Share

Share

Articles:Snowballing in the Group Stages of the S3 World Championship

From Leaguepedia | League of Legends Wiki
Jump to: navigation, search
Snowballing in the Group Stages of the S3 World Championship


LeadWinRateGroupStages-alt.png

I wanted to a follow up to the popular snowballing in competitive play article. So we are going to look at the first half of group stage so far and how these international teams have clashed on the snowballing side.

Overall, in the group stage so far there has been a distinct lack of comebacks. Gold lead at 25 minutes and first inhibitor both have 100% win rates, as no team has ever come back from such a disadvantage yet. The two biggest comebacks we’ve seen have been Vulcun vs SSO and SKT T1 vs Lemondogs, which have both only been mid-game comebacks. Even gold lead at 15 minutes is extremely hard to come back from for teams, having 90% of teams who have even a gold lead of 100g go on to win. Despite the very close and intense matches (especially of group B), these numbers are very telling so far in the group stages.

HorizontalObjectiveWinRate-alt.png


The next thing I wanted to cover is what happens when a team gets not just one first objective, but 2 or more. I thought it was interesting to see that after claiming 2 first objectives, no matter which they were, boosted the win rate to about 80%! I also wanted to present how often a team gets 2 objectives, so that is represented by the “17 games” inside the bars. Since this is out of 20 total games you can see how often these situations arise (85% of games the team who gets FB also gets first tower).

GoldWinRateGroupStages-alt.png


The third snowballing effect I wanted to look at was gold lead at various times in game. This lead can be anything from 100g to 10000g (ties were excluded). But it still presents an interesting pattern and interesting individual data. We have never seen a comeback from a gold lead at 25 minutes and even 15 minutes in 90% of the games go to the team with the gold lead. The two exceptions to this 15 min advantage was the comeback of SSO against Vulcun on day one, and the comeback of SKT T1 against Lemondogs. Some other possibly interesting data is that 80% of the teams who have the gold lead at 8 min go on to keep it at 10 min. 90% of teams who have it at 10 minutes go on to keep it at 15 min, and 90% of teams who have it at 15 minutes go on to keep it by 25 minutes.

ObjectiveWinRateMinGG-alt.png


The last chart has to do with the difference between this data when including Mineski and GG.eu games. These teams have been unfortunately mismatched in the WC. I wanted to present the snowball data with and without these teams as their games can somewhat skew the data. Without these teams you can see that the matches are a bit closer fought, but there is still an prominent advantage for teams who get things like first dragon or first blood. Without these 2 teams the sample size went down from 20 games to 12 games though, which is why I wanted to keep them in for the original graph at the top.

GG.eu MNK TSM LD SSO VUL SKT FNC GMB OMG
First Blood 0% 25% 25% 75% 50% 75% 50% 50% 50% 100%
First Tower 25% 0% 75% 50% 100% 75% 25% 25% 50% 75%
First Dragon 0% 0% 50% 75% 25% 75% 25% 75% 75% 100%
8 Minute Gold Lead 0% 0% 25% 100% 25% 75% 25% 50% 75% 75%

The last thing I wanted to cover on this article was the individual teams and the % of games they got the first dragon, first blood, first tower, or gold lead at 8 minutes. I just thought some people might be curious about it so I included this. I don’t think the data is very telling though, as most teams have only played 4 games, but we can see how it develops in the future, and it was interesting that Mineski / GG.eu have such low percentages.



Written by Spellsy - @SpellsyLoL
Graphics provided by Eric “Vesca Violette” Womack - @VescaViolette
Gold Stats by @3leven



Anonymous user #1

364 days ago
Score 0+-
looks like SKTT1 is nearly the worst team by statistics xD can't be !

Anonymous user #2

364 days ago
Score 0+-
didnt matter have faker

Anonymous user #3

364 days ago
Score 0+-
worst team in early games .... maybe

Anonymous user #4

364 days ago
Score 0+-
i think its because everyone knows for a fact that they will win if nothing is done by early mid game. so they throw themselves at skt early game to try and get a lead

Anonymous user #5

363 days ago
Score 0+-
I think you're seeing those numbers because they've had a lot of comeback wins where they didn't have first tower or first blood... but still won anyways

Anonymous user #6

358 days ago
Score 0+-
They do have a weak early game recognized by many analysts and teams.

Anonymous user #7

364 days ago
Score 0+-
can we have a region based graph which is showing how the specific regions are doing stats wise etc?

Spellsy

364 days ago
Score 0+-
there are too few games for this currently. i will try to do one after group stages or maybe just after the whole WC (latter more likely)

Anonymous user #8

364 days ago
Score 0+-
I thought it said religion lel

Anonymous user #9

364 days ago
Score 0+-
skt is good at counter snowballing they give first blood intentionally

Anonymous user #10

364 days ago
Score 0+-
haha it's bad to take the first tower

Anonymous user #11

364 days ago
Score 0+-

This is obviously flawed methodology. First, you see correlation and claim causality. It is different to say that having first dragon "causes" teams to win as opposed to saying that better teams usually take objectives first, but it is the teams being better (individual skills, map awareness, composition, teamfights etc.) not the gold/exp from first objective that makes them win.

If teams are equally matchetd, this totally brakes. For instance Lemondogs vs TSM - lemondogs got first blood at level 1, and they also got first dragon. Since these things have such a huge impact in your stats this should basically mean lost match and no chance for TSM. And we all know how it went.

Anonymous user #12

364 days ago
Score 1+-
Pretty sure there's no claim of causality here. Simply tendencies (aka correlations). Further, the percent chance of winning after securing both first blood and first dragon is 82%, which explicitly states that this did NOT win the game 18% of the time.

Anonymous user #13

364 days ago
Score 0+-
It s probably because Faker insists that the enemy team get a handicap before he destroys them.

