11/18 - Miserable Monday

Discussion in 'Daily mTurk HITs Threads' started by Mothra, Nov 18, 2019.

Thread Status:
Not open for further replies.
  1. lilmsmisses

    lilmsmisses Active Turker

    Messages:
    508
    Gender:
    Female
    Ratings:
    +325
    Elizabeth Howlett [A3KV5BJ1V6DBH1] Numerical Judgment - $0.30 | PANDA


    Unrated

    Unrated

    Unrated
    $7.50 / hour
    00:02:24 / completion time
    Pros:
    Cons:

    Qualifications: HIT approval rate (%) > 90; Location EqualTo US;
    To read Elizabeth Howlett's full profile check out TurkerView!
     
  2. morweeg

    morweeg Triple Double Certified Organic Batch Master

    Messages:
    40,571
    Ratings:
    +54,131
    Amazon have lots of money. They no pay taxes. I have no money. I have to pay taxes. Seems to me that Amazon should pay more in taxes so that I don't have to.
     
    • LOL LOL x 1
    • 5/5 Pay 5/5 Pay x 1
    • Love Love x 1
  3. morweeg

    morweeg Triple Double Certified Organic Batch Master

    Messages:
    40,571
    Ratings:
    +54,131
    Also @ChrisTurk change your pinned tweet already.
     
    • Like Like x 1
  4. jessers

    jessers Survey Slinger

    Messages:
    14,696
    Gender:
    Female
    Ratings:
    +44,726
    So, each panda is like searching mturk manually. If you go to the hit and press "accept" manually, the panda is trying to do that for you. So imagine each panda saying "I want that hit" and trying to press that "accept" button.

    Your account is only allotted so many searches in a period of time. I don't have the specific numbers, so this is completely for the sake of an example, but let's say you can get five mturk searches a minute. those five searches are 2 hitforker refreshing every 30 seconds, and 3 pandas running. 5 pandas a minute (edit from CT: Just a minor clarification (b/c it actually makes examples easier in the future). Pandas & Searches are no longer considered the "same thing" on MTurk and so the page request pools are isolated from each other. At least, if you're doing anything remotely normal. Frankly I'm not sure about like 3ms interval pinging of either endpoint.)
    That number is ridiculously wrong, but again, just for example's sake. So if you have 30 going at once, it's only trying to press the "accept" button on each one, one at a time, so every couple minutes.

    And if you "search" too frequently, too quickly, panda crazy will slow itself down.

    [​IMG]

    Here, each panda is trying to accept every 1165ms. PRE'S is mturk saying "you're going too fast" because I went over that allotment 46 times. If you're hitting that allotment, the "elapsed" will go up, sometimes to a few full seconds.

    I feel like I made it more complicated now.
     
    • Today I Learned Today I Learned x 3
    Last edited: Nov 18, 2019
  5. IcyTundra

    IcyTundra Tundra Thundra

    Messages:
    9,022
    Gender:
    Female
    Ratings:
    +17,080
    Thanks for answering these questions @Tracy Wethington - I have been only using MTS and been afraid to add Panda Crazy and Hit Forker because I think early on I thought I saw something that said it would screw things up and that they wouldn't work together. I'm still not sure which ones are the best to be using and whether I will blow up my computer into a million pieces.
     
    • Love Love x 1
  6. ChrisTurk

    ChrisTurk Administrator

    Messages:
    56,679
    Ratings:
    +162,927
    Honestly that was just a complete dumbass luck example I clicked over to twitter to look and it was legit the 2nd thing.

    Prolific has grown to what is now probably close to 1MM MRR just by saying "we're ethical" and then one academic repeats it and 30 believe the first person who said it even though it had a referral link attached to it.

    I don't think any of these folks have lived in like normal reality where your broke ass 2nd cousin is hocking herbalife on FB to realize anything with a referral link probably isn't something you should actually read seriously. I have zero actual explanation beyond that to understand what I've been watching play out the last year or so. And @THFYM can vouch I've taken a serious deep dive and it's.. scary.
     
    • Love Love x 2
  7. turks24

    turks24 Survey Slinger

    Messages:
    6,720
    Gender:
    Male
    Ratings:
    +13,777
    If you're honestly curious as to why it happens, picture this scenario.

    2017 - you lose $500; you pay no taxes because you didn't make any money
    2018 - you make $400; you pay no taxes because you get credit for the money you lost the year before

    compare that to:

    Jan - October: no work
    November- you lose $500
    December - you make $400

    pay no taxes because you didn't make money this year

    Taxes are wayyyyy messed up, for sure. Carry over loses? In the vast majority of situations I don't think there should be any objections
     
  8. Melting Glacier

    Melting Glacier PE: $30.01 - That's over $1.25/hour! ┬┴┤( ͡° ͜ʖ├┬┴

    Messages:
    6,157
    Gender:
    Male
    Ratings:
    +11,423
    oh hey I caught you, should I do a tuesday thread or not? Can't tell if weekly rundown mentioned is for Monday or a TV review Tuesday thingy too :flail: and also WHERE DID THE SUNDAY THREAD GO AHHH :untzuntzuntzflail:
     
    • LOL LOL x 2
  9. Tracy Wethington

    Tracy Wethington The Bone Crusher

    Messages:
    13,955
    Gender:
    Female
    Ratings:
    +22,085
    I prefer forker and crazy. Forker is a bit better in my opinion. :)
     
    • Today I Learned Today I Learned x 1
  10. morweeg

    morweeg Triple Double Certified Organic Batch Master

    Messages:
    40,571
    Ratings:
    +54,131
    For a lot of beginners, it's easier to just use MTS scraper + catcher because it doesn't overwhelm with the amount of features. Once you're familiar with what the scripts do, there's no real reason not to go Forker + PC.
     
