so they don't know any better they are just happy to make chump change

 

Oh absolutely...TO is annoying. I used to give just a cursory glance at the numbers for pay without reading the reviews much. But these people are ridiculous with their ratings. I have standards damn it!
(albeit low enough to work on porn hits for pennies, but standards nonetheless)

 

Yeah, my mom does the 411Richmond tasks because she thinks it's fun to see what people are buying. No need for the money, and she probably breaks 50cents a day tops. It's just for fun.