I write about this topic basically once a year, and sometimes I find it boring.

The reason why I write it every year is because there are new microblogs or new blogs popping up every year.

Blog, light blog, Weibo, long Weibo, G+, etc.

So many. . .

My history of blogging probably starts in 2004.

I liked a girl back then and she started a blog on an education website, so I followed suit.

Later in 2005, I liked another girl and she used SPACE, so I also used SPACE.

SPACE was the first one I used for a long time. I used it until MSN abandoned SPACE. After switching to WordPress, I felt it didn’t look good and gave up...

Then I used BlogSpot to synchronize updates. Go to Douban Q space and Sina blog on Renren campus.

Now it is basically used for public disclosure, and Evernote is used for private archiving-oh, remind me, I have several things written in it and then forgot to sync it to Evernote. . .

For light blogs, I have used Sina Qing and the originator Tumblr.

In terms of Weibo, domestic ones include Sina Weibo, Tencent Weibo, NetEase Weibo, and foreign ones include Plurk and Twitter.

After using so many miscellaneous things, there will always be some opinions.

Definition 1: The weight of information refers to the combination of the amount of information itself, the operating costs of information browsing and communication media, and the complexity of the information presentation method.

In terms of distance, Weibo and long Weibo are basically the same in all aspects, but Weibo has a maximum of 140 characters, and long Weibo can hold up your browser at most... So Weibo and long Weibo The weight is different.

Moreover, Weibo can be viewed directly. For long Weibo, you need to click to display the larger image. From this point of view, even if the same Weibo is used, it can be converted into a long Weibo image with the same content. Finally, the weight will also increase - because the cost of your browsing operation increases.

Let’s talk about something cross-border.

There are often people’s real-time updates on SNS. These updates are usually within 140 words (a beautiful classmate of mine once said to me: Isn’t Weibo just about everyone’s personal signature? Why is it so popular? ? Then I explained to her the principles here while imagining her deep V coat, and then there was no more - friends, take this as a warning), so in terms of the amount of text information! , the personal status of SNS is the same as that of Weibo. But I don’t know if you have noticed that if you want to interact with personal status in SNS, you generally need to click in, then comment, or forward - of course, likes can be done directly on the homepage TL.

Therefore, the operating cost of personal signatures and other real-time updates in SNS is higher than that of Weibo, so it is heavier than Weibo.

Blogs are even more so.

Generally, people who write blogs have far more words in their posts than blogs - didn't Mr. Jian recently call on everyone to "read a thousand words of writing every day, and think like a blowout pen and pee"? (How about it, my sentence is very smooth, right? Hahahahaha~~~~~) What’s more impressive is that generally you can’t see everything on the homepage, and you can’t see everything in the list of all articles (SPACE can do that) See the whole thing, so watching TL becomes a disaster...), you need to click in to see the whole thing and interact - so the operating cost increases again. So blogs are more important than Weibo...

So how do blogs compare with SNS?

This involves the third dimension - the complexity of presentation.

The only content in a blog is posts. The content in SNS is diverse, including posts, polls, advertisements, etc.

The variety of interactions determines that the entire SNS system has extra weight - users need to spend extra energy to distinguish what content is what, and this extra classification causes extra operations. Cost - But unlike the second item, the operating cost here is automatically completed by the brain and does not require you to click with the mouse.

This is a latent operating cost, but it is also a cost.

So, SNS is more important than blogs.

Of course, it’s not so certain. For example, if a blog has dense posts, but the SNS layout is relatively loose, then the SNS may actually be lighter for users - which brings up another point. Richness is not only related to the type of content, but also related to the density of the content.

The more types, the heavier the information; the smaller the density, the lighter the information.

Of course, richness is not limited to these contents. For light blogs and blogs, there is almost no difference in information type and information density, but the overall UI of a blog is richer than that of a light blog, which results in the additional weight given to information by the entire platform - but this Part of it is hard to quantify.

Of the above three weight indicators, one is from the information itself, and the last two are given by the platform for information dissemination. However, without a platform for information dissemination and production, there would be no information - but this reminds us that this weight can be corrected through a third-party APP.

Roughly speaking, on a device d, the weight W of a piece of information m on an information platform p can be written in the following form:

W(m, p, d) = Len(m) / 100 + (Ob(p) + Op(p)) × 2 + (1 + K(p)) / 2 × N(p, d) / M(d) × UI(p)

Where Len(m) is the length of m, Ob is the number of operations required to browse a piece of information, Op is the number of operations required to spread a piece of information, K is the number of content types provided by the platform, and N is the number of pages that can The number of information displayed - this value is not only related to the platform p, but also related to the device used by the user (PC? Mobile phone? Tablet?), but it is only recorded as p here. Correspondingly, M is the optimal display number of a page on a specified device, such as 10 on PC and tablet, 2 on mobile phone, etc. The final UI is the extra weight caused by the platform's UI.

Of course, some weight parameters, such as that M, are currently chosen randomly and may not necessarily be good values.

Now, for example, a 100-word Weibo and a 50-word Weibo have different weights - for Sina Weibo, Ob(sina)=1, but Op (sina)>1, because if we want to forward the original message of a forwarded message, additional operations are required (click "Original Weibo"). Let's assume that Op(sina)=1.1. For Twitter, we can think of Op(twi)=1. Regarding the types of information, Sina Weibo has Weibo, promotion accounts, advertisements, and friend activities. Therefore, counting the frequency weight, we can think that K(sina)=2.5 and K(twi)=1. The number displayed on each page varies depending on the theme. Let’s assume that N(sina, d) / M(d) = N(twi, d) / M(d) = 1.

