Project Euler Forum

I think that maybe some folks need to stop and take a breath. Euler made a good-faith effort in good faith to add a little whimsical fun to the ratings display, not conscripting us all to join a death squad. While I respect that some people have reservations about real-world military service, such absolute animosity to merely bearing a title in an undefined (and certainly non-martial) fictional organization of mathematicians and computer scientists seems insensitive to those members of the community who did serve in their nation's armed forces.

For a substitute that is equally open-ended, more mathematical, and less emotionally laden, perhaps we could go with animated gifs of the Platonic, Archimedan, and Catalan solids? There are 31 of them, so it would probably last nearly as long as the last system, and arguably stellations could be added after that to keep it going if need be.

Tirian wrote:For a substitute that is equally open-ended, more mathematical, and less emotionally laden, perhaps we could go with animated gifs of the Platonic, Archimedan, and Catalan solids? There are 31 of them, so it would probably last nearly as long as the last system, and arguably stellations could be added after that to keep it going if need be.

I like this idea. You get more faces as you do more problem. How about a sphere for 100%?

Edit: Actually, I think non-animated versions of the solids would be less cluttered.

Also, maybe the scores page should show people within 500-1000 rank (the closest 1000 ranks) of the viewer? This way, newcomers will not feel put off by being "off the radar."

Hmmm. I figured that animation would give a better sense of the symmetry of the figures, but I guess the clear ones are absurdly large. Pity.

Thumbo wrote: Academic ranks?

Freshman, Sophomore, Junior, Senior, Alumni, Lecturer, Associate professor, Professor, Dean, Bursar...

The problem with this is that it isn't universal. Those designations are not used in the UK and while I can guess what they mean, if they were jumbled I'd struggle to put them back in the correct order. Better to use system that's common to everyone around the world (if that's possible ).

The military ranks were clever and well-implemented. I appreciate the thought and effort that went into the implementation.

I mean no offense to any particular individual, but I don't see how any reasonable person could find offense in something so clearly meant to merely make the site more interesting and engaging.

As Tirian suggested, i think we should have "campaigns".. or "challenges" as you could call them.
Like as said before, you could group all the 'gaming' problems together.. the 'prime' problems together.. etc.
You could call them.. well, I'll let you come up with a name!

As for rankings, since we are all math-oriented more or less, how about we going by mathematicians names? It sounds a little lame, but who knows?

The academic rankings as suggested by Thumbo appeal to me also.

Another suggestion is using the names of precious minerals (opal, diamond, gold, etc.) as rankings...

I didn't find offense at the military rankings; I thought they were pretty cool.

euler,
First off, I love project euler and I have been trying to complete all the problems for a month or two now. Thanks for the great content.
Don't get too discouraged by the comments on the ranking system. I think the reason people are so vocal about their dislike of the new system is that so many people really enjoyed the site the way it was. I personally think its great, and will continue solving problems regardless of the ranking system.
That being said, I love the Catalan/Archimedean solids idea. While I am not averse to the military ranks, I think the solids fit the bill quite a bit better. You could even get the users to submit their own renditions.

However the ranking system turns out it won't detract from the site enough to make me stop coming back, but for what its worth my vote is for the Catalan/Archimedean solids.

-brad dunbar

euler wrote:I must admit that I couldn't think how that could be done. Unfortunately we don't keep track of when users register, only when they solved the last problem.

Maybe I wasn't clear. Let S_t(u) be the total number of problems that user u has solved by the start of week t. The weekly improvement is Δ_t+1(u) = S_t+1(u) - S_t(u). So once a week (at the time the new problem is published), you retrieve last week's scores, calculate the improvements, compile web pages for the top 100-or-so, then save off this week's scores ready for next week's update. Easy.

greekcandyman wrote:As Tirian suggested, i think we should have "campaigns".. or "challenges" as you could call them.
Like as said before, you could group all the 'gaming' problems together.. the 'prime' problems together.. etc.
You could call them.. well, I'll let you come up with a name!

Like mentioned before, this might divulge some crucial information on some problems. If we group together all multiplicative functions problem or all cf, dp, or Diophantine equation problems, that would make the site lose a bit of the problem-solving aspect.

KenMiller wrote:I mean no offense to any particular individual, but I don't see how any reasonable person could find offense in something so clearly meant to merely make the site more interesting and engaging.

