Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make square-ness or non-square-ness of the number easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Is the page BP384 really about "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots likewise arranged helpfully"?

It is pedantry, but the difference might affect how people sort the BP page within meta-BPs.

The OEBP shoos aside this philosophical issue by titling BP384 "square number of dots vs. non-square number of dots". The "help" keyword is just a fluffy administrative tag on top of that, which says we don't want users to submit any examples that are hard for humans to parse, because that would clutter up the page with irrelevant noise.

If a user insisted, no one would stop them from creating another BP page for that slightly different solution: "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots likewise arranged helpfully." This would be an example of an "overriddensolution" (left-BP568); the original solution is simpler. If a person were given that BP and were told to sort a new, non-helpfully arranged picture of a square number of dots, it would still seem correct to sort it left.

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make square-ness or non-square-ness of the number easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Is the page BP384 really about "square number of dots vs. non-square number of dots," or is it "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots also arranged helpfully"?

It is pedantry, but the difference might affect how people sort the BP page within meta-BPs.

The OEBP shoos aside this philosophical issue by titling BP384 "square number of dots vs. non-square number of dots". The "help" keyword is just a fluffy administrative tag on top of that, which says we don't want users to submit any examples that are hard for humans to parse, because that would clutter up the page with irrelevant noise.

If a user insisted, no one would stop them from creating another BP page for that slightly different solution: "square number of dots that are arranged in a helpful way so as to communicate the square-ness vs. non-square number of dots also arranged helpfully." This would be an example of an "overriddensolution" (left-BP568); the original solution is simpler. If a person were given that BP and were told to sort a new, non-helpfully arranged picture of a square number of dots, it would still seem correct to sort it left.

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make square-ness or non-square-ness of the number easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution.

Is the page BP384 really about "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots (ditto)"?

It is pedantry, but the difference might affect how people sort the BP page within meta-BPs.

The OEBP shoos aside this philosophical issue by titling BP384 "square number of dots vs. non-square number of dots". The "help" keyword is just a fluffy administrative tag on top of that, which says we don't want users to submit any examples that are hard for humans to parse, because that would clutter up the page with irrelevant noise.

If a user insisted, no one would stop them from creating another BP page for that slightly different solution: "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots (ditto)." This would be an example of an "overriddensolution" (left-BP568); the original solution is simpler. If a person were given that BP and were told to sort a new, non-helpfully arranged picture of a square number of dots, it would still seem correct to sort it left.

BPs should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make square-ness or non-square-ness of the number easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it may become unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots (ditto)"? The OEBP kicks away this philosophical issue by outright specifying the page BP384 is for "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse, because that would clutter up the page with irrelevant noise.

If you insisted, no one would stop you from creating another different BP page for that slightly different solution: "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots (ditto)." This would be an example of a Bongard Problem with an "overriddensolution" (left-BP568). If a person were given a new, non-helpfully arranged example having a square number of dots to sort, it would still seem intuitive to sort it left.

Problems should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make square-ness or non-square-ness of the number easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it becomes unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots"? The OEBP kicks away this philosophical issue by outright specifying the page BP384 is for "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse, because that would clutter up the page with irrelevant noise.

If we really wanted to, we could create another different BP page for that slightly different solution: "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots." This would be an example of an "overriddensolution" (left-BP568). If one were given a new, non-helpfully arranged example having a square number of dots to sort, it would still seem intuitive to sort it left.

Problems should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make square-ness or non-square-ness of the number easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it becomes unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots"? The OEBP kicks away this philosophical issue by outright specifying the page BP384 is for "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse, because that would clutter up the page with irrelevant noise.

If we really wanted to, we could create another different BP page for that slightly different solution: "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots." This would be an example of an "overriddensolution" (left-BP568). If one were given a new, non-helpfully arranged example having a square number of dots to sort, it would still seem intuitive to sort it left, alongside the more helpful examples of squares.

Problems should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

Helpfulness can be a spectrum; most Bongard Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged recognizably. Instead, the dots in the examples are always arranged in shapes that make square-ness or non-square-ness of the number easy to check without brute counting.

When all examples in a Bongard Problem are helpful, it becomes unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots"? The OEBP kicks away this philosophical issue by outright specifying the page BP384 is for "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse, because that would clutter up the page with irrelevant noise.

If we really wanted to, we could create another different BP page for that slightly different solution: "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots." However, this would not be a very interesting addition to the OEBP--we don't want loads of slightly different copies of Problems. Furthermore this would be an "overriddensolution" (left-BP568). Even when all examples in BP384 are helpful, with dots arranged in ways that make it easy to see whether they are or are not a square number, it still seems intuitive to sort left, along with those helpful examples, any new, non-helpfully arranged example that has a square number of dots.

