Universal Origami Grading System ?

General discussion about Origami, Papers, Diagramming, ...
User avatar
Tavin
Junior Member
Posts: 61
Joined: February 1st, 2009, 1:01 pm
Location: Germany
Contact:

Universal Origami Grading System ?

Post by Tavin »

Hello,
as some of you know I'm making Origami-tutorials on Youtube . In my Videos I try to say first what severity the model has.

Now my problem:

I want a grading-model in that does to things:
1: every folder (knowing the system) can grade a model accurate
1: one can tell by the grade if one is skilled enough to do the model.

Is there a universal grading system for Origami?


For example: Let's say the Crane is "easy". But compared with some pure-land origami you also could call it "intermediate".

The often use system of "very easy" => "easy" => "intermediate"=> "high intermediate" => "hard" => "very hard" is vague.

Too vague for my taste.

What I came up for my videos was a scale from 1 to 10 , low numbers being easy and high numbers being hard. I set the traditional Crane as a reference-point and gave it the rating (3/10).

So everything easier than a Crane got a rating below three and everything harder got a higher rating. You get the idea.

But I'm still not satisfied with that system because it is still a matter of my test how much higher or lower I rate different models.


And now to my new attempt of a solution:

My 1-10 System has to major advantages compared to the "adjective-system":
  • it is more precise by having much more grades (you also can rate a model 3.6)

    it has a commonly known reference-point: the crane
The main flaw is the fact that it just has one reference-point.


What I came up with:

I called it universal Origami Grading System (UOGS)

The System also has Numbers.
The Crane is still a reference-point.
The System will have a table of criteria in Order to rate a model.

example:

the model contains a closed-sink (If yes: rating is at least X)

the model is pureland origami (If yes: the rating is lower than X)

the CP consists of less 30 creases/lines (If yes: the rating is lower than x)

tissue-foil is necessary to fold the model. (If yes: rating at least X)

(each X stands for a number)





What are you guys thinking about my Idea?

Is anyone interested in helping me to define the table of criteria? (replacing all the X's with actual numbers)
Moog
Junior Member
Posts: 53
Joined: March 17th, 2010, 9:40 am

Post by Moog »

the model contains a closed-sink (If yes: rating is at least X)

the model is pureland origami (If yes: the rating is lower than X)

the CP consists of less 30 creases/lines (If yes: the rating is lower than x)

tissue-foil is necessary to fold the model. (If yes: rating at least X)
I think that your idea is very interesting, but really difficult to fulfill: I know people that makes a closed-sink in a second and cannot manage a twist, how do grade they which model is difficult ?
and a Byzantine question: how to rate a CP with less than 30 creases/lines, but with a closed sink?
I'd prefer a numeric grade, but I think that "difficult" is too much a subjective "impression", or, maybe, we may arrive a some kind of scale like that one for rock-climbing, that is always follow by a word description of the most important passages for every climbing.
Juanfran
Newbie
Posts: 23
Joined: January 27th, 2010, 11:05 am

Post by Juanfran »

I also thought about this problem, I tried to divide the puntuation, assessing separately several parameters: difficulty of folding, difficulty of final modelling, size of paper (bigger the paper=more difficult models)

Another idea was the same you had, to use models of reference. Crane is a good one, find more. Ryuzin 3.5 or whatever the top, pegasus by Kamiya could be a 6 (for example), and more...

Common known models can be used as a reference. You say: this model is harder than Kamiya's pegasus, and with that information a person could tell aproximately whether he can fold the model or not.

Good luck!
User avatar
Tavin
Junior Member
Posts: 61
Joined: February 1st, 2009, 1:01 pm
Location: Germany
Contact:

Post by Tavin »

If it was simple someone already would have done it! :D

But now seriously:
Yes I also had some thought on the problem that there are different impressions of how difficult one and the same action (e.g. a sink) is.

My Solution to that is: we ask several folders and take the average.