I think most of us are only concerned, not offended. Certainly that is reasonable, don't you agree?

braddunbar wrote:That being said, I love the Catalan/Archimedean solids idea. While I am not averse to the military ranks, I think the solids fit the bill quite a bit better. You could even get the users to submit their own renditions.

I think so too. It's only natural that a math/computing website get a more "mathy" theme.

neverforget wrote:

greekcandyman wrote:As Tirian suggested, i think we should have "campaigns".. or "challenges" as you could call them.
Like as said before, you could group all the 'gaming' problems together.. the 'prime' problems together.. etc.
You could call them.. well, I'll let you come up with a name!

Like mentioned before, this might divulge some crucial information on some problems. If we group together all multiplicative functions problem or all cf, dp, or Diophantine equation problems, that would make the site lose a bit of the problem-solving aspect.

What about grouping some problems together that were proposed by a certain member? I don't know if anyone has proposed 10 to 15 problems as the "source" is never mentioned, but there could be a <insert username>-challenge or so...

Here is a comprehensive list of uniform polyhedra:
http://en.wikipedia.org/wiki/List_of_uniform_polyhedra

One option would be to store the number of faces, edges, vertices, and T=F+E+V in a database. Then depending on how many problems a member has solved, S, they can select any of these polyhedra for which S[ge]T. So a member could not select any until S=14 (tetrahedron) and you couldn't pick, say, the great dirhombicosidodecahedron until S=424. However, apart from the "privilege" of picking more elaborate polyhedra as you solve more problems I wonder if there would be any significant reward in this?

Alternatively, we use T to order them for the ranks?

One of the reasons for having a ranking system is that it presents short term targets for members, plus it allows members to find recognition in one of the score tables.

So perhaps we forget about clever rank titles and for, say, every fifty problems solved you advance a level. This way a different score table could be created for each level: Novice (0-49), Level 1 (50-99), Level 2 (100-149), and so on. This would reflect the "academic" class proposal made earlier, where members could continue advancing from one level of experience to the next as they solve more problems.

The other option I presented earlier in this thread is allowing members to pick a famous mathematician/scientist when they've solved, say, fifty problems. Then every, say, fifty more problems they solve they can add another to their profile. It would allow for some interesting statistics relating to the most admired/popular mathematician/scientist among the Project Euler members. But like allowing members to pick polyhedra, I don't think there is any intrinsic reward in this. Maybe this could work alongside the "Level ranking" idea?

I think that the solids idea is on the right track, but there are a few small problems:
1) The order of the higher ranked solids are less apparent. For example, it's not so easy to remember "I'm currently a great inverted snub icosidodecahedron! Only 7 more solutions until I'm a great retrosnub icosidodecahedron!" I don't know how this would compare to the ease of using other systems, though. Some people don't know if a Private or a Major comes first (a bit drastic, but only because I don't know enough about the military to use closer ranks!). Others don't know if it's cooler to catch a Carp or a Bass. Some don't know the difference between their pre-algebra teacher and Carl Gauss! So, that's a problem that must be considered for ANY system. The seemingly simple solution is to say: "You're a Great retrosnub icosidodecahedron, having solved 164 problems." However, then couldn't you just say "You've solved 160 problems?" This is where it seems plausible simply to use levels: "You're at level 16, having solved 164 problems."

2) The solids, as they're drawn on wikipedia, are mindbogglingly ugly. They would need to be rendered with much less Windows 3.1-esque colors. That, however, would probably require the use of much more similar colors (light silver, dark silver) instead of red, green, yellow, blue, etc. That, in turn, would make them harder to see, harder to distinguish, etc.

All in favor of the immediate scrapping of any scoring system overhauls, say "I..."

I think it would be much more worthwhile to stick with the current system, but work on little awards/badges to recognize people beyond the normal ranks/scores (solving many problems, quickly; contributing a problem; brilliant idea on a problem forum; donating to the site...). And the Mathematician vs Computer Scientist feature. I haven't forgotten that, yet!

All just ideas,
StalePretzel

At higher levels, being able to choose among the vast number of polyhedra would seem silly (and also graphically clumsy). I suggest to select a few simple and beautiful polyhedra for the rank icons for completing certain number of problems (maybe after every 20-25) as a trophy or milestone.

Is anyone contributing to this discussion a "low-ranking" player? Or are the 90%+ crowd just enjoying a thought experiment, I wonder. None of the latter need any motivational boost, surely.

I've not solved any.
http://projecteuler.net/index.php?secti ... file=euler

euler wrote:One of the reasons for having a ranking system is that it presents short term targets for members, plus it allows members to find recognition in one of the score tables.

