Hacker News Reader

Comment by Macha | Hacker News Reader

The number at the end is unclear - it says I agreed with the majority 11%, but then it shows a bunch of charts. Yet the three I said were vehicles (the car, the police vehicle, and the ambulance) are the only 3 above 50% support, so it seems I agreed with the majority 100%. I even opened a second session in another browser and hit yes to everything to see if the numbers in the chart was the amount that agree with me, and no, the numbers didn't invert so it seems the chart is measuring yes answers.

View on HN

Same, the only three I marked as violating the rule were the car, the police and the ambulance. And it told me I agreed with 11% but showed a bar chart showing that basically everyone agreed with me. It was confusing.

Also breaking a rule is fine for an emergency vehicle.

I also took away the opposite of what the author tried to convey.

But I also got a bad impression of their argumentative integrity because they tried to use a strawman to illustrate their point - only it backfired anyway.

An ill defined rule that lacks examples and definition is not a good way to prove people interpret a good faith attempt at rules differently.

And the longer explanation at the end simply dismisses the notion of giving any examples or even trying to give a clear rule by hand waving and basically saying a motivated person can find ambiguity in anything.

So because a rule or law can’t be defined to perfection without the slightest ambiguity then we should just have anarchy? I’m sorry for the bluntness but that’s asinine.

I also got 11%, but I only selected 3 answers that did _not_ violate the rule: The matchbox toy car, the ISS, (and something else).

This entire thing is showing that people have different interpretations, you can't resort to "that did _not_ violate the rule" to say that it's super obvious and the only answer.

I phrased my answer a bit difficult. What I meant was exactly that, instead of stoplisting a few items, like most people appear to have done, I allow-listed only a few (which is emphasizing the fact that vastly different interpretations exist).

Hah, I got 11% and I selected the car, the police, the ambulance and the tank. Does everyone just get 11% no matter what?

Edit: reading further, yes, everyone gets 11%.

I got 48%, so no not everyone does.

I think it's giving you a percentage of people who answered exactly the same as you on all questions.

Edit: Okay I have no idea what it's meant to be (other than "11")

The problem is that I also got 11% just saying yes to everything. It would seem surprising to me to have the same percentage of maximalists as there is for people who are in the majority on every question

I answered "is a vehicle" to 15 out of 27 questions and also got 11%, so this is either a crazy coincidence or the number's effectively hard coded.

I said yes to about half and also got 11%

I answered yes to one and also got 11%

Suspiciously, I also got 11% (no to everything except police, ambulance, and car)

I got 23% but I included defunct tank memorial.

I also had 11%, but included the tank memorial.

Exact same for me. Well played!

I got 11% but only with the car

11% here too…

I got 11% as well, with no to everything (including the emergency vehicles) and yes to the Honda civic

Same.

I got 11% as well.

I also got 11% with a mix of yeses and nos. I'm guessing something broke. Hopefully the data is being logged correctly at least.

also 11%

I got 11% voting yes on only the Honda Civic and the police. (I accidentally clicked no on the ambulance)

I also got 11% answered no to everything but car and horse...

Not sure given the other responses if this number is acurate?

I answered no to every single scenario as a test, and I also got 11%.

I came to the comments to find out what's going on with the 11%. I was worried that if I agreed with the majority just 11% of the time I must have reversed the buttons or something. But my choices mostly agree with the graph, so I don't think I made an error. I guess everyone gets 11%?

I got 12%, and after also not knowing what it means, and seeing no units whatsoever on the y-axis of the chart, have to leave feeling the author completely and utterly missed their opportunity to convince me of whatever argument they were making. You lose me completely at inability to convey meaningful data from the experiment.

Am I crazy/dumb or is the chart super confusing? I really had to stare at it to figure out how to read it, and I'm still just guessing the vertical scale is percent of the whole that think each item is a car.

But also I got 11% here.

I agree. This has to be the worst, most unintentionally confusing graph I've ever seen. And I don't solely blame this website because many people make this mistake too: For crying out loud, how can you make the simple mistake of not labeling your axis? People online who make graphs—stop not labeling your axes! It's the whole point of the graph!