E.g.
A thinks a twist is a 5, B votes for 6 and C thinks its just a 4.
So (5+6+4)/3 =5, and a sink gets the rating 5 . (All numbers are just examples)

And finally a model gets the rating of the highest criteria. Because that rule is necessary for the "I tell by the rating if I can do the model"-part


Could you explain the rock-climbing part more? I'm German I think I'm not getting it right yet.
User avatar
Falcifer
Super Member
Posts: 249
Joined: November 3rd, 2009, 10:43 pm

Post by Falcifer »

I like the idea, but in all honesty, I don't see how feasible it really is.

I'm sure you could generate some sort of criteria-matching scale like you mentioned, but would that really indicate how difficult a model is and would it create a truly comparitive scale?

Number of steps, type of folds used, type of paper needed, size of paper, etc could all be used to create a score. But some models require difficult folds on smaller areas than others, or areas which are obscured by other flaps/parts of the model, which would make them harder, so they would have to be taken into account, too.

On top of everything, there are plenty of models which are created using shaping more than folding, either completely or in the finishing steps, which can be hard enough, even more so when the model is wet-folded. And I don't see how a model like Giang Dinh's cat, which works best with wet-folding and shaping, can be compared to something like Roman Diaz's prehistoric bird, which (quickly glancing at the diagrams) doesn't seem to have (m)any difficult folds. But in "Licence To Fold" they're both marked as Intermediate.
Personally, I would consider the cat to be more difficult, especially because some of the folds are judgement folds. But the bird has more than 3 times the number of steps, which might suggest that it's much harder.
I would have to fold the models, though, with the recommended paper and method, to get a real idea of how difficult each model is, though.

I just think there are too many variables to create a scale in which two different models could be properly compared.

Also, using other models as references would be too hard, since they would have to be models which would be known well enough among beginners and experts alike. I don't think there are many models which are known well enough to beginners - besides the crane - so using them would be meaningless.

As you say, if it was simple then someone would have come up with it already.

Also, I don't really know if such an accurate and detailed system would be necessary. I certainly don't know if it would be worth all the effort.
The current simple/intermediate/complex system works well enough, I think. I have a pretty good idea what to expect from each type of model, and anyone not familiar with origami can still understand what the words "simple", "intermediate" and "complex" mean.
Using numbers would require some knowledge of what the numbers mean, and using the criteria means that people have to know what a "closed sink" is and how difficult it is for them. (Just an example).

It also occurs to me that using numbers is not so different from the vague system you describe;
1 = Very Simple
2 = Simple
3 = Low Intermediate
4 = Intermediate
5 = High Intermediate
6 = Complex
7 = Very Complex

Just that you have a few more grades. Having more grades may not be that much better, and I believe that the current grades all have some requirements, albeit vague and not-set-in-stone requirements.

Finally, I can imagine that using a detailed list of criteria would lead to some misleading grading; one model might score high due to one aspect, but llow due to another, which would give it an average score. Another model might score average on all aspects, giving it a similar grade, suggesting that the two models are equally difficult, but that may not be the case.
It's hard to explain exactly what I mean, so I'm sorry if this doesn't make sense.

Anyway, that's my $0.02
Moog
Junior Member
Posts: 53
Joined: March 17th, 2010, 9:40 am

Post by Moog »

Tavin wrote: Could you explain the rock-climbing part more? I'm German I think I'm not getting it right yet.
I'm Italian , so probably I have been a little confusing don't using my own language...
I would say that when you prepare a rock-climbing you look which grade is that climb, the main indication is how much difficult is the more difficult of the steps (the "grade" of that climb), but every step has his own grade and there are descriptions of the most crucial steps.
Tavin wrote: Yes I also had some thought on the problem that there are different impressions of how difficult one and the same action (e.g. a sink) is.
... not only, for me it's important even how good is the result, I may arrive at the end of a super-complex model having only a crumpled ball of paper, though having done all the steps of the diagram

Tavin wrote:My Solution to that is: we ask several folders and take the average.
In this case we must know how much expert is every folder, to give a weight to his vote
User avatar
Falcifer
Super Member
Posts: 249
Joined: November 3rd, 2009, 10:43 pm

Post by Falcifer »

Moog wrote:I'm Italian , so probably I have been a little confusing don't using my own language...
I would say that when you prepare a rock-climbing you look which grade is that climb, the main indication is how much difficult is the more difficult of the steps (the "grade" of that climb), but every step has his own grade and there are descriptions of the most crucial steps.
If I understand this correctly, an example for origami would be;

Crane - Simple. Contains petal fold (Low Intermediate).

