Are All Championship Runs Created Equal?

Published on 12/11/2018 06:59 PST by Topher Doll

Serral has just accomplished the impossible by winning the WCS Championship at Blizzcon. In doing so he torched through the best Korea could throw at him. From Zest and sOs in the groups stages to the Zerg masters Dark and Rogue with the most consistent player of Legacy of the Void, the man with more Premier finals appearances than any other player, the Shield of Auir who took the god Serral to seven games at GSL vs the World, Stats. But even Stats fell to him, marking the greatest upset in Starcraft history. But looking at the names he beat, this was no weak run where he got lucky on one matchup or facing bracket luck. No, this was truly a great run, regardless of who won it. Had Stats, Dark or Maru made this run it would still be incredible. 


This run got me thinking, are all tournament runs created equal? The answer is clearly "no" because sometimes a player faces weaker competition or fewer opponents. But I wanted to examine this, what are some of the greater runs in recent memory? What was a weak one? I wanted to delve this but knew it would be hard to do without some measurement because otherwise subjectiveness slips in. 


The Methodology


While not perfect I felt the need to use Aligulac ratings. I understand there are odd rankings at times and I also understand that there are weaknesses in that it weights offline and online events the same but on the whole, the big picture, it is a fairly effective measure of who is the better player by order of magnitute. So once I figured this would be the best option going forward the easiest option would be to just take an average of the opponents the champion beat along his way. This was just a pure way to measure the quality of opponent they faced. I then also felt that it might be interesting to see a cumulative  accomplishment of a tournament run so I took the sum of the Aligulac scores of all of their opponents. This measure doesn't necessarily reward playing the highest quality opponents but it does reward players who had a long run to win their championship. So a tournament like WESG or the IEM World Championship may see the winner play weaker competion but they have to play far more series than say the champion of GSL vs the World. These two metrics provide very different, but important measures of the quality of a tournament run and obviously the greatest tournament runs are both long and full of high quality opponents.


Lastly I wanted to include two counting metrics, the number of opponents faced who had over 2,500 (a top foreigner or very good Korean) and the number of opponents over 2,250 score (mid tier Koreans and good foreigners). For example the WCS champion in 2018 (so basically Serral) had an average opponent Aligulac score of 2,310 while the average GSL champion (Maru) had an opponent Aligulac score of 2,512. This also doesn't measure how well the player performed in that tournament run, they could have not dropped a single game or it could have been match point the whole way.


I have been using Aligulac for a number of years now and while they are incredibly helpful the numbers also suffer from inflation. That is to say it is hard to compare the Aligulac score of a WCS run in 2013 to 2018, the numbers have grown too much. I am still working on a way to create a floating average so we can compare across eras right now I am going to limit this study to just 2017 and 2018. It wouldn't be fair to say TaeJa's 2014 Shenzhen run where he beat Solar, Jaedong, Zest, Life and MMA being considered weaker than any WCS event this year. Even going back two years to ByuN and Dark dominating Korea in 2016 looks weak compared to the modern Aligulac score. 


Having said that I may look at the strongest tournament run in two year groups so I may do an article looking at the toughest tournament runs for 2015 and 2016.


The Data


First let us look at the 10 weakest Premier championship runs since the start of 2017 by measure of average Aligulac score:


Event
YearPlayersWinnerAvg Aligulac of OpponentsCombined AligulacOver 2250% of OpponentsOver 2500% of Opponents
WESG201724TY200216019337.5%00%
WCS Valencia201732Elazer211712700350%00%
WCS Austin201832Serral212912774116.7%00%
WCS Montreal201732Neeb216512988350%00%
WCS Jonkoping201732Neeb218213093116.7%116.7%
WCS Austin201732Neeb222313336233.3%233.3%
IEM XI Championship201724TY227820500777.8%00%
WCS Valencia201832Serral226313701466.7%116.7%
WCS Leipzig201832Serral229313759466.7%116.7%
HSC XVI201732Zest229918388562.5%225%


Should be no surprise that 7 of the 10 events are WCS events while two others were global events that locked in players from different regions and lastly HomeStory Cup, while awesome, rarely has only top players. Maybe the most interesting thing is while TY's 2017 year was incredible for being one of the highest earning years in Starcraft history but he did it through two relatively weak tournaments.


If we examine the weakest tournaments by combined opponent Aligulac strength we get a quite a different picture with a few notable changes:


Event
YearPlayersWinnerAvg Aligulac of OpponentsCombined AligulacOver 2250% of OpponentsOver 2500% of Opponents
GSL vs the World
2017
16
INnoVation
2417
9666
3
75%
2
50%
GSL Super #1
2017
16
herO
2469
9874
4
100%
2
50%
GSL Super #2
2018
16
Classic
2471
9883
3
75%
2
50%
GSL Super #2
2017
16
Rogue
2527
10108
4
100%
3
75%
GSL vs the World
2018
16
Serral
2563
10251
3
75%
3
75%
GSL Super #1
2018
16
Stats
2682
10729
4
100%
4
100%
IEM XII Pyeong
2018
16
Scarlett
2417
12085
3
60%
2
40%
IEM XII Shanghai
2017
16
Rogue
2464
12318
4
80%
4
80%
WCS Valencia
2017
32
Elazer
2117
12700
3
50%
0
0%
WCS Austin
2018
32
Serral
2129
12774
1
16.7%
0
0%


What we see here is the result of facing far fewer opponents. In a 16 person tournament a player can face between 4 and 6 opponents (depending on if it has a group stage or not) while in a 32, or larger, tournament you face a minimum of 5, but often 7 or more, series. The GSL Super and vs the World tournaments are pure bracket stages so the winners of those events, while facing incredibly high opponent quality, had fewer series which lead to less accumulated opponent strength which measures potential for upset. 


