Number of solvers for the latest 160 problems

Announcements, comments, ideas, feedback, and "How do I... ?" questions
Post Reply
openBook
Posts: 10
Joined: Tue Dec 08, 2020 4:39 pm

Number of solvers for the latest 160 problems

Post by openBook »

I glanced at the number of solvers for the last 160 problems, it seems like most problems have less than 500 solvers. Why is this the case? Also, would it be possible to get the number of unique solvers for the last 160 problems? Is Project Euler now just catering to a few who can solve the problems in the interval [600, 750]? What about the majority of the solvers who start on PE? Are they all petering out before they reach a stage where they can solve the recent problems?
User avatar
gaufowl
Posts: 5
Joined: Tue Sep 22, 2020 10:36 pm
Location: MD,USA

Re: Number of solvers for the latest 160 problems

Post by gaufowl »

A lot of people do peter out. At this link https://projecteuler.net/problem_analysis you can see the % of people who have solved x problems or more, when considering registered users who have solved at least 1. So the amount of people who have solved at least 1 problem is 100%, whereas the number of people who have solved at least 2 drastically drops to 80%. Meaning 20% of people just register to solve 1 problem and then stop. It seems to me the amount of active users, my metric being solve a problem once a month, is probably on the scale of a few thousand. There does seem to be a significant population of a few hundred people that can pretty consistently solve the newest problems quickly, but there's also a lot of people like me who are still working through a lot of the older stuff.
Image
1691991_rIEOKCNEDBtm7EzRUeWtIZDvhFNxQVp1
openBook
Posts: 10
Joined: Tue Dec 08, 2020 4:39 pm

Re: Number of solvers for the latest 160 problems

Post by openBook »

I am interested in the number of unique solvers for problems in the range [600, 750]. This is something for the PE admins to think about, are the new PE problems serving the entire PE community or only a small fraction of them.
User avatar
hk
Administrator
Posts: 11257
Joined: Sun Mar 26, 2006 10:34 am
Location: Haren, Netherlands

Re: Number of solvers for the latest 160 problems

Post by hk »

The number of unique solvers in the range [600,750] I don't have.
For some other reason I processed the solvers of problems 726 through 747 yesterday. As you know #747 is the most recent problem at the moment.
The number of unique solvers with a public account for these problems is 708. That's more than I thought.
All unique solvers for those problems are 809. That's if I include unlisted accounts and delisted accounts.
New members start with an unlisted account by default these days.
(I looked up all solvers, not only those in the fastest solvers list)

For your leading questions I don't have an answer. What is small in this context?
Image
openBook
Posts: 10
Joined: Tue Dec 08, 2020 4:39 pm

Re: Number of solvers for the latest 160 problems

Post by openBook »

I am assuming the number of unique (X) users solving problems in the [1, 100] interval is much more than the number of unique users (Y) solving problems in the [600, 750] range. So, is (Y / X) < (1 / 10000). Is it a good thing or a bad thing that so few solvers are solving the recent problems. Is PE in the danger of becoming a small community of solvers (i.e. like a clique) who attempt the recent problems?
Last edited by openBook on Thu Feb 18, 2021 5:56 am, edited 1 time in total.
User avatar
euler
Administrator
Posts: 3609
Joined: Sun Mar 05, 2006 4:49 pm
Location: Cheshire, England
Contact:

Re: Number of solvers for the latest 160 problems

Post by euler »

I'm guessing English is not your first language, but the term "in-bred" is deeply offensive in the English language. I would suggest you edit your post and find a more appropriate term.

Comparing the unique user solve count for the most recent problems with the first 100 is not a fair comparison. There is no doubt that the number of distinct solvers will diminish as the problem ID increases.

For example, currently these are the distinct solvers in the following intervals of 100 problems.
101-200: 51752
201-300: 37894
301-400: 18793
401-500: 11550

For 650-747 (most recent problem), the number of unique solvers is 8203.

I am very fortunate to speak with many members of the community about their experiences at Project Euler and the vast majority use it recreationally. They simply enjoy coming back regularly to tick off another problem as they working through the problem set from the ground up. They are not racing to "get to the end". The more recent problems, on one level, cater for those who have already mastered the skills necessary to tackle the bulk of the problem set and/or enjoy adding a competitive element to solving problems. But on another level they simply add more problems to choose from, that will provide almost unlimited fun and challenge for many years to come for the majority of our members.
Image
impudens simia et macrologus profundus fabulae
skoczian
Posts: 6
Joined: Sat May 25, 2013 4:43 pm

Re: Number of solvers for the latest 160 problems

Post by skoczian »

I think it's more interesting to see how many new members start with Project Euler in a given interval. And, of course, if they continue with it (solving one problem and then stopping is a sad thing). I don't find anything about this in the statistics, but I just looked into the forum for problems 1 - 50 and the last entries there are quite recent (2 weeks at the most).

