CatZ on: Serral #1 on KR, heading into Blizzcon

Published on 10/09/2017 10:37 PDT by Paulo "CatZ" Vizcarra

Intro

From time to time I find myself writing LONG ass responses on forums, from here on when I do, I will try to adapt them and post them as an article of opinion. This one is a response to on-going conversation around Serral being #1 on the Korean ladder and what that means or implies for his chances headed into Blizzcon or his overall status as one of the best in the landscape of competitive StarCraft 2.


SpeCiaL (Blizzcon participant, one of the best non-korean Terrans) was recently quoted in a stream chat, proc-ting someone to make a reddit thread based on his opinion titled "MajOr's Opinion of Serral being #1"


I will now take a few opinions / reactions on said thread and attempt at a detailed response:


For starters I would argue that being #1 on the ladder absolutely DOES hold meaning, indisputably, it means at least the following:

The Match Making Raiting (MMR) system (excusing anyone who has left GM) has determined that based on your average online performance over hundreds of games, you are out-performing everyone else in the server at a given point in time, in this instance - currently; present, on going. While I would consider this one of the strongest and most reliable factors to take into consideration when assessing someone's (in this case Serral's) chances of success at Blizzcon or claims to being "The Best" currently, It is also indisputably and as Nathanias implies: not the only thing that matters, next we'll dive into some examples that could be factored into a player's overall performance.

[Note: on a tangent related to Nate's comment above, Ladder rank / MMR these days has to do less with grinding and more with performance, and while there is a co-relation between 'grinding' and performance because we get better as we do more of anything, the MMR system won't care much about how many games you play when contrasted to who you beat and who you lose to]

Ladder vs Tournament


While it is fair to say that ladder performance =/= tournament performance, ladder is without a doubt where most games of Starcraft take place and thus the largest sample size for a player's skill at a given point in time. For most of us who don't actively compete in tournaments it is the best and the only measurement/reflection of our own skill contrasted to everyone else who chooses to play on the ladder plus there isn't one single pro-gamer I know of who doesn't ladder.

Tournaments to varying degrees will introduce other factors such as preparation or intel gathered on your opponents and their trends - much less common factors in a ladder game.

In a tournament setting especially for the round of 16/group stage you can take into account these factors to modify and adapt your play according to your perceived strengths and ideally your opponent's weaknesses.

Other (perhaps smaller) factors to take into consideration are things like experience on stage, nerves, jet lag, or if you are sick vs feeling well on that given day, hell, what you had for breakfast or the RedBull you did or didn't drink will have an effect on your performance for better or worse (Not to be understood as will win or lose you the game, but it certainly could have a say on the matter alongside a sea of factors).

Even overall mentality and believing in your own ability to succeed can have an impact in your performance - another one of my recent articles explains my detailed thoughts on that: here

One could also make the argument that people "try harder" in tournaments or perform best under pressure, but an opposing argument for comfort can just as easily be made.



Who is, in fact, "The Best"?

The following comment is alluding to MajOr's initial thoughts, which state being #1 on the ladder holds no meaning / doesn't mean you're the best.


We often think of the best tournament performers as the best players, I imagine we put a larger emphasis on Tournament game-play because it's where the money is and it is how most competitive industries are built with spectators/witnesses in mind.

Individually however, we all have different ideas to take into account when we form our opinions on who "The Best" player, athlete or expert of any given discipline is. Can one say the best performing player on the ladder is the best player? Well, you can say whatever you want!

Many people like to think MMR systems are often bad, the SC2 one doesn't get complained about as much as other games where you can blame your team for losses and dilute yourself into thinking you're better than everyone in your bracket. And while perfection may be unattainable as far as MMR systems are concerned (or most if not all things for that matter), "MMR" is still the best approach we have, in StarCraft, to ranking players on a consistent basis based on their performance / play pitted against others and it takes anyone that wants to be a part of it into account.

Ultimately, "The best" is a relative term and thus it's impossible to assert that anyone is "the best" at anything aside from personal opinion. While personal opinion can be based on facts/past happenings, our ideas of "the best" don't have to align.