Theelkspeaks

364 days ago
Score 0+-
Well, the above comments are a bit harsh on SKT, a more useful reading of the data would be to say that SKT wins because of their teamfighting, whereas the rest of the teams win by securing early objectives.

Spellsy

364 days ago
Score 1+-
i support this analysis :D

Anonymous user #14

364 days ago
Score 0+-
skt just wins at midgame and this is mostly not before 8mins

Spellsy

364 days ago
Score 0+-
there are too few games for this currently. i will try to do one after group stages or maybe just after the whole WC (latter more likely)

Meristematic

364 days ago
Score 0+-
Really interesting stats again, will be cool to see what happens with them

Gandalf45435

363 days ago
Score 0+-
None of the teams are "bad" They are at worlds...they are world class. Whether they go never lose a game or just the opposite they are the BEST from their region.

Anonymous user #15

359 days ago
Score 0+-

Well, GamingGear are good players of course. Their team has been in Challenger tier in Europe and they absolutely can play this game. However, when going to Worlds you have to be compared with the top teams of the world and doing this it can't be denied that GamingGear did benefit from the system. I do not mean this offensive, I am sure they are a good team, but being unbiased I have to say that they did not really deserve to go to Worlds. You said "they are the BEST from their region", but they aren't. The thing is that Riot has caused a problem by establishing this CIS Zone, which overlaps with Europe. GamingGear also tried to qualify for LCS in Europe, but they did not even manage to qualify for the LCS Qualifiers, neither in Spring nor in Summer nor for the Challenger tournaments at TLP and Gamescom, although they tried for every single one of these tournament. The only thing they could qualify for was Dreamhack, where they lost all of their games. Taking this into account I would say that GamingGear is not in the Top 20 of Europe.

Being not good enough to qualify in Europe (again, I don't want to offend these guys, I just have to compare them with the best teams of Europe) they tried to qualify at CIS Championships in Moscow. They won this tournament, but it has to be said that in comparison to European Challenger tournaments the participating teams were really weak. Warhunter from Instruments of Surrender was the only player besides the GamingGear-players I did actually know. And to be honest: Neither in the CIS they are "the BEST from their region". The best team from CIS of course is Gambit, who did not participate at this tournament, because they already were qualified for LCS Summer. So GamingGear had a quite easy way to go to Gamescom.

To be fair, their victory at Gamescom was kind of remarkable. At least they managed to defeat teams like Pain and Dark Passage, wich really surprised me. Nevertheless this qualification still was a very easy one in comparison to the other regions.

And of course I do not want to grudge them their participation at Worlds for personal reasons. I just find this system treats some other teams quite injustly. The idea behind this Wildcard for Worlds was of course that regions like South America and Oceania, who could not qualify for Worlds in the 5 established main regions and I approve that also teams from South America and Oceania should be given a chance to compete at Worlds (also Africa btw, but I am not sure how many people play LoL there), but the problem is, GamingGear is not from one of those regions. They are form Lithuania, from Europe and they did benefit from a flaw in the system and therefore they went to Worlds. They just got an unfair advantage because of their origin in comparison to other European competitors. The new Copenhagen Wolves for example did prove several times that they are a lot better than GamingGear, but still GamingGear was favoured due to their origin. And I am not even speaking of teams like Evil Geniuses, Ninjas in Pyjamas, Alternate, SK Gaming. And there would come over 10 further European challenger teams to my mind, which would have deserved this place at Worlds more than GamingGear did.

Riot really needs to clean up this confusion regarding to the different regions. Regions MUST not overlap, otherwise the system is destroyed. Imagine, what if Gambit lost their battle for the 3rd rank at Gamescom against EG. Then Gambit would not have been allowed to go to Worlds, GamingGear instead would have been the "best" team from CIS. If I were Gambit, I would leave LCS Europe to go through the CIS Championship and then through the Wildcard tournament, which would be a lot easier and safer. You see, how broken this system is?

Anonymous user #16

358 days ago
Score 0+-

If Gambit went through the CIS championship, they would not have the practice against the strongest teams in europe meaning they would likely win wildcard, but get stomped due to lack of practice against less than the highest calibur teams.

What I think is very interesting about the LCS is the the lack of international practice and competitions forces strong regiong to stay strong, and weak reagions to remain weak (Like c9 looking incredibly strong, almost unstoppable, all lcs, yet losing a series to fnatic quite decisively). It definitely seems like the NA LCS will remain a joke until we get better teams to practice against.

As a side note, I think single elimination Bo3 for the World champ quarterfinals is complete bullshit. All the asian teams qualified through double elimination or at least Bo5. 3 games is almost a tossup, and even 5 games can be very onesided depending upon how warmed up the players are. Teams coming in cold (3/4 lost their quarterfinal matches) has had a huge effect and the hype leading up to the worlds ends so quickly (speaking about c9). Making the tournament a little bit longer and allowing double elim. would allow for the teams that got an unlucky draw in the quarterfinals (OMG, arguably, Gama Bears and any team really) to prove that they might have just had an off day. I have seen so many tournaments where a team drops out early into the losers bracket only to come back and win the entire thing(Gama bears had a similar road to worlds).

Anonymous user #15

357 days ago
Score 0+-
Yeah, I'd also prefer Double Eliminiation.

ShadyEX

357 days ago
Score 0+-
Useful information. I was shocked about statistics about double objectives...

SugiStlye

356 days ago
Score 0+-
Nice statistics keep this up :)

Anonymous user #17

150 days ago
Score 0+-
This is why league is shit.
Add your comment
Leaguepedia | League of Legends Wiki welcomes all comments. If you do not want to be anonymous, register or log in. It is free.


Click Here to return to the Articles Index.