Our Six-Week Esports Analysts Interviewing Process
Great Esports organizations (players, coaches, & analysts) are built on top of great talent and culture. Whereas the best players can be identified by their competitive track record in the game and in tournaments, great analysts are harder to identify.
The top 5–10% of analysts will already be known through their participation in the competitive scene and community through activities like making videos and blogging, there are many talented individuals who are not only undiscovered, but also need some level of grooming to reach their full potential. This is exactly why we (GankStars) created a six week, competitive interview process to bring on new analysts.
The goals of this process were three fold:
- Allow open participation from the community, so even people who won’t make it the end get a chance to practice and interact with us (you never know who will be a late bloomer).
- Include several rounds that encompass all the skills an analyst needs, so that we can evaluate all areas of potential candidates and push them to improve upon their weaknesses.
- Introduce competition into the process, because in the end commitment will always beat talent when talent isn’t committed.
At the end of this process, we accepted two analysts: Vyzeox and Wavernot. We said we would only accept one but they were both so close and complementary we decided to take both. Below I cover how both did throughout the process and their respective strengths and weaknesses.
Part One: The Application
Part one of our process was the analyst application. Applicants were told they should be skill tier Simply Amazing (7 out of 10 in Vainglory) or higher, though I read through the answers of everyone who submitted an application regardless and would have accepted someone of a lower tier if they submitted the right answers (one person came very close). Applicants were asked to provide some background information on themselves and to provide answers to three simple but difficult questions to test their understanding of the game. View full answers here.
The first question:
Calculate DPS (Damage Per Second) for basic attacks for lvl 9 Fortress with one Sorrowblade and Blazing Salvo.
Is a question with a numerical answer that touches on many areas of the game but has a couple tricks.
The second question:
Is it better to buy Sorrowblade or Tyrant’s Monocle for Taka lvl 10 with a Serpent’s Mask?
Is a why question that requires you to weigh more than one factor and explain your thinking.
The third question:
Is there a better purchase for the same gold?
Is a broad question meant to see how far an applicant would go and how many different scenarios he would to consider to find the most correct answer possible.
We selected three individuals out of eleven applications to move on to the second part of the process. Almost all of the eleven applications we received showed strong knowledge of the game, and the top four were all close to perfect. Below is the spreadsheet I used to review all the applicants and check for the specific tricks I put into the application questions.
How Vyzeox Did
Vyzeox was one of the first people who reached out to me at all. He messaged me directly on Line where all of my guild mates hang out. From the very beginning I was impressed by his hustle. He asked me a bunch of questions to clarify what I was looking for in the app and proceeded to crank out the answers far ahead of the actual deadline. Because of his quick work, I was happy to provide feedback and point out a couple of mistakes he made to which he quickly corrected. This significantly improved his analysis. I was also impressed when I learned about his involvement with VGL as their video editor. In fact, he even used his video editing skills to get some of the numbers he used in his application. Nothing gets you more brownie points in my book than testing numbers first-hand in the game. All in all Vyzeox came across to me as a strong applicant and I was excited to see him continue into the next round.
How Wavernot Did
Wavernot was a very interesting candidate from the beginning. His application was very poorly written. He also submitted it twice which confused me (he decided to add more details before the deadline). His app probably took me the longest to read through and understand, but he ended up having the best answers and reasoning out of anyone. I was honestly surprised it turned out so well because the first impression I got before I digging into his app was not favorable. Wavernot also tested many of his numbers first hand in the game like Vyzeox did and he attached several in-depth spreadsheets to his application which got him brownie points and helped show me what he was capable of.
Part Two: Two Essays
Three candidates from part one advanced onto Part 2. Each candidate was provided direct and actionable feedback by me on their strengths, weaknesses, and what he needed to improve in Part 2 to be successful.
The assignment this time was for the candidates to choose two areas of the game they wanted to analyze and write up to five pages on each area including recommendations for the players. They had a little less than a week to complete them. I told them to read the posts on my blog as examples as level of depth required and provided them the build optimization research I did for CullTheMeek.