For example, If I ask a bunch of imaginary people who is the best StarCraft 2 player we could reasonably expect to have some answers along these lines:

"Innovation" and provide some reasoning such as: "he has been most dominant (In tournaments, often times in ladder) for the last few years"

or: "Rogue" He has impressed me the most recently and he is the most recent champion in Korea, considered the best region.

or: "Stats" He is rank #1 in Korea Standings and thus had the most dominant year in tournament play in Korea.

or: "Neeb" He is the #1 ranked player on Aligulac which takes into account offline AND online performances and non-'televised' tournaments.

or: "soO" He hasnt won a GSL but he's been in the finals most consistently!

or: "MVP" because he was the most successful player at the time when most people played the game

or: "MC" Because he's made the most money from the game

or: "ByuN" because he is the current world champion / won the most prestigious event, which happens just 1 time per year.

or: "Serral" because he currently rank 1 in the KR ladder, which has the largest amount of recorded interactions between players and is regarded as the most competitive ladder.

or: "llllllllllllll" because I just saw him play against innovation rogue serral and neeb on stream and he beat them all

To many different people, "The best" might mean different things or combinations of factors the person considers relevant to their equation, think of this:

Working under the assumption that MVP was the best player at X point in his career, when did MVP stop being the best player? Was it in the finals of the last tournament he managed to win? Was the last person to beat him in a finals or in his last game of the ladder better or worse than MVP minutes before beating him? (that person's skill likely didn't increase significantly in 30 minutes, it's just that you learned of his capabilities at that point)

Ladder is a metric, much like in all of the examples mentioned above we could use different other metrics or combinations of them to form our ideas on who the best player is, none of these factors mentioned throughout mean anyone is "the best" simply because "who is the best X?" isn't a question that has an objective answer. This doesn't mean that you should stop saying what you think or who you think is the best at something using whatever limited information you can gather to justify your reasoning, but your opinion =/= fact.


A Blizzcon/Math Experiment:



I am no mathematician, in fact I always sucked at math, but hey, I have google and a calculator, so let's give this a shot.

IF

we are to ignore all other factors and use the ladder as a sole indicator for one's skill level or strength in an off-line tournament such as Blizzcon. Being #1 on the ladder DOES NOT mean that player will win, and while the player's chances to win would be higher than everyone else's (with this sole metric in mind)... In a game as competitive as StarCraft (meaning there are many people playing at an extremely high and similar level) anyone's chances to win are relatively slim. As an example betting sites who offer lines of play on "Who will win the event" have Serral paying x11, or for the Soccer World Cup an extremely competitive sport, you have EVERY team paying at LEAST x5 or x6 to win the whole thing.

Why? 

Well simply put a favorite or whoever is widely considered "The best" won't (almost ever) have a 50%+ chance to win a tournament in a highly competitive field, especially when interacting / competing directly against others in a tug-of-war type situation as opposed to a 'best record' based discipline.

Yea, yea, but Why?


Ok, so lets take Serral's ladder case as an example:

Serral currently has a 74% winrate on the Korean Ladder, I think it is fair to assume that the field at Blizzcon is far more competitive than the Korean Ladder in that every player attending is capable of being ranked high up there (For example KelaGOD, considered one of the biggest underdogs, currently sits very close to the top 10) and on the ladder you WILL play many players who aren't nearly as good, thus inflating your win %.

For the sake of simplicity, lets work with 74%, which is a very generous estimation given what I just explained.

IF

Serral has a 74% chance to win against any/every other player at Blizzcon (which again, he probably doesn't).

There are 3 scenarios in which he advances from groups:

To Advance First:

1. WIN (74% of 100%)->WIN (74% of 74%)=54% of the time

To Advance Second:

2. WIN (74% of 100%)->LOSE (26% of 74%)=19%->WIN (74% of 19%)=14% of the time
3. LOSE (26% of 100%)->WIN (74% of 26%)=19%->WIN (74% of 19%)=14% of the time


Overall with Serral having a 74% winrate over anyone at Blizzcon (to stress, is not the case), his chances to advance from groups would be a stunning 82% 

Now he would be in the RO8, where he would have a 74% of 82% chance to advance to the Semis giving him a 60% chance.

His chances to advance to the finals would then be 74% of 60% = 44%

and his chances to win in the finals would be 74% of 44% = 32%

In this example Serral would be by FAR, "the best" player in the world (having the highest chances out of anyone going into the tournament) because of the inflated win-rates from ladder, in reality 55-60% winrate is already insanely high for SC2, but as you can see, even with a ridiculous 74% chance to win over anyone, he would stand less than a 1 in 3 chance to win the whole thing.

So naturally, while being #1 in the ladder or ANY other factor individually won't always or even usually translate into "out performing" everyone else at any isolated event, but that doesn't mean that it is insignificant / meaningless or that a player's chances should be evaluated independent of these factors, all of them and however you want to weight them SHOULD have a say in the matter if you are a betting man (or woman). And out of these many factors mentioned throughout this article I hope that it is clear that I believe ladder performance is by far one of the most meaningful ones, while it may not be determining on it's own, it certainly doesn't "say nothing".



...is he, though?


[Note: For simplicity's sake the fictitious math is based upon BO1 competition, the longer the series, the more favored a favored player becomes because they stand a higher chance to win each individual game, to re-iterate, I don't think ANYONE at Blizzcon has close to 74% chance to win over everyone else]

CatZ 168