I’m not good at math, but people keep passing this chart around in reddit-logo

But I keep seeing people in the comments debating whether this is actually “impossible” or not, using stats terms I’m not familiar with. Also idk the source of this graph so maybe it’s bs.

  • rhubarb [he/him]@hexbear.net
    link
    fedilink
    English
    arrow-up
    40
    ·
    4 months ago

    Those numbers of votes are almost definitely wrong and calculated from the percentages instead of the other way around. It doesn’t necessarily mean the percentages are fake, but someone at some point did fuck up.

  • Tabitha ☢️[she/her]@hexbear.net
    link
    fedilink
    English
    arrow-up
    34
    ·
    edit-2
    4 months ago

    alright so I’m bit a of a genius so I’ll try one explanation (BTW I’ve never looked at seen these numbers or this controversy until now).

    Somebody rounded the percentages for normal aesthetic reasons (no idea if done by a person manually or spreadsheet formal in excel, intentionally or not).

    The spreadsheet software is configured to show extra zeros (significant digits, trailing zeros) or somebody typed in those zeros for padding (no idea if done by a person manually or spreadsheet formal in excel, intentionally or not).

    Either way, you have the raw numbers on the left so you can do the math and see 4,445,978 / 10,058,774 = 0.4419999893.

    This is definitely being parroted by people don’t understand math or stats, but they understand that George W Bush has 11 letters (therefor did 911), and this is just another one of those for their collection of parrot songs.

    • Vampire [any]@hexbear.net
      link
      fedilink
      English
      arrow-up
      20
      ·
      edit-2
      4 months ago

      The spreadsheet software is configured to show extra zeros (significant digits, trailing zeros) or somebody typed in those zeros for padding (no idea if done by a person manually or spreadsheet formal in excel, intentionally or not).

      Either way, you have the raw numbers on the left so you can do the math and see 4,445,978 / 10,058,774 = 0.4419999893.

      No. This is wrong. Don’t confuse people. To give five decimal places after the number of percentage points it’s 51.20000% etc.

      In each of the three cases, it is within one vote of the exact 51.2%, 44.2%, 4.6%. It couldn’t be more on-the-button without fractional votes, where you chop Venezuelans up into bits and take the bits to the polling booth.

  • moujikman [none/use name]@hexbear.net
    link
    fedilink
    English
    arrow-up
    26
    ·
    4 months ago

    I worked in government contracting as my first job out of college and the guy who sat I next to in the cubical worked on curating reports for various gov agencies. They would often come back to him and tell him that they number weren’t right and it should be more like ‘X’. He would then just change the numbers and send it back because he didn’t give a fuck.

  • ChaosMaterialist [he/him]@hexbear.net
    link
    fedilink
    English
    arrow-up
    25
    ·
    4 months ago

    It’s not impossible, it’s called padding. You can get your spreadsheet software to display 5 decimal points and it will happily convert ‘51.2’ to ‘51.20000.’ If it were originally ‘51.234’ it would display ‘51.23400’

    Aka this is a nothingburger.

  • Vampire [any]@hexbear.net
    link
    fedilink
    English
    arrow-up
    23
    ·
    edit-2
    4 months ago

    I see it, give me a minute to edit the comment up

    Basically the problem is the divisions are too clean.

    If 10,058,774 votes were divided naturally, you would expect something like candidate A gets 51.19817%, but he got 51.20000%. This would happen if someone said “Make 51.2% of the votes be for Maduro”


    10058774 × 0.512 = 5150092.28, or to the nearest whole number 5,150,092

    10058774 × 0.442 = 4445978.10 or to the nearest whole number 4,445,978

    10058774 × 0.046 = 462703.604, or to the nearest whole number 462,704

    • Vampire [any]@hexbear.net
      link
      fedilink
      English
      arrow-up
      10
      ·
      edit-2
      4 months ago

      No. That wouldn’t explain why exactly* 51.20000% of the votes go to one candidate.

      *The level of exactitude in question = within one whole vote

        • Vampire [any]@hexbear.net
          link
          fedilink
          English
          arrow-up
          6
          ·
          edit-2
          4 months ago

          Five.

          Votes are always natural numbers. There are no fractions of votes.

          For 10,058,774 votes, you need 10-⁷ accuracy. That way you can see the smallest number of votes: i.e. one vote

          Dividing the vote by 7 orders of magnitude is the same as subdividing each percentage point to 5


          As I have already shown in this thread, the match is exact to the nearest vote. It couldn’t be closer without chopping voters into bits.


          Ask yourself this question: if there are 10,058,774 votes, how many votes is 51.2% of that? No fractions allowed.