The crane is a simple model, but the petal fold is not a simple step. Either the entire model is elevated to the grade of the hardest step, or the model is judged on the whole sequence, with the hardest step(s) mentioned separately.

Unicorn - Intermediate. Contains spread squash and unwrap (High intermediate).

In this case, I don't know if the unwrap would be considered H.I., but the model is of the Blade Runner unicorn, so it's usually folded from small-ish paper, which makes it slightly harder, even though it's only unwrapping one side of a bird base.

This way would be good, since it gives an indication of the model's overall difficulty, but also points out anything which may cause a problem, or be of greater difficulty than the rest of the model. One hard step in an otherwise easy model shouldn't be enough to mark the model as hard, though. But it's definitely a good idea to mark steps which don't fit in with the model's grade.

But then, you have to decide how many steps are enough to change the grade of the entire model; one step shouldn't, but should two steps? Three, five, ten? Less than 10% of all steps, or what?
I don't think it makes things any simpler in the grand scheme of things, even though, it's not a bad idea...
User avatar
Tavin
Junior Member
Posts: 61
Joined: February 1st, 2009, 1:01 pm
Location: Germany
Contact:

Post by Tavin »

@Juanfran
Juanfran wrote:[...]I tried to divide the puntuation, assessing separately several parameters[...]
Very good Idea, especially to take "shaping" into account too.
I'll totally include that.

You could make one bigger Parameter "Material" to make it more simple
Another idea was the same you had, to use models of reference.
More people having the same Idea is often an indicator for the idea to be a good one.

I think the reference-models should be available for free, so we have an open standard. But more reference-models is definitely a good point. [/quote]
User avatar
orislater
Buddha
Posts: 1211
Joined: November 5th, 2009, 3:57 am
Location: somewhere with a piece of paper in my hand
Contact:

Post by orislater »

another way could be like this

overall folding sequence:* or ** or ***
shaping:* or ** or ***
steps that could be difficult:(insert step number)
and then give a paper recomendation
my flickr tissue foil is for noobs! mc FTW!!!!
User avatar
gachepapier
Forum Sensei
Posts: 533
Joined: June 17th, 2008, 11:22 am
Contact:

Post by gachepapier »

It would make more sense to point out the most difficult steps of a model:
e.g.

crane : basic petal fold
praying mantis : unsink, wrap, closed sinks, multiple in/out sinks.

that way, a beginning folder can assess whether he masters the techniques required by the model or not.

EDIT : much like orislater says :)
mes p'tits plis (now also in English)
User avatar
orislater
Buddha
Posts: 1211
Joined: November 5th, 2009, 3:57 am
Location: somewhere with a piece of paper in my hand
Contact:

Post by orislater »

wouldn't it be so cool if my idea was used in origami books of the future?
i think it would -smug look- haha
my flickr tissue foil is for noobs! mc FTW!!!!
User avatar
Tavin
Junior Member
Posts: 61
Joined: February 1st, 2009, 1:01 pm
Location: Germany
Contact:

Post by Tavin »

@Falcifer
Thank you for your very usefull thoughts on this.
After carefully reading your posts, I have to agree with you.
Maybe you're right and it's actually not worth all the thinking, but I'd like to try solving this interesting puzzle any ways.
Falcifer wrote: [...] I believe that the current grades all have some requirements, albeit vague and not-set-in-stone requirements.
I see what you mean.
Do you think it could be useful to try narrow those vague requirements?
User avatar
Tavin
Junior Member
Posts: 61
Joined: February 1st, 2009, 1:01 pm
Location: Germany
Contact:

Post by Tavin »

I really fancy this "point out the difficult parts"-idea.

