Differences

This shows you the differences between two versions of the page.

--- statistics [2026/05/11 12:43] – created ultracomfy
+++ statistics [2026/06/09 08:54] (current) – ultracomfy
@@ Line 1: / Line 1: @@
-{{page>Templates:Page Incomplete}}
 <WRAP column 18% right>
 {{page>Templates:Secularization}}
@@ Line 9: / Line 7: @@
 <fs xx-large>Statistics</fs></WRAP>
-====== Lying with Statistics ======
+Before I go talk about statistics and what to do with them, I'd like to talk a bit about how statistics are used incorrectly. I want to do it that way because the way statistics are abused gives us major hints for what to do to use statistics responsibly.
+====== Statistics: How to Not ======
 At least superficially, most humans claim that they can be persuaded by facts and numbers. Therefore, statistics are your weapon of choice when trying to convince other humans that you are right about something. Statistics compile the complexity of reality into easily digestible numbers that, ultimately, help assign a numerical value to things. For example, statistics can show us the change of GDP of a country over time, and we can try to interpret that based on the things we think influenced that change. This interpretation can be used to say things about the influences. For example, when rich countries with low birth rates allow immigrants in, this generally positively affects that country's GDP.
@@ Line 22: / Line 22: @@
 An easy way to lie about this is to use the raw number for shock factor. If a town of 10,000 gets 10,000 immigrants, it's now a town of 20,000. Do you see how this gives opportunity to lie about the numbers? It's so simple you'll even remember the necessary maths from school. Twice as much crime! The number of break-ins are going through the roof! 50% of our crime is committed by migrants! And that, even though migrants are actually committing the appropriate amount of crime and nothing about our statistics has actually changed. Using absolute numbers when relative numbers are more appropriate is a classic way to lie with statistics.
-===== Crime: Relative =====
+===== Crime: Relative Numbers =====
+<WRAP column right 25%>
+{{ ::crime_change_infographic.png?nolink |}}\\
+Look at this, left extremism is much worse than right extremism.
+</WRAP>
 But, fear not, it works the other way around as well! Today is May 11, 2026 and I just stumbled over a fresh-off-the-press example of relative numbers being used to distort reality when using the absolute numbers was more appropriate. One of Germany's conservative parties is currently trying to make cooperation with the far right Nazi party palatable, and they do that with the standard method of pointing fingers at "the left". So, the infographic goes as follows:
 "Change of extremist crime in Bavaria:", followed by a bar graph: "Islamism +6.5%", "Right wing extremism +24.8%" and "Left wing extremism +135.7%"; "Source: Federal Office of the Protection of the Constitution 2025". Hmmm, I bet the constitution protection office wouldn't sign off on this static as such, but putting their name there anyway lends these numbers crecedence.
-If you're not prepared, what follows from this statistic would be that //the real problem// is left wing extremism. They're going completely out of control! Well, no, they're not going out of control. I find this chart to be weak in general because even with at least a little bit of life experience you'd recognize that the normal growth numbers for public policy things like crime generally moves in comparatively small numbers. If your numbers explode like this, this would typically raise questions about your methodology or your interpretation. For endemic types of crime, +6.5% is a natural amount of growth. Even when prompted due to certain events (like right now the Israel - Palestine situation) you can have spikes, but 135.7% is basically a number you just do not see. The reason for this is that left wing crime is not (or at least used to be) not endemic. Even the +24.8% for right wing extremism could be within the margin for endemic right wing extremism, the reason we have such an explosive development in left wing extremism is because it is/was NOT endemic. It basically wasn't a thing before and it's now because of Israel - Palestine that we see movement on that front. And of course, if there's only been very little of a thing before, spikes like these can easily happen when a conflict like that reaches a melting point.