  • JohnBrownsBussy2 [she/her, they/them]@hexbear.net
    link
    fedilink
    English
    arrow-up
    18
    ·
    edit-2
    4 months ago

    Where does this chart come from? Was it released by the CNE? They haven’t released the tabulated results yet, so it looks like someone took the estimates that the CNE announced in their preliminary bulletin, and then multiplied those by the vote total.

    EDIT: Checking the CNE website, it is currently down, mostly likely a part of the cyberattack campaign against Venezeula’s electoral infrastructure to try and discredit the results.

    EDIT2: Telesur hasn’t posted anything like this either. It’s pointless to debate the plausibility of these totals unless you can verify that these are the totals published by the CNE.

    EDIT3: I haven’t seen any Venezuelan government (or pro-government) source claim a those vote totals either, just a 59% turnout. I wouldn’t be surprised if that estimated was multiplied by the estimated registered voters to get these numbers. It makes sense that the numbers would look jank from reverse construction, but I don’t think that the CNE or government claimed these numbers in the first place.

  • jackmarxist [any]@hexbear.net
    link
    fedilink
    English
    arrow-up
    16
    ·
    4 months ago

    Some Venezuelan channel had a shitty infographic where they mentioned smaller parties multiple times and labelled each of them as 4% instead of their actual vote share so the total added up as above 100%.

  • driving_crooner@lemmy.eco.br
    link
    fedilink
    English
    arrow-up
    9
    ·
    4 months ago

    Because out of ballot surveys gave 70% of the votes to the opposition. Those surveys are nor perfect, but is pretty rare that they have a difference so big, unless the election or the survey was faked.

    • jack [he/him, comrade/them]@hexbear.net
      link
      fedilink
      English
      arrow-up
      44
      ·
      4 months ago

      Exit poll surveys are illegal in Venezuela and have never been practiced before this so I don’t think we should put any stock in them - especially because the company performing them (Edison Research) literally works for the US government. Its main contracts are with Radio Free Europe/Asia and Voice of America, US State Department run propaganda news outlets.

    • Rod_Blagojevic [none/use name]@hexbear.net
      link
      fedilink
      English
      arrow-up
      38
      ·
      4 months ago

      As a general rule of thumb, if someone tells you the ghettos of a nation are rising up and demanding that the oil fields be privatized and given to the richest American, you can be extremely confident that they’re full of shit.

    • Barx [none/use name]@hexbear.net
      link
      fedilink
      English
      arrow-up
      12
      ·
      4 months ago

      It’s very easy to get such discrepancies when your paycheck depends on it. Just poll opposition stronghold sites and start saying things like “70% from our survey”. That pollster’s main clients are US State Department propagandists like Radio Free Europe.

      Imagine saying Trump actually won 70% of the vote in 2020 because you only polled rural Texas, for example.

      The same gang of imperialists and their lapdogs used exactly this logic to support the right wing coup in Bolivua, claiming it was statistically impossible for votes tallied later to swing so heavily for MAS. But any child familiar with what was happening could tell you that the later votes were from rural MAS strongholds and earlier votes included some urban right wing strongholds.

      Boring, Common things spun into " troubling" results as early as possible using hidden and usually incompetent math.

      Re: Venezuela in particular the official results haven’t even been released and these media campaigns were already calling it illegitimate hours after the basic percentages were stated. This was in contrast to what the election observers were saying, which is that the election was conducted fairly and without widespread issues. Venezuela uses a much more auditable and open voting process than the US.

      • SevenSkalls [he/him]@hexbear.net
        link
        fedilink
        English
        arrow-up
        6
        ·
        4 months ago

        So it’s like when Trump wanted to stop the count because he was losing votes from mail-in ballots or whatever it was? That was hilarious.

    • TechnoUnionTypeBeat [he/him, they/them]@hexbear.net
      link
      fedilink
      English
      arrow-up
      8
      ·
      4 months ago

      the survey was faked

      Oh come on, don’t insult your own intelligence

      The survey is probably real, but without having access to key data like sample size and who was surveyed

      If you survey 10 people in an area where the opposition is stronger, and 7 people say they voted for the opposition, that’s 70% - but certainly not a valid number to base a coup off of

  • bubbalu [they/them]@hexbear.net
    link
    fedilink
    English
    arrow-up
    7
    ·
    4 months ago

    Actually, the vote share does round to 51.20000% it looks silly but did we really need to go to six decimals to satisfy these bozos?

  • sexywheat [none/use name]@hexbear.net
    link
    fedilink
    English
    arrow-up
    5
    ·
    4 months ago

    I think a Venezuelan news agency posted the results with the last three candidates all at 4.9% or something, when you tallied everything together it was more than 100%, when in fact the last three candidates together received 4.9%.