Oh! Thanks, @GoldbergData @linuxenko (nice job by the way! You got a follower!) - be aware I don't consider myself an expert but probably that is a better reason my experience might be important to some.
Please don't think I don't have any interest in outliers.
One thing that could happen is that people copied and pasted automated messages.
However I went through some messages showing outlying behaviour just now and found that could be related to other factor too.
Although crowded, the actual number of seriously seriously engaged participants in the chat is not thaaaat big. The very engaged participants have been the main drivers of the characteristics of the messages and the topics.
One clear example is @abhisekp who was very very active in FCC and the chatroom and was a frequent emojier. He is, according to my list, one of the first "outliers". @CEREBR4L has been another one. There are some more.
They also served as moderators of the room and sometimes and more specially @abhisekp they used emojis to indicate the importance of the message they were posting (something new I am learning here thanks to your question, @huyingjie!!!).
I haven't seen all messages so I can't completely confirm this, but the few I checked suggested so.
This might require a deeper analysis for sure. Do you want to do that, @huyingjie? ;)
And welll... this is going to prove as a true trend... The list of those that made the top 5 at least once (:tada:!!!):
Some of them could have made the top 5 probably once. I haven't seen when - context could be relevant. But in general:
Looking at this, it might be that moderators could have had a larger impact than I was expecting and that there is a widespread use of "freecodecamp-branded" emojis (stars, sparkles, cookies, etc).
This, assuming that the list and counts are both correct (:scream:).
By getting rid of possible emojis that were "freecodecamp-branded", those which stood to the podium of the top5 were:
I think fire should be in the "freecodecamp-branded" list. Hmmm... Anyway.