EPT![[image loading]](http://www.teamliquid.net/staff/treehugger/proleague_590x300.jpg)
Graphics by shiroiusagi
Hello TL!
Since the Unholy Alliance was announced, I've been religiously following EG-TL's run through KeSPA's Proleague. Two of my favorite teams, joining forces to try and conquest the super-prestigious team league in Korea; no way I was going to miss it.
Now that the regular season has come to a close, I have taken some time to analyze the team's performance throughout the rounds. I hope this will prove interesting, and that it will help understand the team's progression through the league. Now, let's jump to the graphs.
The Big Picture
![[image loading]](http://i.imgur.com/CfdILVV.png?1)
Cumulative matches won by EG-TL, compared against all other teams
EG-TL started the first two months of Proleague with an OK showing. Many of us expected the team to crush all opposition, since the team had many SC2 seasoned veterans which would be facing BroodWar pros, who had only been practicing SC2 for a few months. However, EG-TL encountered very well prepared opponents, opponents who made up the lack of SC2 experience with hard work and prepared tactics. From the beginning of the league until the 3rd week of January, EG-TL won 6 out of the 13 matches played, earning them a respectable 4th place, tied with SKT1 and Team8.
However, after that week, their slump begun.
![[image loading]](http://i.imgur.com/ffKWzDm.png)
During these weeks, EG-TL players would see this screen way more often than not.
In the following weeks, EG-TL would win only a single match of the 8 that were played. They went from the 4th position to the dead last, and started to become the laughing stock of Proleague.
Fortunately for us the fans, the team decided not to call it quits just yet, and instead took some action. During the mid season break, Evil Geniuses added the famed Coach Park as the new EG-TL head coach. Only two days after that, PuMa was released from EG due to his poor overall performance. And less than two weeks after that, aLive and Oz joined EG in order to reinforce their roster.
The so-called Coach Park Effect took action very quickly, with EG-TL breaking out of their loosing streak and winning their first 2 matches under the new coach. From the start of Round 4 until June the 13th, the team won 7 out of their 10 matches, their best showing yet, giving hope to all of us fans, and even switching out of the last place from time to time.
Finally though, on June 13th the team was mathematically eliminated from the playoffs; the big (and somewhat unrealistic) goal that the new EG-TL had made for itself. With that, it seems like the team relaxed and lost focus, proceeding to win only 2 of their remaining 6 matches and finishing their Proleague run in the last place, tied with Team8.
Overall, the team showed that they can compete in Proleague. If their long mid-season slump and their lack of motivation during the final weeks had not occurred, the team would most likely have made it to the playoffs. If they can avoid those loosing streaks and their players can improve a bit (or they get better players), and if the Unholy Alliance holds for the next Proleague season, I think EG-TL will no doubt become a legit contestant for the playoffs.
Honorable mention to STX Soul (and INnoVation), begining their run in the bottom part of the table, only to pull an impressive recovery in rounds 4-6, going from 7th to 3rd place and securing their place in the playoffs.
The Players
But how well has each player performed? Who is EG-TL's best player? Who is the worst? Let's look at their season performance:
![[image loading]](http://i.imgur.com/FJm1R8W.png?1)
Cumulative map wins for each player that played for EG-TL. Foreigners marked with black borders.
As we can see, HerO and Jaedong have been the team's most performing players, with JYP in a remarkable third place. These three players have won 56% of all maps won by EG-TL. TaeJa and Oz are tied for 4th place at 9 maps, TaeJa having gone through wrist problems, and Oz having amassed an impressive amount of wins for having joined only mid season. The foreigner award goes to Stephano, not only for being tied in the 6th place in total maps won after only playing for two rounds, but also for begin the most efficient EG-TL's player, with a 55% win ratio compared to HerO's 51%, TaeJa's 50% or JYP's 48%.