Problems should be marked as helpful when the OEBP wants most examples (on at least one side) to be helpful, rather than when just one or two uploaded examples are helpful.

Left examples have the keyword "help" on the OEBP.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged in a helpful way.

Helpfulness can be a spectrum; most Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

When all examples in a Bongard Problem are helpful, it becomes unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots"? The OEBP avoids that "philosophical" issue by outright specifying the page BP384 is for "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse, because that would clutter up the simplicity of the Problem with irrelevant noise. If we really wanted to, we could create another different BP page for that slightly different solution, "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots." However, this would not be a very interesting addition to the OEBP--we don't want loads of slightly different copies of Problems--in particular this could be considered a "overriddensolution" (left-BP568) page. Even when all examples in an image of BP384 are helpful, with dots arranged in ways that make it easy to see whether they are or are not a square number, it still seems intuitive to sort left along with the helpful examples of square numbers of dots any new, non-helpfully arranged example that has a square number of dots.

Problems should be marked as helpful when the OEBP wants all examples on at least one side to be helpful, rather than when just one or two existing examples are helpful.

BP Pages on the OEBP where users are advised to upload examples that help humans (by hinting at the solution) vs. other BP Pages.

BP Pages on the OEBP where users are advised to only upload examples that help humans (by hinting at the solution) vs. other BP Pages.

Left examples have the keyword "help" on the OEBP.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged in a helpful way.

Helpfulness can be a spectrum; most Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

When all examples in a Bongard Problem are helpful, it becomes unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots"? The OEBP avoids that "philosophical" issue by outright specifying the page BP384 is for "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse, because that would clutter up the simplicity of the Problem with irrelevant noise. If we really wanted to, we could create another different BP page for that slightly different solution, "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots." However, this would not be a very interesting addition to the OEBP--we don't want loads of slightly different copies of Problems--in particular this could be considered a "nonsolution" (left-BP568) page. Even when all examples in an image of BP384 are helpful, with dots arranged in ways that make it easy to see whether they are or are not a square number, it still seems intuitive to sort left along with the helpful examples of square numbers of dots any new, non-helpfully arranged example that has a square number of dots.

Problems should be marked as helpful when the OEBP wants all examples on at least one side to be helpful, rather than when just one or two existing examples are helpful.

Left examples have the keyword "help" on the OEBP.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged in a helpful way.

Helpfulness can be a spectrum; most Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

When all examples in a Bongard Problem are helpful, it becomes unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots"? The OEBP avoids that "philosophical" issue by outright specifying the page BP384 is for "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse, because that would clutter up the simplicity of the Problem with irrelevant noise. If we really wanted to, we could create another different BP page for that slightly different solution, "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots." However, this would not be a very interesting addition to the OEBP--we don't want loads of slightly different copies of Problems--in particular this would be a "nonsolution" (left-BP568) page. Even when all examples in an image of BP384 are helpful, with dots arranged in ways that make it easy to see whether they are or are not a square number, it still seems intuitive to sort left along with the helpful examples of square numbers of dots any new, non-helpfully arranged example that has a square number of dots.

Problems should be marked as helpful when the OEBP wants all examples on at least one side to be helpful, rather than when just one or two existing examples are helpful.

BP Pages on the OEBP such that users are advised to only upload examples that help humans (by hinting at the solution) vs. other BP Pages.

Left examples have the keyword "help" on the OEBP.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged in a helpful way.

Helpfulness can be a spectrum; most Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

When all examples in a Bongard Problem are helpful, it becomes unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots"? The OEBP avoids that "philosophical" issue by outright specifying the page BP384 is for "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse, because that would clutter up the idea of the Problem with irrelevant noise. If we really wanted to, we could create a BP page for the slightly different solution, "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots." However, this would not be a very interesting addition to the OEBP--we don't want tons of slightly different copies of Problems--it would be a "nonsolution" (left-BP568) page. Even when all examples shown in BP384 are helpful, with dots arranged in ways that make it easy to see whether they are or are not a square number, it still seems intuitive to sort left along with the helpful examples of square numbers of dots any new, non-helpfully arranged example that has a square number of dots.

Problems should be marked as helpful when the OEBP wants all examples on at least one side to be helpful, rather than when just one or two existing examples are helpful.

Left examples have the keyword "help" on the OEBP.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged in a helpful way.