The purpose of this part was to test if candidates could come up with a fully completed deliverable. If they could ask the right questions and come up with in-depth answers. One of our pro teams’ captains, FooJee, joined me in coaching and providing 1 on 1 feedback to each candidate throughout the process. We directly commented and picked apart the formatting and analysis of their essays through Google Docs. I’d estimate we easily provided over 100 comments per candidate across the two essays.
After the essays were handed in, the judging was turned over to the competitive players in GankStars. The identity of the candidates was kept a secret to make sure the players evaluated the essays without any bias about the background of the candidates. Neither FooJee or I were part of the judging for this reason. This is a matter of process and not because anyone specifically had anything against any specific candidate.
Six Players provided feedback and rated every single essay 1–5 on three dimensions: Presentation, Analysis, and Impact. These ratings were then averaged to score each essay.
1. Presentation: How easily were you able to understand the arguments and data used in this article?
2. Analysis: To what extent do you agree with the conclusions in this article?
3. Impact: To what extent will you change your strategies or gameplay based on this article?
Three out of six essays received perfect scores on presentation. I really hammered the candidates on this and told them insights are worth nothing if players can’t quickly understand them. Next, when it came to analysis, four out of the six essays had players agree with 87% or more of the recommendations. Another key thing to get right and evidence that the candidates left no stones unturned and considered all potential angles of their analysis. Finally, the hardest part to get right and where most struggled was when it came to impact. The top essay had only 53% of players say they would change their strategies based on it, and most scored in the 40s.
Impact is definitely the hardest metric to succeed at but it’s also the most important for analysts. We intentionally did not suggest topics for candidates to choose so that they would have to use their intuition to decide what mattered and what didn’t. Several essays were rated highly on analysis but low on impact because the players already understood what the essay showed, just didn’t have the numbers to back it up. Another factor that hurt some essays was that they were too specific/narrow, ie. they had a big impact on a few players but not the rest. Ironically, the essay that was rated the lowest on analysis could have been rated the highest on impact if it was more convincing. This makes sense when you realize that an essay that changes someone’s strategies must set out to prove ideas that players currently disagree with.
How Vyzeox Did:
Vyzeox won this round with flying colors. His essays were #1 and #3 overall. The most striking pattern across this round was that the essays that received the most comments and feedback from FooJee and I also scored the highest. There was a direct correlation. Vyzeox hustled more than anyone else in this round. He finished his essays first and pushed FooJee and I for more feedback than any other candidate. He also implemented the feedback extremely fast and did not get hung up when we pushed him outside of his comfort zone. Both of his essays received perfect scores on presentation and were close to the top in analysis and impact. Overall he continued to be a well rounded candidate and improved in all areas.
How Wavernot Did:
Wavernot came in second this round and did very well. His essays were #2 and #5 overall. From the previous round I told him I was extremely concerned about the presentation dimension of his essays. It was his clear weakness and the thing that would prevent him from being able to join our team. We spent a TON of time on it going back and forth, rethinking how to present certain ideas. This definitely held up the process and prevented FooJee and myself from spending enough time hashing out his second essay. While his essays scored the lowest among the three candidates in presentation (93% and 83%) I’m really proud of the huge improvements he made. He easily improved more than anyone else when it came to presentation and communication. And when it came to analysis, Wavernot performed the best like he did in round 1. His essays were #1 and #2 on the analysis dimension. FooJee joked that Wavernot’s Gold Bounty Analysis may have revealed insights even SEMC (the creators of Vainglory) were unaware of. Finally, Wavernot was even with Vyzeox when it came to impact. Overall, Wavernot’s essays were only 1% and 4% behind Vyzeox’s so it was very close.
Part Three: SWOT Analysis
Two of the three candidates from Part Two moved on to Part Three: Vyzeox and Wavernot. They were provided with the full results from Part Two and I reviewed with them their standing and what factors placed them in that standing.