On the other hand, we can clearly see why PuMa was booted from EG, winning only 1 map out of 8 that he played. Proleague has also proven to be very inaccessible to the non-Stephano foreigners, with ThorZaIN being the only one able to grab 1 win out of 5 maps, and HuK and Snute grabbing none so far. Also, if I was Zenio, I would be pretty scared.
Evil Geniuses vs Team Liquid
But who performed better, EG or TL??
![[image loading]](http://i.imgur.com/S9Xxkei.png?1)
Cumulative number of map wins for the EG roster (Jaedong, JYP, Oz, Revival, Stephano, aLive, ThorZaIN, PuMa, HuK) vs the TL roster (HerO, TaeJa, Zenio, Snute). And Flash. For scale.
Counting the amount of maps won, we see that the EG players have won almost twice as many maps as the TL players. On the other hand, the full EG-TL roster consists of 9 Evil Geniuses players and only 4 Team Liquid players, so that accounts for the difference. In terms of win rate, the EG players won 41% of their maps, while the TL players won 45%.
Overall, it seems that every team contributed to the Unholy Alliance in their own way; you could say EG providing mostly the quantity and TL providing mostly the quality, but in general their performances are very comparable.
Honorable mention to KT.Flash, who by himself won more maps than all TL players combined. If Flash would join EG-TL, the team would hold the 1st spot in the league without any doubt.
Race performances
Finally, I grouped the players by their race of choice, to see if we could spot some differences.
![[image loading]](http://i.imgur.com/TjjchwU.png?1)
Cumulative number of map wins for EG-TL Protoss players (HerO, JYP, Oz, HuK), Zerg players (Jaedong, Revival, Stephano, Snute) and Terran players (TaeJa, aLive, ThorZaIN, PuMa).
The team has been carried by the Protoss (46% of all maps won) and Zerg (37%) players, with the lack of Terran victories (17%) becoming pretty obvious in the graph. TaeJa having to face wrist issues throughout the season, and aLive joining the team half way (plus his not-so-stellar record) account for the differences. And it is not a problem of Terran balance with the Proleague maps; 2 of top 3 performing players of the whole league are the Flash and INnoVation, and the cumulative win ratios for the three races within EG-TL are not very far off: 37% for Terran, 40% for Zerg, and 46% for Protoss.
Overall though, it looks like EG-TL needs the addition of some other good Terran players to strengthen their roster, or they risk becoming a 2-race team.
Conclusion
EG-TL had an up and down wild ride this Proleague season. They won 6/13 and 7/10 matches in their good weeks, showing that they are able to compete in the league, but lost 7/8 and 4/6 matches in their two slumps. Their slumps can not be attributed to any single player, instead being a consequence of the whole EG-TL not performing. Luckily for us, the managers identified this and added coach Park, who brought the team together and broke their long slump. Separately, EG and TL's players performances are roughly equal, so no team can be said to be carrying EG-TL on their own. Finally, it's clear that EG-TL is in need of some Terran additions, to round the roster and make the team more competitive overall.
Let's hope the Unholy Alliance continues to play next season. If they solve their few problems, I am confident EG-TL can be a serious contender for the playoffs.
Finally, shoutout to Liquipedia and all its contributors, for providing all the data I used in this analysis. You guys rock! And thanks to TL for being such an awesome community!
Edit: Fuck it's late.. Any problems/typos you can spot, I won't be able to fix them until tomorrow.
Edit2: To all people asking why didn't I use win ratios in the graphs: I thought briefly about it, but win ratios only have meaning after a good number of matches, once they have converged to show the player's skill value. And after that, they would remain mostly flat since even if there is a skill change in the player, the calculation would still take into account all previous matches, dampening the change. So half the graph would be meaningless noise while the other half would be boring flatness, so overall the graphs would not be very informative. Cumulative wins on the other hand show progression, and also show losses (when there is a point in the graph but the number does not increase). It's not the perfect measure, but it's the better option of the two imho. If you are interested in each player final win ratio, you can check Liquipedia.
Also, thanks to TheMango for the TL+! Very much appreciated!