I think it is time we move onto the strongest tournament runs though, starting with measuring by average opponent strength:


Event
YearPlayersWinnerAvg Aligulac of OpponentsCombined AligulacOver 2250% of OpponentsOver 2500% of Opponents
WCS Global Finals
2018
16
Serral
2684
13422
5
100%
5
100%
GSL Super #1
2018
16
Stats
2682
10729
4
100%
4
100%
IEM XII Championship
2018
24
Rogue
2620
20957
8
100%
6
75%
GSL Season 1
2018
32
Maru
2597
15583
6
100%
5
83.3%
GSL Season 2
2018
32
Maru
2589
12944
4
80%
4
80%
GSL vs the World
2018
16
Serral
2563
10251
3
75%
3
75%
WCS Global Finals
2017
16
Rogue
2561
12806
5
100%
3
60%
GSL Season 3
2018
32
Maru
2531
17717
6
85.7%
5
71.4%
GSL Super #2
2017
16
Rogue
2527
10108
4
100%
3
75%
GSL Super #2
2018
16
Classic
2471
9883
3
75%
2
50%


The surprising thing, to some, is to see Serral's WCS Finals run at the top but it makes sense. On his way to his ultimate achievement he also got some bracket luck, meaning he got the harder bracket, in that he faced no foreigners. This helped because it kept his average high because he faced off against the best Protoss in the world in Stats, the two best Korean Zergs in Dark and Rogue and two weird Protoss in Zest and sOs in the group stage. Had Serral faced a foreigner in the group stage it would have lowered his average. Stats' GSL Super Tournament in 2018 run was crazy as well where he beat Dark, herO, Rogue and ByuN. Overall this category of ranking favors Korean tournaments because they have higher Aligulac players in general competing but Serral's two wins are worth noting. Maru does stand atop with four runs in the top 10. Rogue has three while Serral has two.


Event
YearPlayersWinnerAvg Aligulac of OpponentsCombined AligulacOver 2250% of OpponentsOver 2500% of Opponents
WESG
2018
32
Maru
2416
21747
7
77.8%
2
22.2%
IEM XII Champ
2018
24
Rogue
2620
20957
8
100%
6
75%
IEM XI Champ
2017
24
TY
2278
20500
7
77.8%
0
0%
HSC XVI
2017
32
Zest
2299
18388
5
62.5%
2
25%
GSL Season 3
2018
32
Maru
2531
17717
6
85.7%
5
71.4%
GSL Season 2
2017
32
GuMiho
2439
17074
6
85.7%
4
57.1%
GSL Season 1
2017
32
Stats
2411
16879
5
71.4%
3
42.9%
GSL Season 3
2017
32
INnoVation
2405
16837
6
85.7%
3
42.9%
WESG
2017
24
TY
2002
16019
3
37.5%
0
0%
GSL Season 1
2018
32
Maru
2597
15583
6
100%
5
83.3%


Here we can see how a tournament filled with great players and also has more than 16 players can pay off. While the WESG may seem like a weak #1 but in that run Maru beat Dark, Serral (peak Serral no less), Reynor, Scarlett, Nerchio, ShoWTimE, Elazer, Semper and TIME. While that is a lot of foreigners that is pretty much a who's who of foreigners with only Semper and TIME being considered truly weak competition for Maru. So he basically beat the best Korean Zerg, the best foreigner and pretty much the rest of the top 5 foreigners. Combine this solid quality of opponents with the fact he had to play nine, NINE, series to win puts this is a great run. 


Obviously neither ranking system is perfect but a few tournament runs rank well in both. Rogue's IEM Championship run ranks 3rd in average opponent strength and 2nd in accumulated opponent strength. Maru's 2018 GSL Season 3 run ranks 8th in average opponent strength and 5th in accumulated opponent strength. If we create an average rank between the two main metrics and percentage of opponents with 2,250 and 2,500 Aligulac we get this chart.


Event
Year
Winner
Average Rank
IEM XII Championship
2018
Rogue
4.00
WCS Global Finals
2018
Serral
5.13
GSL Season 1
2018
Maru
5.25
GSL Season 3
2018
Maru
7.75
GSL Super #1
2018
Stats
7.88
GSL Season 2
2017
GuMiho
9.75
GSL Season 2
2018
Maru
9.75
WCS Global Finals
2017
Rogue
10.00
GSL Super #2
2017
Rogue
11.50
IEM XII Shanghai
2017
Rogue
12.50


While I understand this doesn't factor in matchup strengh, players who might be entering a tournament hot or other subjective points that greatly impact each and every Starcraft tournament I do think these two ranking methods combine to capture which tournaments were a bit easier for the winner and which may have been a mountain to climb.


So what are your thoughts, do you prefer average opponent strength or prefer to see how many points they left in their wake on the way to the top? Do you think a combined version is the best way to go or that both are a waste of time? Let me know!