Both Candidates were told that this was the final round and we would only take one of them in the end. This was our intention going into this round and I will explain later how that changes and why we took both of them.
This round required both candidates to study the VIPL Tournament and create a SWOT Analysis for GankStars Sirius (IraqiZorro, CullTheMeek, and Gabevizzle). SWOT stands for Strengths, Weaknesses, Opportunities, and Threats. It’s a simple framework used in many industries to analyze any organization and make recommendations for how that organizations can perform at it’s best. In this SWOT the candidates were required to study Ardent Alliance and Sirius to provide recommendations.
This assignment is much broader in scope than the round before it and required a lot more maturity and understanding of the game. They were told to focus on video analysis and player analysis, citing specific matches and game times for each of their arguments. I did a SWOT Analysis for GankStars Vega after the VGL so they were able to reference it as an example. There was no restriction on the length of this assignment and again the candidates had access to FooJee and I for feedback and review.
How Vyzeox And Wavernot Did:
When I received the reports from both candidates, I had no idea who was going to win. It was incredibly close in my eyes. They were about equal in the amount of feedback they requested from FooJee and I, so in that sense Vyzeox didn’t out hustle Wavernot as much as he did in the prior round. However, Vyzeox’s report was twice as long as Wavernot’s and surprised me with a new section he came up with on his own. When it came to presentation, Vyzeox’s report was like a work of art. He put a lot of time into making his report easy to comprehend as a whole and jump from part to part so that each player could absorb 100% of his work. But Wavernot didn’t do so bad on presentation this time. Actually, compared to his past work he completely knocked it out of the park in presentation. I was shocked that I was reading a report from the same person. In addition, seeing Wavernot’s strength in round two on the analysis component, this was sure to be a tight race.
To review the reports GankStars Sirius and I met up at their apartment in Korea (I was able to be in Seoul during VIPL) and we all went to a nearby Burger King to read the reports together. I personally prefer authentic Korean food but I digress. We spent an hour and went line by line through both SWOT Analyses. At this point, the players still had no idea who wrote which report so the reports were judged on their own merits. At the conclusion of the review, every member of Sirius was incredibly impressed. In fact, we all wanted to take both candidates and had difficulty deciding. But I forced the players to choose one report and vote by writing down their answer on a piece of paper so that they wouldn’t be influenced by each other. One candidate DID beat the other by a small margin. And YES the person who lost probably could have won if he put in a couple more hours or had more feedback. But I’ll leave it at that :). You all can speculate who won and you can follow both of them on Twitter to see who’s work you like better. I’m encouraging them to blog their work from Part Two of the process.
Part Four: Culture Fit
In addition the trials above, both analyst candidates were required to go through the normal GankStars casual member application process. They were required to play with our casual members to ensure they were not only a good analyst, but that GankStars would be the right guild for them to join and be successful in. Even if someone is the most talented analyst I’ve ever met, he or she will not be accepted unless he or she is also a good fit for the GankStars organization and culture. This is for his or her own benefit as much as it is ours.
Why We Took Two Analysts
In all honesty, I want to take every talented individual that applies to our team. There’s no reason why only one person had to win. For me, more important is that every person on our team is at a high enough level. I would have rejected every single candidate if no one was able to prove to us they could do the job. However, the reason why we were only going to take one more person initially was because we only had two teams. So two analysts (including me) made sense.
But right as VIPL ended we were given news that SEMC and VGL had changed the rules, allowing guilds to enroll up to three teams in the VGL tournament. Since GankStars has a deep roster of talented players, a new team was assembled immediately and the amount of analysis needed to be done by the Analysts Team increased by 50% overnight. After consulting my partners, it was decided that we could take on both candidates. Had this not happened I would have stayed in touch with the candidate we did not take and hoped to bring him on in the near future. I also plan to stay in touch with EVERYONE who applied to help them grow and improve. As GankStars expands competitively so will our Analysts Team.
