How likely is it that Emily has reviewed or will review my nomination?

General Discussion
1

Since I’m in the US, we can assume that 0.26% of people are Emily and that reviewers are a representative sample of the population matching this statistic. To figure out how likely it is that Emily has reviewed my nomination, I need to know how many people reviewed it, which Niantic does not disclose.

If we make some assumptions about how many reviews are required for resolution, we can calculate the probability that a nomination was reviewed by Emily. The formula is quite simple: 1-0.9974^n

So, if my nomination requires 100 reviews to resolve, then there’s a 22.921% chance that Emily reviewed it. If only 25 reviews are needed, then it drops to a 6.301% chance.

If, instead of being reviewed by the community, my nomination is reviewed by Niantic’s Machine Learning AI model, that number drops precipitously to 0%.

1 Like
2

What about Emile?

1 Like
3

Unfortunately, Emilie is closer to 0.01%, so even at 100 reviews, there’s only a 0.995% chance that Emilie reviewed my nomination.

4

Now, if I could spell, I’d know that Emile is even less common :disappointed:

3 Likes
5

Us bad spellers can untie!

5 Likes
6

Hi @NvlblNm
I understand, that you wanted to keep this as simple as possible, but I think you should take some variables into your calculation🤷🏻‍♀️:

  1. how is an upgrade changing (status in queue) the calculation
  2. same as 1. (status in voting)
  3. If Emily and Emile are siblings (little brother or so) and Emile is learning from his sister, how is the future of humanity?

Sure ther’re more aspects, but that’s all for that time.

1 Like
8

On point 3 - If this goes Game of Thrones then we are all doomed.:scream:

1 Like
9

What if the batch of reviews used for training the ML model included reviews submitted by Emily?

10

Or MichaeL.
Everyone knows they are the worst at reviewing.

1 Like
11

Look, I wasn’t expecting some statistical analysis here on a calculus level here.

If we have to do a chi-squared distribution to figure out anything on here, I’m out

4 Likes
12

Sometimes I think you’re all not real and niantic decided to give me a personal game

1 Like
13

Someone with Alice and Wonder in their name is asking me if this is a simulation? :thinking:

14

Thats a lot of assumptions for a mathematical based post. I feel a moment of deep philosophical thoughts coming on I’ll try to resist it for now.

2 Likes
15

who is emily?

16

Emily is the name given to Niantic’s Machine Learning by a bored 1d1ot.

Some people dislike using the name because it’s confusing, death robots, silver feathers and goats being pushed off buildings.

Some people find it helpful to give it a name simply as it’s easier to say.

There are many studies on the positive and negative effects of personifying AI.
I like to call it Steve.

17

A fancy name for ML - machine learning

18

LOL, Steve, huh… isn’t it a man name? Emily is a woman name

19

Nah, not for a long time.

20

Not Steve, ALAN!

1 Like
21

Time spent in queue doesn’t affect how many reviews the nomination will receive, so the math remains the same.

Upgrades only broaden the reviewer pool to the rest of the country and don’t change the number of reviews required, so the percentage of Emily remains the same as does the calculation for the chance that Emily reviews my nomination.

It doesn’t matter if they’re siblings or not, humanity is still doomed.

Maybe it hastens the timeline by a couple seconds, but a bit more inbreeding isn’t going to significantly impact things.

This is simple single-variable algebra. No calculus classes required.

The only real assumption is around possible values for a variable, everything else is factual.

Emily is many people, each with their own physical corporeal form.

Please stop trying to derail my topic. It’s about Emily, not about Machine Learning.

2 Likes