2018 NALCS Changes, Real Data, BO1's, and Competitive Integrety

So this is less a "discussion" post and more a place I can post some real data and then extrapolate. WARNING: This post will be long. I decided to see what BO1's would have done to the 2017 NALCS Competitive Scene. I started by going back through every match in the Spring and Summer Split of NALCS, all 10 teams, and recording whoever won the FIRST game, regardless of actual outcome...and there are several things that stick out. First, the premise of my points: More Data = Better Data. Playing and winning 50% of your games is more telling if that's 50 out of 100 than if it's 5 out of 10 which is better than 1 out of 2. As such, Spring and Summer Split standings (with Bo3's in play) should be a BETTER and MORE ACCURATE representation of how the teams ACTUALLY played League of Legends. Again, because of this that means that any wildly off data between Bo1's and Bo3's constitutes a POTENTIAL problem in Competitive Integrity. Now, on to the data: The format will be "Team", "Real Ranking", and "Bo1's Record", and "Bo1's Ranking" Spring Split 2017: ------------------------------ TSM - Real: 1st Place - Bo1: 9-9 - Bo1's Rank: 3rd-5th (Tied with IMT and DIG) C9 - Real: 2nd Place - Bo1: 13-5 - Bo1's Rank: 1st-2nd (Tied with P1) CLG - Real: 4th Place - Bo1: 7-11 - Bo1's Rank: 9th - RELEGATED FOX - Real: 8th Place - Bo1: 8-10 - Bo1's Rank: 6th-7th (Tied with FLY) P1 - Real: 3rd Place - Bo1: 13-5 - Bo1's Rank: 1st-2nd (Tied with C9) TL - Real: 9th Place - Bo1: 7-11 - Bo1's Rank: 8th IMT - Real: 7th Place - Bo1: 9-9 - Bo1's Rank: 3rd-5th (Tied with TSM and DIG) DIG - Real: 6th Place - Bo1: 9-9 - Bo1's Rank: 3rd-5th (Tied with TSM and IMT) NV - Real: 10th Place - Bo1: 6-12 - Bo1's Rank: 10th - RELEGATED FLY - Real: 5th Place - Bo1: 8-10 - Bo1's Rank: 6th-7th (Tied with FOX) Observations: * Bo1's may not provide enough games to enable enough separation between the teams without playing a BUNCH of playoff games and without adding additional separation criteria. 1st and 2nd were tied (13-5 and 1-1 head to head), 3-5 were Tied (9-9 and all teams were 1-1 against each other), 6th-7th were tied (8-10 and 1-1 head to head). That means AT LEAST 4 playoff games...and how do you play a playoff game between 3 teams? * The VAST MAJORITY of team match record week to week was 1-1. * Bo1 2-0 weeks were NOT dominated by the strongest teams week to week. For instance week 9's 2-0's were C9 and TL, week 8 was CLG, P1 and DIG. TSM (Winner of the split) had only 2 weeks with 2-0 wins (Week 4 and Week 5) while C9 (second place) had 6 of 9 weeks where they went 2-0 against their opponents. * One total team out of 10 ended up where their Bo3's rank actually put them. The rest of the teams shifted up or down... * CLG Slipped from 4th (in Playoffs) to 9th (In Relegation Tournament) in switching formats from Bo3 to Bo1. * IMT Jumped from 7th to all the way potentially up to 3rd. * TSM Slips from 1st Place to the potential of 5th. * Finally, lets face it, it doesn't matter what format LoL is played in, NV still sucked. Concerns: * Consistent 2-0 game play was not standard for teams that proved themselves to be top tier teams. Of the top 4 teams, 2 of them had only two 2-0 weeks. This gives credence to the conversation that there might be a Competitive Integrity problem...the teams that actually proved themselves to be the better teams in spring split did considerably worse in a Bo1 format. * The biggest "Competitive Integrity" problem I see with the above data right now is the outlyer of the 4th place team (CLG) dropping INTO THE RELEGATION TOURNAMENT in the Bo1 Format. A team should NOT be able to slip from 4th to 9th by changing the number of matches played. Because, scientifically, MORE data points = better data, that means this outlyer is a STRONG indication that Bo1's may not provide enough matches to determine a team's actual strength of play. * And it's not just CLG. TSM was down between 2 and 4 slots, IMT was up by 2 - 4 slots...those are LARGE Swings in standings/rankings/team placement and tends to call into question whether or not Bo1's actually shows a team's strength. Summer Split 2017: --------------------------------- TSM - Real: 1st Place - Bo1: 14-4 - Bo1's Rank: 2nd C9 - Real: 4th Place - Bo1: 11-7 - Bo1's Rank: 4th Place CLG - Real: 3rd Place - Bo1: 13-5 - Bo1's Rank: 3rd Place FOX - Real: 8th Place - Bo1: 7-11 - Bo1's Rank: 7th Place P1 - Real: 10th