+If you're not prepared, what follows from this statistic would be that left wing extremism is much worse than right wing extremism. They're completely out of control!\\
+Well, no, they're not going out of control. I find this chart to be weak in general because even with at least a little bit of life experience you'd recognize that the normal growth numbers for public policy things like crime generally moves in comparatively small numbers. If your numbers explode like this, this would typically raise questions about your methodology or your interpretation. For endemic types of crime, +6.5% is a natural amount of growth. Even when prompted due to certain events (like right now the Israel - Palestine situation) you can have spikes, but 135.7% is basically a number you just do not see. The reason for this is that left wing crime is not (or at least used to be) not endemic. Even the +24.8% for right wing extremism could be within the margin for endemic right wing extremism, the reason we have such an explosive development in left wing extremism is because it is/was NOT endemic. It basically wasn't a thing before and it's now because of Israel - Palestine that we see movement on that front. And of course, if there's only been very little of a thing before, spikes like these can easily happen when a conflict like that reaches a melting point. That's why spikes like 135.7% are //possible//, but then obviously don't say very much.
+This graphic is a very smart use of relative numbers to distort the perception of the actual reality that was measured. The actual rates are still lying underneath. Left wing extremism used to be not represented, //now// there are roughly as many left wing extremist cases as there are right wing extremist cases (per capita). One other things this graph fails to mention is the type of crime typically committed. Right wing extremism is often targeted at individuals and frequently sees bodily injury. Left wing extremism expresses itself in vandalism and the destruction of property, //not// people (mostly((Yeah I know, that poor health insurance fucker from the United States. This graph talks about Germany though. And also the guy killed people himself, whether gun or flick of the pen doesn't matter. Doesn't justify it, but it's more compliated than just "libevil". If you think crime against migrants is at least understandable because they make the country worse, then you should understand crime against pen killers as well, whether you agree with it or not.))).
+----
+===== Climate Change: Absolute Numbers =====
+What's the point in reducing our carbon footprint if China is so selfish? There is a very common graph (which I will show further below) that is used to give the impression that China creates so much emissions that other countries can't do anything about climate change, even if they wanted to. Part of the argument here is that China creates more emissions than other nations. And while it is true that China has the largest share of global emissions, there is often the implication that China is a particularly dirty nation. However, that isn't true. As usual, the graph being used to this end shows absolute numbers, which is misleading. There are reasons for why the absolute numbers are the way they are, and it's not that China is particularly selfish. The reason for these absolute numbers is that China hosts more than a third of the world's population. More civies means more consumption means more emissions. Additionally, China also hosts the world's largest industry. Most of our stuff comes from there because it is cheaper to import from China than to manufacture locally. China doesn't just produce for itself, it produces most things for most of the world.
+Contrast and compare the graph being used maliciously on the left with the graph that accounts for per capita and trade on the right.
+<WRAP group>
+<WRAP column half centeralign>
+{{ ::total_co2_emissions_per_country.png?nolink&300 |}}\\
+Yes, China has high total emissions...
+</WRAP>
+<WRAP column half centeralign>
+{{ ::per_capita_consumption_based_co2_emissions.png?nolink&500 |}}\\
+...but the reality is that China is no worse than most other countries.
+</WRAP>
+</WRAP>
+Climate change denialism is one of the easiest offenders to spot. Just ask yourself, "what does this argument imply?". If "[...], [[https://rationalwiki.org/wiki/Climate_change#The_denialist_staircase|therefore we don't need to change anything]]" sounds like a reasonable deduction from the argument given, you have found climate change denialism. Climate change denialism doesn't necessarily deny the existence of man-made climate change, but all its arguments are always, **always** aimed at trying to argue that, for one reason or another, nothing or nothing substantial needs to be done. The goal in every instance is to maintain the status quo with minimal concessions. Everything a climate denier can say will always amount to " - and therefore we don't need to change anything", even though the scientific consensus has been for over 200 years now that, no, we do actually need to change something, and rather rapidly. I really do encourage you to give [[https://rationalwiki.org/wiki/Climate_change#The_denialist_staircase|this]] a read.
+One of these steps, by the way, does include something like "We can't do anything about this, therefore we don't need to change anything". Now, I don't want to exculpate China from its increased responsibility in getting to net zero. If all our emissions are imported, it is still China from which these emissions are imported. However, as with most things, things previously said to be impossible somehow always end up seeming quite possible indeed when you convince the right people that something does need to be done.
