By Dio Ariadi | datawizart.com, August 2021
For whose doesn’t know about Shout Korean music show this is the
place where Kpop artist will perform their song and compete each other
in the broadcasting place such as SBS, MBC, and other. In the end of the
show they will announce who win the show based on some criteria.
Winning the music show for the first time can be leading indicator a
greater chance for the artist and agency that public notice the artist
and the song they produced. The positive side effect can be be generated
more fans and translated into increasing sales album or merchandise.
There are multiple group debut every year and unfortunately not every group will win the show eventually. We take a look on some variable that available that might effect the chance the group to win the show, some of the variable provided by pudding.cool on this article Why are K-pop groups so big?, such as international casting, from survival show or not, size of group, and gender. During the research we noticed that every article mentioned about big 3 SM, JYP, and YG. These 3 along with BigHit (BTS’s agency) will become the variable for this analysis.
From what we gathered there are 251 band where 71 of them have win some of music show. Some of the agency can have more than one group that win, the way they created the group itself can be different. Agency can create group using survival shows and also international casting.
If we split the proportion win and never/still try to win using other dimension such as type of agency and the other, then we can see clearly which variable that might be a good indicator who will win the music show. Here is how it’s look like the proportion where the bigger the green color means more group winning in the music show for that respective dimension or grid. For example grid Type of Group, the group that came from Top 4 Agency, 95% of them was win at least one music show, while the other (non Top 4 Agency) only has 22.5% that win music show (this analysis done on 18 August 2021).
The sequence of the order is based on importance variable, means Type of Group is variable that can differentiate really well which group will win and not compared to the other. See Methodology tab for more detail
Legend Win Trying/Disband
From previous analysis we see that when it comes who will win, type of agency will have the greatest impact. Big agencies will have more capital to spend on marketing and music producer. The second substantial impact is that if there is an international casting at least one even though music shows we count is music shows in South Korea, one of financial reason is that International casting (member outside South Korea) can also help the group expand their reach into a larger market. The latest significance is whether the group came from survival or not. Some of advantages for agencies using survival show are early exposure to the public, where the public can participate in choosing the idol they want. Meanwhile group size and gender does not give significant effect compare to others three.
Is there any group have combination of these three? Big agency, survival show and international casting? Yes, one of them is Twice.
If we count all the winner show the median days for them to wait until the first win was 384 days or 1 year from their debut. Interestingly if we split based on type of agency the difference is staggering Big Company need to wait less than one year or 238 days while the median other company is 483 days.
Note: Each dots represents each group, hover the dot to know the detail for each observation. Y-axis is days they had to wait to get the first win, the higher the dot means more time they need to wait.
From five variable that we gathered, it cannot describe really well variance why certain groups are faster than the other. The most significant difference is whether group comes out from survival show or not. A group band came from survival have median 168 days while the other is 565 days. Additional variable or information might needed to explain the variance between the group, it can be budget during promotion or the competition while they promoted the song.
Three other variable have the least influence on how fast the group will win. For Gender we remove mix due to small number of observation.
The data set is combination between group that mentioned on this article https://pudding.cool/2020/10/kpop/, with additional new groups that debut in 2020 and at least have more than two members. We also exclude Trouble Maker, GD X Taeyang, ToHeart, WJMK, Super M as this group created from existed group or combination from different group.
The sequence of order in Who win the music show variable based on random forest output. The first sequence means the most importance variable that able to differentiate between group who will win or not.
The sequence of the tab order in For those who win how long they need to wait? is based on linear regression where we pick based on the highest R-squared. The first tab means the most gain in term of R-Squared.