Slingshot Readers,

We NEED your support. More specifically, the author of this article needs your support. If you've been enjoying our content, you know that a lot of work goes into our stories and although it may be a work of passion, writers gotta eat. If just half our readers gave 1 DOLLAR a month, one measly dollar, we could fund all the work from StuChiu, DeKay, Emily, Andrew (and even Vince). If you contribute 5 DOLLARS a month, we invite you to join our Discord and hang with the team. We wouldn't bother you like this if we didn't need your help and you can feel good knowing that 100% of your donation goes to the writers. We'd really appreciate your support. After all, you're what makes all this happen. Learn more

Don’t talk to me about scrim results

With the scrim leaks of Team SoloMid and SK Telecom T1 gracing the front page of Reddit this month, I want to urge you not to take them at face value.

About a year ago, I had the opportunity to coach an amateur team of North American players ranked from diamond to challenger. It was my first time being in a role so closely involved with the players and having authority over them. Because of that, the two co-owners — both of which had coaching experience of at least the Challenger Series level — focused on trying to teach me how to coach. They offered me ideas for exercises, how to communicate with the players, and when I needed to put my foot down.

The team had a problem with picking fights when they shouldn’t have. They suffered from “see champion, kill champion” syndrome, giving little thought to what objectives on the map they were fighting for. After a hilariously bad scrim with about two kills per minute in a 30-minute game, I was instructed to deal with the matter. One of my mentors told me to give the following exercise:

The next game, the team was only allowed to have one fight every five minutes.

They could trade and respond to engages, but the entire team could only engage a fight or go for in all in once every five minutes. If a player did, the entire team had to wait before it could do so again. If a player died, then the timer was reset.

Needless to say, the team slowly lost the game, and that was fine. The point wasn’t to win: it was to learn the weight of each fight and realize its importance; to think of how to maximize that precious engage to get as much as they could out of it.

I don’t know if this kind of exercise is good, or even utilized at the professional level. I only know it was recommended to me by someone with coaching experience. If this sort of exercise┬áis normal, then you should always be suspicious of any scrim result. You simply do not know what mad science experiment a team is doing.

Cover photo: Screenshot


Leave a Reply