Blue side vs Purple: Inside the startling numbers

I recently interviewed CLG.EU's support player Krepo, who brought up the fact that he believes blue side is a decided advantage bottom lane and in picks/bans. Blue side gets first pick and ban, as well as what most pros believe to be the superior side of the map to control. Many people wondered aloud after reading Krepo's beliefs what the actual breakdown was for blue vs purple. I took a sample of 145 competitive games played on LAN at MLG Summer Arena, IGN Face-Off, European Regionals, MLG Raleigh, and the NA Regionals to find out what the discrepancy is between the two sides in terms of overall win percentage and win totals in extended series (more than 2 games played). The numbers may be even more surprising than you'd think, and Riot might want to think about separating first pick and blue side for the Season 2 World Championships.

http://bearsgrowl95.files.wordpress.com/2012/05/summoners-rift1.jpg

The advantages of being blue

The following stats and statements must be made knowing that a majority of the games played had the team with the higher seed starting off as blue side. This means that the teams starting blue side were "supposed" to win according to rankings, but the numbers may be a little lopsided even for that.

Krepo: Bot lane has double golems and can help jungle start red buff with 2 people allowing him to transition him well into a quick jungle with early red (minimal difference with blue, insignificant ) but the EXP on lane decides everything. If purple has to help jungle on blue they will lose exp and be forced to either shove lane to equal it or be forced to level slower and be poked off or suffer hard from exp imbalance. In some bot lane combo's, falling behind in exp by being not early on lane/not pushing can lose you the lane entirely. If purple freezes you can always go to your golems to "fake base" and force them to push. In essence: blue bot lane has way more control.

If purple pushes they're WAY easier river ganked than blue. If blue pushes then purple jungle has to commit to entering "enemy jungle" which holds a severely higher risk.

Mid-late baron is easier accessible but that's open for interpretation. Map imbalance resides mostly in bot lane laning phase and exp imbalance. It's also slightly easier to steal wraiths over wall if you're blue side AP MID.

That was Krepo's response to why blue side had an advantage, particularly in terms of bottom lane. Arguments could also be made that top lane is safer for blue side because of no jungle path to enter from behind. Looking back at the map, it's much more likely that a support will have the tribush in bottom lane warded than purple side top will meaning another lane that is potentially "safer". Now let's take a look at those numbers for all of the pro games played in August.

  • In 145 games played the blue team won 86 of them or 59.3%
  • That leaves purple side winning just 59 of 145 possible games for 40.7%
  • 22 series were extended past 2 games (best of 3's or 5's)
  • Of those 22 series purple won the majority of the games only 6 times or 27%
  • In fact, the team that was blue won all 3 games of the series an astounding 10 times.
  • In series where each team won at least 1 game, the deciding game was won by the Blue side 13 out of 17 times or 76.47%

So what do these numbers tell us? Teams are MUCH more likely to win from the blue side particularly in decisive games. The sample size isn't large enough to draw absolute conclusions yet, but it is certainly something to watch for as we move into the World Championships where every team will be evenly matched. Which side a team gets, especially in the group stage could determine who walks away with first place all that prize money. The group stage format is a round robin 1 game playoff where the top two teams from each group advance. It is of my opinion that Riot, at the very least for tournaments, should implement a way for teams to win a "coin toss" where they elect to either have first pick, or which side they want much like in American Football. What are your thoughts RoG'ers, are these stats enough to warrant a change, and if so what should be done about it?


Follow me on twitter for more news and updates!

 

59

Comments

  • #59 currycaptain

    The comments on this are horrendous...most people concluded things without fully reading the article.....

     

    For starters...ANY statistical analysis is invalid as the picking of sides was skewed to the higher seed, according TO THE POST (read people).

    Second, if we choose to do a test (which only one person did) we don't use Chi^2 (i'm not gonna explain why, just review your statistics), but rather a one-proportion z-test. The p-value is about.024 for a two-tailed test, which would be significant, if the condition of randomness was met.