And 11%, why is everyone getting 11%? Is it a backend error? Who knows! If that 11% had a label, it would be easier for us to know whether it's working as intended or not.

I too got 11%. But I think it's actually emergent, not an error.

I answered that 4 were vehicles (contra OP, I included the tank), and still got 11%. I think there must be some issue with the calculation.

Given that “kite” is the lowest and “Honda civic” is the highest, I’m assuming the bar chart shows % of people who classified it as a violation, regardless of color. For me, the last three are red, and they are also the only ones above 50%. To me that means I agreed with the majority 100% of the time.

But it says I had 11% agreement with the majority.

I got 11% - did anyone not get 11%?

I also couldn't see two different colors on my dimmed phone display until I zoomed way in. I'm not even colorblind.

Maybe it is meta commentary on how broken automatic content filters are when applied to forum posts.

I too got 11% but stuck assiduously to powered vehicles on the ground of or launched from the park.

I did not include the tank.

I also got the exact same percentage saying 'yes' to everything.

I also got the exact same percentage saying 'no ' to everything.

27 * .11 ~= 3, fwiw

What, exactly, is the graph showing? % of people who said "Yes this is a vehicle in the park" or % of people who agree with you ? It's very difficult to tell as I said yes only to the car and the memorial...

Your first interpretation is correct. No one thinks a kite is a vehicle. Everyone thinks a private car is a vehicle. Everything else is in between.

If your answers were consistent with a consistent population, there would be a threshold above which all your bars turn red. Any churn in that consistency (my last six bars are a mixture of red and green) shows a deviation of some kind, but I don’t know how this accounts for when there is no overall consistency and everyone disagrees.

I see. This graph is pretty confusing imo. My first idea for a visualization is to use a stacked bar chart that always adds up to 100% with "No it's not a vehicle" above "Yes it is a vehicle." Use 4 colors, dark red, light red, dark green, light green, and emphasize your choices with the dark ones.

Add a couple checkboxes with interactivity:

- Show only "Is vehicle" bars

- Show only "Is not vehicle" bars

- Show only answers that agreed with me

- Show only answers that disagreed with me

And a "toggle colorblind mode" although I'm not sure what colors would be ideal for that.

Any other ideas for how to make this thing actually readable?

We are all the 11%

Did anyone not get 11?

You can get not-11% by choosing to skip before the end. IF you skip after 7 questions, it's "You agreed with the majority: 29%" — same whether I say yes to all or no to all.

So I don't think it's related to yours answers. It's just the weighted average of what people said for the set of questions you answered, or something like that. Not actually your score at all.

i got 29 percent but i stopped after 7 questions

I got 11 when I did all the questions

18 when I stopped after just the first 7 questions

The charts show your answer, and how popular is your answer, the bigger the more popular.

The "majority 11%" means that 1/9 of your answers were above 50% (as you can see it on the charts).

What it appears to be actually calculating is that 1/9 of items had majority consensus that they were against the rule. It doesn't appear to take your actual selection into consideration, only the amount of questions answered

You are right, the "all violate" and "all do violate" options both give 2/7 = 29% for the first 7, which is the majority opinion (car and ambulance violate, the others do not).

I also got 11% and think that number may be a result of some kind of flawed calculation. The final chart distribution of each "vehicle" paints a much clearer picture.

Yeah, I don't get what the score means. Also saw 11%. Clearly the number means something other than what people think it means, which is maybe the real test.

I am also in the 11%, I don't think this it is accurate or I'm misunderstanding what the number actually represents.

I think that the 11% majority are the people that had the exact same answers as you for all questions.

I also got 11%. I suspect foul play..

I also got 11%

I think that number is just bugged

It seems unlikely that we both agreed with 11%, yet that's what it says.

11% club checking in

We're all 11% on this blessed day.

I only selected one "yes" answer (the car), but still got 11%, so not sure what this may mean.

11% as well (all questions answered).

I answered yes to all of them and it said I was in the 11% majority as well.

Also 11%

The only one red in my chart is the car and I too got 11%.

Hey me too. Another 11 here.

Reply on news.ycombinator.com