# What percentage of all life do humans make up on earth?

A quick google said 10,000 trillion ants alive at any time.
From a math point of view, if the Thanos snap was designed to remove 50% of life on earth alone then how many humans would be expected to die if every animal from single celled to plankton etc and upwards was considered life?

I’m guessing statistically less than 10 people would be forecast to die

Percent by what? Mass? Number?

About 3.8 billion. Why would it be anything different?

Not mass. Number. If 50% of all life were to die in one moment and all animals from single celled upwards counted as 1 life then how many humans would be expected to die?

Because it would be random. Not 50% of each species but 50% of all living creatures.

So for example if there were 2 living pink elephants then one of them does not ‘have’ to die. It’s just a random half of all and humans must make up a tiny amount of ‘all’ so statistically the risk to humans alone may be deemed negligible

But half is half. Couldn’t you just as well argue that half would be spared, and humans would only make up a minuscule portion of the half that’s spared?

Two pink elephants isn’t the right comparison, because two is a small number. Out of two pink elephants, you’d expect the number who would die would be about 1 ± 1. Out of 7.6 billion, you’d expect the number who would die would be about 3.8 billion ± 60 thousand.

I don’t think I’m wording it correctly.

If on earth there were only 10 ants and 10 humans and half were to randomly die then you might say that the odds were that 5 ants would die and 5 humans would die

But if every living creature were considered a life and say (I am making this number up) each human life was only 0.0001 of the number that included all individual lives on earth at any given time then the bookmakers would say that if the half taken were completely random then there’s a good chance that very few humans would be touched

I meant to say ‘each human life added together only made up 0.0001 of all the lives on earth’

It doesn’t work that way. If 50% of all living creatures are killed, every human being has 50% chance of being among those that are killed. So imagine flipping a coin for every human, and if it comes up heads, he/she is killed. The end result is, the number of humans killed is almost exactly 50% of all humans.

Ok but if there were 8 billion humans and half were to be killed randomly then the answer would be exactly 4 billion humans die but if you added 100 trillion other organisms then of course there is the potential that all humans could die but it also increases the chances that no humans dies at all.

What I am saying is that even though half die the odds are more in favour of a small group.

So on earth if 100 animals lived and 98 were ants and 2 were humans then more ants die and
Humans cannot be any where near a group that amounts to 2% all animal life in earth if each individual life counts

That point is clear, but it really doesn’t matter once the numbers get big enough. Once the population is over ~100 specimens, no matter how many other species there are, things are, with exceedingly high probability, going to behave in more or less a 50/50 manner. And there are way more than 100 humans.

The coin is not tossed for each individual life. It’s a random pick until half the total is reached

All life is in a bucket, there is a total number

The claw stops grabbing at half that number

Yeah, I get that. Doesn’t matter. It still works out to ~50/50 unless you have a really tiny number of specimens. Like less than 30. And even with 30, it’s still a pretty small probability.

Is it possible? Sure. But it’s also possible to get a run of 10000 heads in a row. At that point, you’re dealing with incredibly tiny probabilities.

It’s a weird thing about statistics. Sometimes, it doesn’t take a whole lot to get to limiting behavior. Getting 4, 5, or 6 heads in 10 flips happens with 67.2% probability (pretty good, right?). Getting 12-18 heads out of 30 flips (40-60% of the 30 flips) gets up to ~80%. Getting between 400 and 600 heads in 1000 flips happens with 99.9999% probability (actually a bit more than this).

Upshot is that as long as there are sufficiently many humans, it doesn’t really matter how many other creatures there are. It works out to 50/50 for each human being with exceedingly high probability. This only breaks down when there are very few specimens, which is why a situation where all the original Avengers survived is not too unrealistic. It’s only 50/50 when you get a sufficiently large pool.

Of course, the flip side is also true. With a really small pool, it’s quite possible for everybody to get selected.

I’m still not understanding why the odds for the little guy don’t increase the more specimens are added.

If there’s just 2 of you and 1/2 go then you have a 1 in 2 of dying. If there are 4 of you then you have a 1 in 4 chance in the 1st pick and then a 1 in 3 chance on the last pick. Surely the odds go up if there’s billions upon billions of other things in the pot and humans are only a tiny proportion of the pot

But then yeah I suppose the more of the other life forms in the pot there are then the more the claw needs to keep dipping so overall maybe it should balance out

Because the number of deaths also increases. If there are a hundred living creatures and half will die, there will be 50 deaths. If there are a billion living creatures and half will die, there will be 500,000,000 deaths. Either way, assuming the deaths are randomly distributed, if you are one of the lives your chances of dying are 50%.

Imagine there are a hundred living creatures, and they are all human. Obviously, exactly 50% of humans will die - no more, no less.

Now imagine there are 100 living creatures, 100 humans and 100 ants. It’s now possible that less than 50% of humans will die - in fact none might die. But it’s equally possible that more than 50% of humans might die - in fact they could all die. The likelihood of none dying is exactly the same as the likelihood of all dying (and in both cases the likelihood is small). On the individual level, the “little guy’s” chances are what they always were - 50%. The individual little ant’s chances are also 50%.

The point is that the “pot” also gets bigger; and if 1/2 of the pot is bumped off, then there are more chances of getting picked. In other words, your odds of dodging any individual bullet go up, but the number of bullets goes up proportionally, and it all works out the same in the end.

If there are 4 people, the odds that you die on the first pick at 1/4. The odds that you die on the second pick are 3/4 (the odds you survived the first pick) times 1/3 (the odds that you get chosen on the second pick), which is also 1/4. And 1/4 + 1/4 = 1/2, just as if there were only two people.

It’s a bit of both.

This is a common fallacy you see a lot in statistics.

If there are 4 of you, only 2 are selected.

You have a 1/4 chance of being the first picked.

That means you have a 3/4 chance of NOT being the first picked. Then, you have a 1/3 chance of being picked second. That’s an overall probably of (3/4)*(1/3) = 1/4. Or:

(1/4) + (3/4)*(1/3) = 1/4 + 1/4 = 1/2

You have a 1/4 chance of being picked first and a 1/4 chance of not being picked first only to be picked second. That’s overall 50%.

You can do this with any number of picks.

With 10 creatures, you have a 1/10 chance of being picked first. That means you have a 9/10 chance of NOT being picked first but a 1/9 chance of being picked second, i.e. (9/10)*(1/9) = 1/10.

That means (assuming the selection is uniformly random) you have exactly the same chance of being picked first or second (or third or fourth or fifth).

Basically, if you have a raffle, the order of prizes awarded doesn’t matter. Psychologically, it doesn’t feel this way, but human intuition is notoriously poor when it comes to probability and statistics.

I don’t know how many living creatures there are on Earth. One estimate I’ve read says that there are 2.4 * 10^24 living creatures on Earth. Suppose this is exactly true. Now give every living creature a number between 1 and 2.4 * 10^24. Suppose we have a computer that can pick a number at random between 1 and 2.4 * 10^24. Start it going and let it run for 1.2 * 10^24 times. (We can’t actually do this fast, but let’s pretend that we can.) Kill all of those living creatures. On average, we expect 3.85 * 10^9 humans to be among those living creatures. That’s half of the 7.7 * 10^9 humans on Earth. That’s because their numbers are randomly spread among the numbers from 1 to 2.4 * 10^24.

There are 10^30 bacteria on earth and even more viruses. Do they count?

There are 10^19 insects, give or take.

10^9 humans.

Either way, about 4 billion humans would die.