Of course, for the time being Corona might push the number of new solvers.
User avatar
hk
Administrator
Posts: 11257
Joined: Sun Mar 26, 2006 10:34 am
Location: Haren, Netherlands

Re: Number of solvers for the latest 160 problems

Post by hk »

The nearest is:
https://projecteuler.net/problem_analysis
By the way: we think most people start at problem 1.
Image
pjt33
Posts: 69
Joined: Mon Oct 06, 2008 6:14 pm

Re: Number of solvers for the latest 160 problems

Post by pjt33 »

openBook wrote: Wed Feb 17, 2021 9:08 amI am interested in the number of unique solvers for problems in the range [600, 750].
Zero. The problem with fewest solvers is the most recent one, which currently has 61.
openBook
Posts: 10
Joined: Tue Dec 08, 2020 4:39 pm

Re: Number of solvers for the latest 160 problems

Post by openBook »

Thanks everyone for your replies. I think I might have asked for the wrong statistic. I am not sure whether this statistic/information could be obtained from the PE database, what I would ideally like to know is how many solvers started PE in earnest and at what stages they gave up solving PE problems and for what reasons.

What I am getting at is PE does a good job of posting problems, but then does absolutely nothing about helping people solve those problems. Since, especially one of the stated goals of PE is "education and entertainment". This could be one reason for some websites showing up which are deeply annoying to the PE team. Maybe the main page PE website should be edited to state that one of the PE goals is "self-education and entertainment".
User avatar
hk
Administrator
Posts: 11257
Joined: Sun Mar 26, 2006 10:34 am
Location: Haren, Netherlands

Re: Number of solvers for the latest 160 problems

Post by hk »

Vamsi,
I've taken a look at your weblog.

Here is something from the entry for problem 1:
The solution is pretty elementary; it does not use the formula for the sum of first n terms of an arithmetic progression. This is by-design, I do not want to use any extra fancy-math when an elementary solution will solve the problem in less than one minute.
The overview PE is giving presents as second solution using "fancy math" as you call it.
Who do you think is helping starting members more? Your weblog or the overview?

PE has written several overviews to show people that there is more than getting the answer in the simplest way.
We've done that deliberately to help members with later problems.

Are we doing absolutely nothing then to help starting members? I don't think so. You are.

Derogative remarks like "fancy math" are denying everthing PE stands for.
Image
echip
Posts: 4
Joined: Sat Feb 15, 2014 6:51 pm

Re: Number of solvers for the latest 160 problems

Post by echip »

openBook wrote: Thu Feb 18, 2021 6:16 am Thanks everyone for your replies. I think I might have asked for the wrong statistic. I am not sure whether this statistic/information could be obtained from the PE database, what I would ideally like to know is how many solvers started PE in earnest and at what stages they gave up solving PE problems and for what reasons.
People gave up solving PE problems at what stages for what reasons. It is none of your business.

I am seeing a person force everybody to accept his point of view. It's really annoying!!! I recently consider to quit ProjectEuler because there are always people spoiling the problem so that I cannot show my talent. See? It is quite different from your answer: "People quit PE because this site doesn't help people solve those problems."

Euler's statistics give a good indication.
euler wrote: Wed Feb 17, 2021 3:23 pm For example, currently these are the distinct solvers in the following intervals of 100 problems.
101-200: 51752
201-300: 37894
301-400: 18793
401-500: 11550

For 650-747 (most recent problem), the number of unique solvers is 8203.
This statistics negate your assumption: New PE problems are serving a small fraction of PE community. By the way, the number of distinct solvers in the range[1,100] has no any meaning at all because most users just do less than ten problems then go away and never come back again. They can not be considered as a part of PE community.
skoczian
Posts: 6
Joined: Sat May 25, 2013 4:43 pm

Re: Number of solvers for the latest 160 problems

Post by skoczian »

hk wrote: Wed Feb 17, 2021 8:45 pm The nearest is:
https://projecteuler.net/problem_analysis
By the way: we think most people start at problem 1.
Which is a very good idea. Solving the problems more or less in the order they appeared often helps with the later, more difficult ones. That means, of course, that you don't work on the latest 160 problems now if you joined the project perhaps a year or two ago. It doesn't mean that you won't get there.