So, on these two different platforms, the weights of these two pieces of information are:

W(100, sina) = 1 + (1 + 1.1) × 2 + (1 + 2.5) / 2 = 6.95

W(50, sina) = 0.5 + (1 + 1.1) × 2 + (1 + 2.5) / 2 = 6.45

W (100, twi) = 1 + (1 + 1) × 2 + (1 + 1) / 2 = 6.0

W(50, twi) = 0.5 + (1 + 1) × 2 + ( 1 + 1) / 2 = 5.5

For comparison, the weight of a thousand-character article is like this:

As far as the platform is concerned, the browsing cost is 2 (click to open and read) . The forwarding cost is higher than 2, estimated to be 2.5 (internal likes are 2, and then if you want to share to other platforms, additional clicks and confirmation processes are required, which is 4. Considering the weighting caused by the frequency of operations, it is estimated to be 2.5, which is about the same... ). The content type is 1 on the homepage, 2 (posting, replying) in TL, and 4 (posting, replying, liking, adding to anthologies or recommendations) in reminders. Generally speaking, 2.5 is good. As for N/M, if it counts as 1 on the homepage, it will exceed 2 in TL and reminders. Overall, it counts as 2. Therefore, the weight of the previous thousand-word article is:

W(1000, JS) = 10 + (2 + 2.5) × 2 + (1 + 2.5) / 2 × 2 = 22.5

It is really heavy...

Douban is probably about the same, but for SNS like Renrenxin, the K value is much higher, and the N/M is also much higher, so everyone applauded.

Let’s compare the weight of the same message as a long Weibo, light blog and blog.

Let’s take Thousand Character Essay as an example.

As a long Weibo, each parameter is similar to Weibo, so there are:

W(1000, pic) = 10 + (2 + 2.1) × 2 + (1 + 2.5) / 2 = 19.95

When the same content is placed on a light blog, if only the spread within the light blog is considered, then its weight is:

W(1000, qing) = 10 + (2 + 2) × 2 + (1 + 1) / 2 = 19

But obviously the communication in light blogs is not considered mainstream. The mainstream still needs to use Weibo to promote, so this part is considered The result is:

W(1000, qing+weibo) = 10 + (2 + 2.5) × 2 + (1 + 1) / 2 = 20

Phase Slightly heavier than Long Weibo.

For blogs, the weight can also be divided into the part that is only spread within the blog system and the part that is assisted by Weibo, respectively:

W(1000, blog) = 10 + (2 + 2) × 2 + (1 + 1) / 2 × 1.2 = 19.2

W(1000, blog) = 10 + (2 + 2) × 2 + (1 + 1) / 2 × 1.2 = 20.2

The extra factor 1.2 here compared to light blogs is the extra weight factor mentioned above because the overall UI of blogs is richer than light blogs.

The front is just a foreshadowing, and the following is the main meal.

Theorem 1: The growth rate of the number of users on a platform is inversely proportional to the average weight of information on the platform and directly proportional to the number of existing users.

It can be written in the form:

ΔU = V × U / Ave(W(m))

Here V is the growth constant. The average weight can be considered as the result of deducting the factors related to the information itself in W.

It can be seen that basically the lower the average weight of the platform, the more people will gather on it at the same time.

Theorem 2: The existence time and spread scope of a message increase with the number of platform users, increase with the increase of information quality, and decrease with the increase of information weight.

Maybe it can be written in the form:

Life(m) = LIFE / W(m) × Q(m) × F(U)

Range (m) = RANGE / W(m) × Q(m) × F(U)

Here LIFE and RANGE are two constants. F(U) is a function related to the number of platform users. At least it cannot be a linear function. It may be a function like this:

F(U) = 1 - 1 / (1 + (U / Limit ) ^ 2)

This means that the more serious the message, the smaller the spread range and the existence time, and the lighter the message, the wider the spread range and the longer the existence time.

For example, the same content will have different dissemination scope as long Weibo, light blog and blog.

Theorem 3: The frequency of user interaction with information is directly proportional to the existence time and spread range of the information.

That is:

Act(m) = ACT × Life(m) × Range(m)

Where ACT is the interaction constant. The characteristic of this theorem is that the quality of information does not obviously appear in the relationship - therefore, a bad message can cause enough interactions if it appears repeatedly enough (for example, a troll repeatedly sends the same message).

The above is only considered within the same system. If multiple platforms exist at the same time and users can choose between multiple platforms, the situation will be different:

< p> Theorem 4: Users are more likely to interact on platforms with more frequent message interactions.

Basically, the stickiness of platform p to users can be roughly expressed as:

Stick(m, p) = Act(m, p) / Sum(Act(m, q) , q)

So, when we compare Weibo, light blogs and blogs, we will find that because Weibo is the lightest, the information on Weibo is spread the most widely and the information exists for the longest time. Users are growing the fastest, and users interact with information the most frequently. As a result, people increasingly like to interact on the Weibo platform.

The above are all statistical guesswork.

Since it is statistics, there will naturally be differences among individuals; since it is a guess, it will always be wrong.

Well, above.

Oye, I finally finished writing all the nonsense, long live~~