Penguin Watch Talk

Too many to mark vs blocking some out

  • Shiphrah by Shiphrah

    APZ0004i9p Every time I see images like this one, especially this cite, I worry a bit more. What should I do about those faraway penguins, so tiny and clustered together that several fit between the crosshairs of each mark? Mark a few, perhaps a dozen, at the edges of the cluster? Mark one or two? Mark as many as I can, knowing that I'll probably be off by a factor of 2 or 5 or more? They're simply impossible.

    Meanwhile, I can do a pretty accurate count of the ones in the foreground. But that number gets swamped by the far bigger number in the background. I keep wondering what use this data can possibly be. There must be huge variations among the different people who mark this image.

    In earlier data, these areas were generally shaded out. That made sense. If for some reason that's not possible now, I wish we had clearer guidance about how to handle the distant parts of images like this one, beyond just clicking "too many to mark".

    I'm sure there are many users chewing over the same questions, and they may have posted about it, bit I missed their posts. Anyway, thanks for listening!

    Posted

  • gardenmaeve by gardenmaeve moderator

    Have you checked the FAQs, Shiprah? Have a good look, and you'll find what you need. 😃 http://talk.penguinwatch.org/#/boards/BPZ0000007/discussions/DPZ0000aay

    Posted

  • Shiphrah by Shiphrah

    thanks, @gardenmaeve!

    I have been studying the FAQs carefully. The relevant section in the one you reference says:

    "If there are more animals on a distant hill who look very small and are probably penguins but they're too small to be marked, just mark at least one of them and then click on the 'There were too many penguins to count' button to let the scientists know that more penguins might be there."

    I'm asking for further definition of the phrase "mark at least one". Does it mean "only one" or "as many as possible"?

    In previous discussions @yshish and I have both said we mark all we can, and that others do as well. In an image like this one, I'm pretty sure most of the ones I see are penguins. When they're spread out, especially when they're on the snow, I have no problem marking them.

    I'll restate my question: In situations like this one, when there are many tiny faraway penguins grouped closely together, which I should I do (in addition to clicking "too many to mark"):

    a. Mark only one or two in the center of each clump

    b. mark several on the edges of each clump to define the approximate area

    c. mark as many as I can discern, even though I know I'm missing many.

    Option c has been my usual procedure in the past, when marking that big crowd of kings for example, in which I am glad the rear is shaded out so I can see which area is more useful to mark. But this new data has a lot more tiny distant penguins and I'm trying to figure out how to manage most usefully. With images like this one, I'd much prefer option a or b, but would like do whatever serves the project best.

    Thanks!

    Posted

  • AvastMH by AvastMH moderator

    Hi Shiphrah,
    Thanks for the question.

    We always like to get as high a count as possible. Doting around the distant section, to mark those penguins that you are confident of, is the best way. And it is certainly what I do. I also use the 'hide old marks' button to help to mark distant penguins - sometimes those crosses can be so bright that counting is affected. I'm very fortunate in that I work with a very large screen. But it is concentrated work getting those '3 pixel' penguins marked up.

    The clue to why we say 'please mark at least one' is that we realize that some people will not wish to mark well over a hundred pengs (for example, children working on a school project), and may not have the same facility of screen size in order to do so. The scientists decided that by offering a limit of thirty, and given that several people mark each picture, they should get a good representation of what's really going on. And working on from that we do wish to encourage volunteers to try to show that those distant penguins are there, so encouraging one to be shown will help.

    I hope this helps!

    Thank you for your splendid work - it is very much appreciated!

    Joan

    Posted

  • yshish by yshish moderator, translator in response to Shiphrah's comment.

    Hi @Shiphrah

    I'm sorry that the FAQs aren't clear enough to you. I really tried my best to keep them short and simple.

    I've asked the scientists to check this thread to make sure I won't mislead you anyhow.

    Please, let me know if there are some more unclear sentences in the FAQs so I could redo them to make them more clear. If you find any more, please, mention them under the FAQs: discussion thread.

    Many thanks!

    Z.

    Posted

  • Shiphrah by Shiphrah

    Thanks to you both! I guess, to sum up, my best policy is #c -- mark as many as I can, even though I know I won't get them all.

    @yshish, I'd suggest a slight modification to that FAQ:

    "If there are more animals on a distant hill who look very small and are probably penguins but they're too small to be marked, mark as many as you can, at least one, and then click on the 'There were too many penguins to count' button to let the scientists know that more penguins might be there."

    I didn't mean to criticize the FAQ, I was genuinely unsure what to do. In general the FAQs are very very good! It's just that I am an editor and former English teacher, and often see implications in a text that others pass over. please forgive me if my comments seem nit-picky. It comes with the job!

    Posted

  • yshish by yshish moderator, translator in response to Shiphrah's comment.

    OK, thanks!! You know that my English is pretty bad and it's not easy for me to express my thoughts as well as in my native language:/ Any corrections are very welcome!

    Z.

    e: That sentence has been changed 😃

    Posted

  • gardenmaeve by gardenmaeve moderator

    Pardon me, yshish, but I must disagree. Your English is quite good.

    All of us want the FAQs to be the best they can be, despite having been created by actual humans exhibiting (shhh) imperfections. 😄

    Goodnight, all!

    Posted

  • mkmcguir by mkmcguir

    @yshish Your English is pretty darn good, and the FAQs are great!! I teach English as a Second language, so while there may be some mistakes, you should be proud of learning English so fluently. Many of my students struggle for many years.

    Posted

  • AvastMH by AvastMH moderator

    Actually - it's my fault. I wrote the FAQs over a very short period of time! Yshish is checking them for technical content and getting them loaded up. It really does help to have an 'unknowing' pair of eyes use them because that's the real test. When you write them you know exactly what you want to say...but getting it all in the right order with the right implications to someone who needs directions is not always easy. I think there is a common question that I missed altogether...need to check back over the last few days to see what it was (I only went and forgot it - sigh!). SO IF you think there's something missing PLEASE do let me know and I'll sort that out via yshish.

    There are more sections on how to use the site around - but that might well change for the better so those FAQs are on the shelf awaiting further action.

    It's been fun doing the FAQs, and it all came out of taking a 5 minute break from OldWeather on Sunday 5th Jan (I remember it well) that lasted 11 hours. Everyone was getting their New Year's Resolution to save the Planet underway! 😃

    Thanks for letting us know about that FAQ!

    Joan

    Posted

  • gardenmaeve by gardenmaeve moderator

    Fault- there is no fault, only shared credit for making the FAQs appear in their infant form. We'll enjoy seeing them as they mature! Thanks to all who have and continue to contribute.

    Posted