Place - Bo1: 3-15 - Bo1's Rank: 10th Place - RELIGATED TL - Real: 9th Place - Bo1: 5-13 - Bo1's Rank: 9th Place - RELIGATED IMT - Real: 2nd Place - Bo1: 10-8 - Bo1's Rank: 5th Place DIG - Real: 5th Place - Bo1: 15-3 - Bo1's Rank: 1st Place NV - Real: 6th Place - Bo1: 5-13 - Bo1's Rank: 8th Place FLY - Real: 7th Place - Bo1: 8-10 - Bo1's Rank: 6th *Observations* * During summer split we had a 100% clean ranking at the end of the season...there were no ties and no tiebreakrs had to be used to rank the teams. * Four of Ten teams placed identically to their Bo3's final ranking. * DIG Potentially would have shifted up 4 slots, IMT would have dropped 3 slots...the rest of the teams (8 of 10) were pretty stable (within 2 slots) between Bo1's and Bo3s. Concerns: * Spring Split had almost everyone tied, Summer Split had almost no one tied. While we don't have enough data points, this feels consistently inconsistent. * Again we have a wild shift between spring and summer split with consistency of the data. Spring Split had 3 teams jump between potentially up to 4 slots away from their actual final placement (IMT up, CLG and TSM Down). Overall -------------------------------- If Spring and Summer split represent the ends of the spectrum with regards to the number of tied teams at the end of the season, it's possible that the general middle ground is acceptable to everyone. Whether or not Bo1's provide enough games in a season to accurately rank one team against the next would require more data points...no clear trend is showing. This is concerning but not alarming. With regards to team placement between Bo1's and Bo3's this is also fairly concerning because we have, again, bracketed the ends of the specturm...8 of 10 teams stayed reasonably close to their Bo3's ranking with 4 teams ending identically to their Bo3's ranking in Summer Split while in Spring Split only 1 team ended Identically (NV at last) and several teams with a 2 slot shift potentially. Only about 3 teams stayed relatively close (within 1 slot) of their Bo3's ranking. Again we need more data points as the two seasons seem to have "book ended" the spectrum. More data points would be needed to see if we generally see a happy medium or if one or the other extreme is more the norm. Conclusion -------------------------------- I think there IS enough data to be concerned about the competitive integrety of a "Bo1's" format. While there is certainly NOT a "clear" competitive integrety problem, it is clear that large data variances between Bo1 and Bo3 formats can (and have) happened. If we were to get too many seasons like Spring Split, "Competitive Integrety" would definately be challenged if not outright broken. Sending the 4th best team to religation and potentially kicking the 3rd best team out of the playoff tournament are not good signs that "the best team" is winning. Understand, while this is a sport, this isn't a sport like basketball or baseball. Normal sports are played over the course of hours not minutes, and normal sports do not GENERALLY have a "point of no hope" within them. League, with it's gold collection, itemization and scaling team comps, is a VERY different game...a couple mistakes 5 minutes into the game and you're pretty much done, which makes this a very different game than other sports. And THAT is where the concern comes in when comparing (and ultimately making decisions on) League vs. Basketball or Football. In "The Dive" podcast talking about the changes it was referenced several times that "normal" sports don't play best of's to determine winners...but what wasn't addressed was that normal sports are very different from league and different games necessitate different methods of determining who the better team is. ALL normal sports play quarters or halfs - football plays basically 4 games against the opponent, basketball plays 2, and baseball plays pretty much 9 head to head games every time a team matches up. In this mentality, playing Bo3's is actually MORE normal than playing Bo1's if comparing to more "normal" sports. Better numbers/comparisons should be gained determing time in match instead of number of games played.

We're testing a new feature that gives the option to view discussion comments in chronological order. Some testers have pointed out situations in which they feel a linear view could be helpful, so we'd like see how you guys make use of it.

Report as:
Offensive Spam Harassment Incorrect Board