     

    And to people saying this doesn't meet a large enough sample size.....you simply need to have 10 failures (blue side loses) and 10 successes (purple side loses) , or vice versa. This test only doesn't meet a randomness condition, which is very serious, and data using games that were decided on a coin toss WOULD be valid and  I would love to see some data on that

  • #58 FatLady2k

    how to fix blue/purple side advantage!

    implement flipped camera

    let dragon spawn randomly in either baron or dragon pit and baron spawns in an random pit or or if occupieded by dragon in the empty pit

    then the only difference is 1. pick vs last pick

    Last edited by FatLady2k: 4/26/2013 8:35:31 AM
  • #57 w00p00

    separate blue side from first pick :)

  • #56 nightcracker

    I'd like to note that I've actually done statistical analysis on random in-game samples as well. See this thread: http://euw.leagueoflegends.com/board/showthread.php?t=698540

    Sadly I can't continue my current research because the tool I used to programmatically get info from riot's servers (riotcontrol) is broken.

    Or I could crawl lolking.net.

  • #53 Kikutar

    the advantage for blue side is very small...what makes it such an issue is the current meta, imo ...the meta we have ( or after the last tournaments (2vs1, ad/sup in mid)the meta we had so far)

  • #45 Gatward

    Shouldnt it be the teams responsibility to work around purple teams difficult nature.

  • #46 Gameguy301

    players can only adjust so much with the tools and strategies they are given, your assumeing a team comp aimed at mitigating blue vs purple imbalances will still be functional as a whole.

  • #42 toastal

    It's funny because Dominion never had this problem.

  • #43 pagan4life

    what a lame attempt to high-jack the nature of this post. die.

    Warning: Posts like these will not be tolerated
  • #41 fedao321

    I remember seeing a thread on the official forums in wich a red stated that there was this disparity between blue and purple sides. He said it didn't occur only on high ELO, but throughout all ELO ranges with greater or lesser impact.

    At the moment, riot was not worried with it, as matchmade games had matchmaking usually placing the stronger team on purple side in order to achieve 50% wins for each side, but as they become more and more concerned with LoL as an e-sport, they'll probably change the map very soon.

  • #49 guy420

    free... MAGMA CHAMBERS?

  • #50 fedao321

    Magma chamber not happening. It became crystal scar.

  • #51 Skip3rd

    never stop believing, one day magma chamber WILL be released!

  • #54 EnaiSiaion

    Quote from fedao321 »

    I remember seeing a thread on the official forums in wich a red stated that there was this disparity between blue and purple sides. He said it didn't occur only on high ELO, but throughout all ELO ranges with greater or lesser impact.

    At the moment, riot was not worried with it, as matchmade games had matchmaking usually placing the stronger team on purple side in order to achieve 50% wins for each side, but as they become more and more concerned with LoL as an e-sport, they'll probably change the map very soon.

    They would have to remake the whole map and make it symmetrical instead of one side being turned upside down for no good reason. This would take effort and yield no extra money.

  • #40 xCry0x

    Just change the meta to be adc/support mid and solo tops & bots then it is all pretty much friging the same aside from camera angle.

    If mid gets exp fastest then having the duo lane mid would help the duo lane keep up with xp =)  Also 1 person can leave lane to gank while other person freezes lane at turret.

  • #39 Misticaltom

    What, we don't get Purple side counter arguments?    But the Counter Argument is crutial in the overall debate of Blue VS Purple!

  • #44 Fixitorit

    I did it :)

  • #38 BenBenBenBe

    It should be stated that blue side is the higher seeded team 2/3 of the time; that is to say, TSM will generally beat LgN regardless of their side, but since they're blue side in a best of 3 that goes to 3 rounds, that'll be 2 wins for blue side.

  • #37 thealdo

    Since this is a digital game and there is no specific home field, isn't this just the equivalent of home field advantage?

    Away teams in traditional sports have to deal with the atmosphere and all the other intangibles that go with playing a team at their home stadium. I feel that is the approach with blue side. You are the home team, you get first pick, advantages for bottom lane, and safer top lane and the purple side have to overcome it. Just my two cents.

  • #47 SpacianEU

    The difference is that (in most cases) the teams have first leg in one stadium and second leg in the other stadium, or, if it's only a series of 1 game (like world cup), you're playing in a totally different stadium. So overall, there is no advantage or disadvantage.

    That's different in League, because, let's say you get blue side first game (win), you gonna have purple side second game (lose) and blue side in the third game again (win). In this way, you get an advantage, which is not supposed to be there.

  • To post a comment, please or register a new account.
Posts Quoted:
Reply
Clear All Quotes