I think that getting to the root of what would make the members feel "accomplished" and what "rewards" they would appreciate for those efforts are important; otherwise these are just solutions in search of a problem. I will echo the comments of others that the "original" system gave me all the feedback I required. I worked my way onto the top 1000 list and then I climbed it. At one point, I knew that I was on page 5 and a week ago I had been on page 7 and I didn't need a title or a graphic to tell me that was good. At the same time, I don't demand or expect that anyone else cares about my progress, because I barely care about the progress of anyone else. (The one exception is that I like looking at the problems solved by people near my "rung" of the ladder to get a sense for the problems that might be easiest.)

So I don't think that there is much sense in the "reward" side of the equation, at least for the standard goal of solving all of the problems. I've solved 165 problems, and that stands on its own in my mind without being called a quartermaster sergeant or a junior professor or a triakis octahedron. The only thing I really want is something that you're not prepared to give: a hint for one of the remaining 37 problems.

What I suspect would be fruitful is creating more goals. One of your ideas is solving the most recent 25 problems, and an idea from my earlier suggestion on "campaigns" might be solving all of the problems that are obviously graph-theoretical in nature. Something that might welcome someone who might currently feel shut out would be solving all of the problems that can be solved without college-level math (or Google). You could implement all these and more by adding tags to problems and then allowing users to search for the percentage of problems that they have solved that have a specific tag. That wouldn't generally require a formal high score table (see above about not caring about other people's progress) so much as just tracking how many people are at my level and (if not completed) how many people are above it. And some sort of glyph for the people who finished -- either a rosette or a picture of a mathematician or a uniform polyhedron -- take your pick. But it would give the newcomers something to aim for while they accumulate their first 100 problems and get on the "main list", and might give all users an easier time in searching through the entire catalog of problems for things that they might specifically enjoy.

euler wrote:I've not solved any.
http://projecteuler.net/index.php?secti ... file=euler

I actually wondered about it... there were times you were always one step ahead

Keeping track of rank deltas could provide a nice incentive, especially for lower ranking players. With many inactive players, it should be possible to move up the ranks by quite a few positions each day. The higher you get, the harder it will be to get a high rank delta. It's also a nice way to compare against other active players.

As for rewards, for me it's the number of solved problems that counts. A reward now and then could be nice, but it's not really what I'm here for. Except perhaps the ability to store notes in my profile. I'm working on these problems during breaks and at other places sometimes, so that would be a convenient reward for solving, say, 50 problems (I'm at 76, heh). That also means only the more dedicated players will benefit from it.

The programmer versus mathematician idea sounds interesting. I'd like to see some statistics on that.

Why not do the following:
Instead of Level 1, Level 2, etc we choose names of mathematicians in alphabetic order.
So instead of Level 1, we could use Abel.
Instead of Level 2, we could use Babbage.
Instead of Level 5, we could use Euler.
And so on.
If we fear that we would not have enough levels we could use several mathematicians with the same first letter.
The alphabetical order still decides.
This way we are saving us a few troubles:
-it is immediately clear which level is the highest
-it saves us the ordeal deciding which mathematician is "smarter"
Here is a list of a lot of mathematicians: http://www-groups.dcs.st-and.ac.uk/~his ... Index.html

euler wrote:
One of the reasons for having a ranking system is that it presents short term targets for members, plus it allows members to find recognition in one of the score tables.

So perhaps we forget about clever rank titles and for, say, every fifty problems solved you advance a level. This way a different score table could be created for each level: Novice (0-49), Level 1 (50-99), Level 2 (100-149), and so on. This would reflect the "academic" class proposal made earlier, where members could continue advancing from one level of experience to the next as they solve more problems.

I don't like the thought of adding shapes or mathematicians to my profile. It doesn't seem like a really good reward..

But I like this level idea. However, I don't think it should be every fifty levels, because wouldn't that mean you could only get to Level 5 or so?

What if you made certain problems worth more "points" (just going towards your level) than others, based on difficulty or the number of people who solved them. That way if you solve lots of "hard" problems you could jump up in your level more... or if for some reason thousands of people solved what was thought to be the "hardest" problem, your ranking would change accordingly (aka drop!).

Oh, and I'm not in the 90% boat either.. I only solved 37! But then again, I havent graduated from high school yet!
(I think you should add an "age" thing to profile while you're at it! )