But the problem_analysis doesn't give any timeframe.
User avatar
gaufowl
Posts: 5
Joined: Tue Sep 22, 2020 10:36 pm
Location: MD,USA

Re: Number of solvers for the latest 160 problems

Post by gaufowl »

hk wrote: Wed Feb 17, 2021 1:07 pm The number of unique solvers with a public account for these problems is 708. That's more than I thought.
All unique solvers for those problems are 809.
That's interesting. Based on the fact that the most recent problems have a range of [63,390] I assumed the elite veterans who were solving these most recent ones had a population of about 400 based on that range, but it's double that. It's encouraging to see that not even the best of the best can regularly solve the new problems. I hope to one day solve a problem in the recent list. D:

Goes to show how much a population can vary with just a few hundred participants, I guess.
Image
1691991_rIEOKCNEDBtm7EzRUeWtIZDvhFNxQVp1
v6ph1
Posts: 131
Joined: Mon Aug 25, 2014 7:14 pm

Re: Number of solvers for the latest 160 problems

Post by v6ph1 »

You should consider time as one reason:
I started around 14 years ago - at this time there were only around 140 problems available.
During this time, I reached at best place 9 within my country and was within the best 200 solvers in the world. My progress was around 70%.
But times change:
I had times with ~ 1 solved problem per day and I had times with 1 problem after 3 months.
As the second 100 Problems are now available for ~15 years and the current 100 are only available for 1 week to 2 years.
And we have only one 6th of the people solved one of the last 100 problems.
This is a great number - if you compare the time these problems are available.

@euler: I assume the total number of solvers of at least one of the problems 1-100 is ~500-800k?

The eulerians and fastest solvers table is only half the truth. There are problems with 50 solvers in the first hour after release.
Incredible fast and nearly impossible to get there within the top 25 if you aren't schedule your weekend after project euler.

The ~400 solvers max vs. ~800 solvers total seems legit:
Not all of us have the same math and coding skills for all subtopics.
So there are geometry problems unsolved by combinatorics persons and vice versa.
And for sure, some of the recent problems are really tough and I'm not sure whether therefore is "fancy math" or which craft needed. :wink:

@gaufowl: You may solve one day one of the recent problems - or be even in the fastest solvers table.
It depends all on your skills, your interest and time and even a little bit of luck to get a recent problem with your favorite topic. But if you solve step by step more and more problems, this will be save bet.
Image
User avatar
euler
Administrator
Posts: 3609
Joined: Sun Mar 05, 2006 4:49 pm
Location: Cheshire, England
Contact:

Re: Number of solvers for the latest 160 problems

Post by euler »

v6ph1 wrote: Thu Feb 18, 2021 9:28 pm @euler: I assume the total number of solvers of at least one of the problems 1-100 is ~500-800k?
Much higher! There are 1079461 unique solvers for problems 1-100. Interestingly, the number who have solved #1 is 1010733, so that means nearly 70000 of those unique solvers decided not to solve the first problem. :o
Image
impudens simia et macrologus profundus fabulae
openBook
Posts: 10
Joined: Tue Dec 08, 2020 4:39 pm

Re: Number of solvers for the latest 160 problems

Post by openBook »

echip wrote: Thu Feb 18, 2021 11:42 am
openBook wrote: Thu Feb 18, 2021 6:16 am Thanks everyone for your replies. I think I might have asked for the wrong statistic. I am not sure whether this statistic/information could be obtained from the PE database, what I would ideally like to know is how many solvers started PE in earnest and at what stages they gave up solving PE problems and for what reasons.
People gave up solving PE problems at what stages for what reasons. It is none of your business.

Euler's statistics give a good indication.
euler wrote: Wed Feb 17, 2021 3:23 pm For example, currently these are the distinct solvers in the following intervals of 100 problems.
101-200: 51752
201-300: 37894
301-400: 18793
401-500: 11550

For 650-747 (most recent problem), the number of unique solvers is 8203.
This statistics negate your assumption: New PE problems are serving a small fraction of PE community. By the way, the number of distinct solvers in the range[1,100] has no any meaning at all because most users just do less than ten problems then go away and never come back again. They can not be considered as a part of PE community.
Agreed, as just another person interested in solving PE problems, why someone else might stop solving PE problems is clearly not any of my business, but to a PE team member interested in the traffic to PE this might be important. There is a whole field called "user success management" which studies how to keep users interested and keep them coming back for more (of course that is for enterprise businesses, while PE is not-for-profit, but the idea is the same. I am assuming the PE team wants a reasonable sized vibrant set of PE users)

Agreed, just the number of unique solvers in the range [1, 100] or [600, 750] may not be the best statistic to support my argument.
echip wrote: I am seeing a person force everybody to accept his point of view. It's really annoying!!! I recently consider to quit ProjectEuler because there are always people spoiling the problem so that I cannot show my talent. See? It is quite different from your answer: "People quit PE because this site doesn't help people solve those problems."
Really?? Are you telling me that you cannot show your talent at solving PE problems because of spoiler websites, could you care to explain how this happens? Please remember that you are just a single data-point. If I may venture a guess, I am guessing you are one of those super-talented PE solvers who think using big integer's is cheating :)
Post Reply