+In our case, the simplest solution would be to stop importing. Don't like it? Well, a more involved... gracious... solution could be some kind of treaty with China wherein we pay more to China for our imports but China uses the extra money to go to net zero((I know you may not like these ideas and I concur. I can only point you towards the [[https://www.ipcc.ch/|IPCC]], who are actually experts in the field and know what they're talking about. But unfortunately I have to agree, there is no way to get to Net 0 today without pain. That's the result of decades, nay, centuries of inaction.)). There are all kinds of possible ideas here. But as far as these statistics are concerned and the way they are used to lie about our reality, it is wrong that we can't do anything about China's emissions, or that China isn't doing anything about its emissions. They're not just releasing uncontrolled amounts of CO2. The reality of China's emissions is that they are roughly on par with the rest of the developed world (except for Australia).
+====== Statistics: How to Do It Right ======
+So what do we make of this? Well, there are primarily two things we will have to learn from this:
-This graphic is a very smart use of relative numbers to create a perception of "rabid" left wing extremism that hides the actual rates lying underneath. Left wing extremism used to be not represented, //now// there are roughly as many left wing extremist cases as there are right wing extremist cases (per capita). One other things this graph fails to mention is the type of crime typically committed. Right wing extremism is often targeted at individuals and frequently sees bodily injury. Left wing extremism expresses itself in vandalism and the destruction of property, //not// people (mostly((Yeah I know, that poor health insurance fucker from the United States. This graph talks about Germany though. And also the guy killed people himself, whether gun or flick of the pen doesn't matter. Doesn't justify it, but it's more compliated than just "libevil". If you think crime against migrants is at least understandable because they make the country worse, then you should understand crime against pen killers as well, whether you agree with it or not.))).
+  - Data is **weak**. As they captured so well in the 2026 video game "007 First Light", any data can be tortured so long and so easily that you can get any confession out of it. Visualized, cherrypicked or scaled correctly, any piece of data can be used to say anything about anything.
+  - It takes real understanding of a subject to understand data correctly. Only if you have the full picture of a situation are you able to fit in any kind of data.
-===== Climate Change: Absolute =====
+In other terms, if someone, anyone posts statistics online or in print or wherever, do not assume that it proves a point or supports anything. Even if the data is used to support causes you are allied with/think are correct, neither you nor I have the necessary expertise to actually meaningfully understand the data/statistic. In general with things like these, it is best to shut the fuck up.
-Another graphic I stumbled over is the amount of CO2 emissions per country. If you understood the situation with crime, this one should be pretty easy to read as well. The argument here is that there is no point in western countries to go carbon neutral if China won't. Well, no, these are absolute numbers. China hosts more than a third of the world's population. More civies means more consumption means more emissions. Additionally, China also hosts the world's largest industry. Most of our stuff comes from there because it is cheaper to import from China than to manufacture locally. China doesn't just produce for itself, it produces most things for most of the world. So, more industry means more emissions. If you adjust this graph for per capita and trade, this is the result.
-This is not to exculpate China from its increased responsibility in getting to net zero. If all our emissions are imported, it is still China from which these emissions are imported. However, it means there are things we can do about this. The simplest solution would be to stop importing. A more involved solution could be some kind of treaty with China wherein we pay more to China for our imports but China uses the extra money to go to net zero. There are all kinds of possible ideas here. But as far as these statistics are concerned and the way they are used to lie about our reality, it is wrong that we can't do anything about China's emissions, or that China isn't doing anything about its emissions. They're not just releasing uncontrolled amounts of CO2. The reality of China's emissions is that they are roughly on par with the rest of the developed world (except for Australia).
+It is one of the many reasons for why I don't like the [[Kurzgesagt]] [[YouTube]] channel. It makes claims about things like climate change and the such and then backs them up with data, but data alone is useless. This inspires a false confidence in viewers who then go touting around their newfound statistics without actually having the knowledge to interpret those statistics correctly. This is made worse by the fact that Kurzgesagt is largely in line with scientific consensus on most things, so it is rather difficult to have a merit-based argument with their viewers. They are polluting the discourse around complex topics with under-complex interpretations of scientific concepts and no understanding of how to interpret the data... all while not even being totally wrong. As I said, the best thing to do is to shut the fuck up. And [[Probably Wrong|not just because of statistics.....]]