I did some more thinking about the function of ratings. The thing most folders want to know by looking at the rating probably is: Can I fold this? Will this model be a piece of cake or a challenge to me?

And that probably can be achieved by some variation of the "mountain-climbing"-rating system.

My criteria could be integrated in some fancy website that has a "show me diagrams that I can probably do"-function. (After you did some sort of short questionnaire)

EDIT: I'd also like to know how you guys grade origami!
User avatar
Falcifer
Super Member
Posts: 249
Joined: November 3rd, 2009, 10:43 pm

Post by Falcifer »

Tavin wrote:Do you think it could be useful to try narrow those vague requirements?
It has its advantages and disadvantages, in my opinion.

One the one hand, it would improve the grading of models and make it easier to identify models which would be foldable by someone given their abilities. At the moment, some "intermediate" models could contain several complex steps which wouldn't necessarily be obvious to everyone.
Also, it's hard to see how difficult certain steps are, even if you look at the diagrams before folding; a sink might not be too challenging in itself, but if the sink in a particular model is performed on a very narrow, sharp point, and the model can't be opened fully at the step, then it can become much harder to perform the sink than usual.
Having these sorts of things mentioned would be very beneficial.

Knowing that models of a certain grade contain certain folds/steps would be a huge advantage, too. Although it does go back to something I was trying to describe earlier; a model may have a grade of 6.5 because it contains a few very difficult steps, but otherwise not be too difficult. Another model of grade 6.5 may not contain many difficult steps, but could require a lot of intricate shaping, so the grades alone would be next to useless in comparing the models.
Using the description of the difficult part (i.e. shaping or difficult folds) does go some way to helping, though.

However, there has to be some limit to the specificity of the criteria. Exactly how narrow should the requirements be?
Two models of similar sequence would obviously be of similar grade. But if one models has 100 steps and the other has 90 steps, how would the grades differ? Is ten steps enough to make them different? What about 11 or 12 steps?

I think that being vague is a necessary part of any grading system, otherwise you could end up making way too many grades; 1-100 for example, instead of 1-10.

I don't think narrowing the requirements is a bad thing, as long as you don't try to narrow them too much, which can be a challenge in itself.
You have to know that there are going to be compromises.

Also, I may be in the minority but I rarely look at the difficulty of models. The number of steps and apparent complexity of the finished model are usually good indications of difficulty.
However, I can imagine that many beginners would benefit from the grading. But then anything too complicated would be overkill for beginners who may not fully understand the degrees of difficulty, especially if they're based on specific folds/techniques.
You would also have to define how difficult sinks, unwraps, unsinks, etc are on the same scale, so that someone who didn't know what they were could judge how difficult they were to perform.

Again, these are just my thoughts on the subject, and everyone else could well have very different ideas.

As for how I grade origami, as I mentioned above, the number of steps and complexity of the finished model are usually good (albeit general) indications.
Also, if I need 50cm x 50cm double tissue to fold it, I know it's not going to be no cakewalk in the tea park.
User avatar
Falcifer
Super Member
Posts: 249
Joined: November 3rd, 2009, 10:43 pm

Post by Falcifer »

By the way, if this is purely for your own videos, I wouldn't worry too much about developing a universal system. ;-)

I know that a lot of books and diagrams use grades, but I wonder how the individual diagrammers/designers judge each grade. For example, do John Montroll and Roman Diaz consider "simple" to be the same? What criteria do they both use for their own models?
For one book, it's easy enough to compare one model to another. But when comparing models from different books and different creators, the lack of consistency could cause a problem with more specific grades.
It would be nice to have one central resource for checking a model's difficulty, but I can't imagine too many people adopting it themselves, especially if it requires checking off a list of requirements and calculating a grade. It's just much easier to use grades which are more vague.
Post Reply