    • Today I Learned Today I Learned x 1
  11. ChrisTurk

    ChrisTurk Administrator

    Messages:
    56,679
    Ratings:
    +162,927
    Yea you can post tomorrow, the stupid thingy failed on me anyway which I knew was gonna happen but I tried

    I moved the thread because :tinfoil:
     
    • LOL LOL x 1
    • Today I Learned Today I Learned x 1
    • WOW WOW x 1
  12. afallstarxfall

    afallstarxfall Well-Known Turker

    Messages:
    8,763
    Gender:
    Female
    Ratings:
    +878
    Title: Past Experience Study | PANDA
    Requester: Rebecca Cipollina [ARJ0V578GVAZL]
    TurkerView: [ $6.61 / hour ]
    Description: Study where you will recall a past experience and report on questions about it
    Duration: 60 Min
    Available: 0
    Reward: $1.00
    Qualifications:
    • Total approved HITs GreaterThan 100,
    • HIT approval rate (%) GreaterThan 95,
    • Location EqualTo US,
    [3EJ1F91AOVX5WELGRU7C2WPIZ71IZR]
     
  13. turks24

    turks24 Survey Slinger

    Messages:
    6,720
    Gender:
    Male
    Ratings:
    +13,777
    Bro, for the sake of your sanity....
     
  14. Razr

    Razr Jive Turkey

    Messages:
    2,270
    Gender:
    Male
    Ratings:
    +2,723
    I started with HF & PC but I ended up getting a rejection bc I had already done a hit so I switched to MTS because of hit tracker
     
    • Today I Learned Today I Learned x 1
  15. WelcomeToFrightNight

    WelcomeToFrightNight Active Turker

    Messages:
    191
    Gender:
    Female
    Ratings:
    +132
    Dear God!!! Will someone Dr Kevorkian me!!!

    I will just have to pay someone cold hard cash for like an hour's tutorial on scripts. I think $40 would be fair right?, and once au fait with these Godammed things I can make that cheddar back in about 3 hours.
    Someone please make a video and sell it to me, it has to be comprehensive though and I will pay you through Paypal...or Amazon gift card:emoji_joy:, who's game?
     
    • Like Like x 1
  16. jessers

    jessers Survey Slinger

    Messages:
    14,696
    Gender:
    Female
    Ratings:
    +44,726
    Luckily for me, I can claim VRAPI and others as an actual job on a resume. But the line blurs fast with trying to claim you work for google just because you did some zoltars once, or work for pinterest because you do pins. The problem shouldn't be "how can I make mturk look better for myself", but more so "why don't people see this as genuine work?" when it is absolutely genuine work.

    And again, most importantly, the requesters who are basing their research/work on here can be thrown out the window because the platform looks like worthless garbage with a bunch of homeless mentally ill people sticking their thumbs up their butt all day for $5. People are contemplating taking their work off of it just because it's been delegitimized, and that's a huge problem for us.
     
    • Love Love x 5
    • LOL LOL x 1
    Last edited: Nov 18, 2019
  17. THFYM

    THFYM Survey Slinger Former MTG MotM

    Messages:
    8,887
    Gender:
    Male
    Ratings:
    +23,050
    I used to think of intelligence as a linear scale. On one end, there's dumb people. On the other, there's smart people.

    Academic Twitter(TM) taught me that it's not like that at all. It's a circle. The smarter you get, the more clueless you might become, apparently.
     
    • LOL LOL x 6
  18. morweeg

    morweeg Triple Double Certified Organic Batch Master

    Messages:
    40,571
    Ratings:
    +54,131
    You can use hit tracker in addition to the other two things. MTS is totally worth installing for other features like that even if you're a Forker + PC main.
     
    • Today I Learned Today I Learned x 1
  19. IcyTundra

    IcyTundra Tundra Thundra

    Messages:
    9,022
    Gender:
    Female
    Ratings:
    +17,080
    This is extremely helpful.
     
    • Love Love x 1
  20. morweeg

    morweeg Triple Double Certified Organic Batch Master

    Messages:
    40,571
    Ratings:
    +54,131
    Also, that rejection was 100% bullshit. Most people won't go back and doublecheck if they did a survey before and when you're a certified A grade Survey Slinger doing dozens of hits/day, you aren't going to remember every survey. That was 100% on the requester for not using exclude quals.
     
Thread Status:
Not open for further replies.