Helpfulness can be a spectrum; most Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

When all examples in a Bongard Problem are helpful, it becomes unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots"? The OEBP avoids that "philosophical" issue by outright specifying the page BP384 is for "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse, because that would clutter up the idea of the Problem with irrelevant noise. If we really wanted to, we could create a BP page for the slightly different solution, "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots." However, this would not be a very interesting addition to the OEBP--we don't want tons of slightly different copies of Problems--it would be a "nonsolution" (left-BP568) page. Even when all examples shown in BP384 are helpful, with dots arranged in ways that make it easy to see whether they are or are not a square number, it would still seem intuitive to sort left any new, non-helpfully arranged example that has a square number of dots.

Problems should be marked as helpful when the OEBP wants all examples on at least one side to be helpful, rather than when just one or two existing examples are helpful.

Left examples have the keyword "help" on the OEBP.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged in a helpful way.

Helpfulness can be a spectrum; most Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

When all examples in a Bongard Problem are helpful, it becomes unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots"? The OEBP avoids that "philosophical" issue by outright specifying the page BP384 is for "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse, because that would clutter up the idea of the Problem with irrelevant noise. If you (a user) really wanted to, you could create a separate BP page for the slightly different solution, "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots." However, this would not be a very interesting addition to the OEBP; we don't want tons of slightly different copies of Problems; it would be a "nonsolution" (left-BP568) page. Even when all examples in this Bongard Problem are helpful, explicitly showing dots arranged in ways that makes it easy to see whether they are or are not a square number, it seems intuitive to sort a new, non-helpfully arranged example, with a square number of dots, on the left.

Problems should be marked as helpful when the OEBP wants all examples on at least one side to be helpful, rather than when just one or two existing examples are helpful.

Left examples have the keyword "help" on the OEBP.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged in a helpful way.

Helpfulness can be a spectrum; most Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

When all examples in a Bongard Problem are helpful, it becomes unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots"? The OEBP avoids that "philosophical" issue by outright specifying the page BP384 is for "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse. If you (a user) really wanted to, you could create a separate BP page for the slightly different solution, "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots." However, this would not be a very interesting addition to the OEBP; we don't want tons of slightly different copies of Problems; it would be a "nonsolution" (left-BP568) page. Even when all examples in this Bongard Problem are helpful, explicitly showing dots arranged in ways that makes it easy to see whether they are or are not a square number, it seems intuitive to sort a new, non-helpfully arranged example, with a square number of dots, on the left.

Problems should be marked as helpful when the OEBP wants all examples on at least one side to be helpful, rather than when just one or two existing examples are helpful.

Left examples have the keyword "help" on the OEBP.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged in a helpful way.

Helpfulness can be a spectrum; most Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

When all examples in a Bongard Problem are helpful, it becomes unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots"? The OEBP avoids that "philosophical" issue by outright specifying the page BP384 is about "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse. If you (a user) really wanted to, you could create a separate BP page for the slightly different solution, "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots." However, this would not be a very interesting addition to the OEBP; we don't want tons of slightly different copies of Problems; it would be a "nonsolution" (left-BP568) page. Even when all examples in this Bongard Problem are helpful, explicitly showing dots arranged in ways that makes it easy to see whether they are or are not a square number, it seems intuitive to sort a new, non-helpfully arranged example, with a square number of dots, on the left.

Problems should be marked as helpful when the OEBP wants all examples on at least one side to be helpful, rather than when just one or two existing examples are helpful.

Left examples have the keyword "help" on the OEBP.

There is a potential harder version of any helpful problem that is not helpful. For example, BP384 (square number of dots versus non-square number of dots) would be much harder if all examples had hundreds of dots that weren't arranged in a helpful way.

Helpfulness can be a spectrum; most Problems are helpful to some degree by not including the most convoluted unintelligible possible examples.

When all examples in a Bongard Problem are helpful, it becomes unclear whether the helpfulness is part of the Bongard Problem's solution. For example, is the solution to BP384 "square number of dots vs. non-square number of dots," or is it "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots"? Since the OEBP is about Bongard Problem solution ideas, we avoid that "philosophical" issue by outright specifying that the page BP384 is about "square number of dots vs. non-square number of dots." The "help" keyword is just a fluffy administrative tag on top of that, which says the OEBP doesn't want users to submit any examples to the OEBP that are hard for humans to parse. If you (a user) really wanted to, you could create a separate BP page for the slightly different solution, "square number of dots (that are arranged in a helpful way so as to communicate the square-ness) vs. non-square number of dots." However, this would not be a very interesting addition to the OEBP; we don't want tons of slightly different copies of Problems; it would be a "nonsolution" (left-BP568) page. Even when all examples in this Bongard Problem are helpful, explicitly showing dots arranged in ways that makes it easy to see whether they are or are not a square number, it seems intuitive to sort a new, non-helpfully arranged example, with a square number of dots, on the left.

Problems should be marked as helpful when the OEBP wants all examples on at least one side to be helpful, rather than when just one or two existing examples are helpful.