20180830 Predictions for NA LCS Semifinals

This is some predictions for various “first” markets (blood/dragon/turret/baron) for the upcoming NA LCS semifinals on the 1st and 2nd of September, 2018. This article was published on the 30th of August, 2018.

Cloud9 vs TSM

Cloud9 vs TSM looks to be an exciting game. Let’s take a look at some data, and what markets might be promising. First, the team’s overall performance:

TSM have an extra 6 games on C9 - a tiebreaker to progress to the playoffs, and their five game quarterfinal match against Echo Fox. At the end of the regular season, C9 had one game up on TSM. C9 has won their last 8 games straight.

C9 and TSM’s win rates over the entire split are 56% and 61% respectively - a weighted average of 53% to 47%, too close to call. However, for the last eight games, the numbers tell a different story: 68% to 31%. Cloud9 is hot stuff. The odds are 1.44 : 2.62 in favor of C9, which reflects this. Both are around 0.8 to 0.9 EV, so I wouldn’t put money on either.

How about first bloods?

Both have achieved a high % of first bloods in recent games. They also follow similar trends through the season. The odds here are 1.72 : 2, for C9 : TSM. Sampling the entire split (displayed as -1 games), last 8 and 4 games gives us:

=== cloud9 vs team solomid - first_blood last -1 games ===
Relative %   53.78 : 46.22
cloud9 : team solomid 1.72 : 2
cloud9  ev: 0.92
team solomid    ev: 0.92

=== cloud9 vs team solomid - first_blood last 8 games ===
Relative %   68.75 : 31.25
cloud9 : team solomid 1.72 : 2
cloud9  ev: 1.18    <------------------- High EV
team solomid    ev: 0.62

=== cloud9 vs team solomid - first_blood last 4 games ===
Relative %   62.5 : 37.5
cloud9 : team solomid 1.72 : 2
cloud9  ev: 1.08
team solomid    ev: 0.75

C9 might be a candidate for an FB bet - the latter part of the split, combined with the odds, gives solid EV.

First dragon isn’t anything right home about, even both ways throughout the season. First turret is a bit more interesting:

C9 have maintained their first turret percentage at well over 50% throughout the entire season! TSM, on the other hand, has been hanging around the low 40s. FT is more often than not in the bot lane, since ADCs are basically chosen for their turret taking abilities, plus they have a second player helping them. This reflects both C9’s stellar bottom lane, especially Sneaky, the definition of consistency, as well as TSM’s underperforming European duo.

=== cloud9 vs team solomid - first_turret last -1 games ===
Relative %   65.33 : 34.67
cloud9 : team solomid 1.72 : 2.0
cloud9  ev: 1.12    <------------------- High EV
team solomid    ev: 0.69

=== cloud9 vs team solomid - first_turret last 8 games ===
Relative %   75.0 : 25.0
cloud9 : team solomid 1.72 : 2.0
cloud9  ev: 1.29    <------------------- High EV
team solomid    ev: 0.5

=== cloud9 vs team solomid - first_turret last 4 games ===
Relative %   75.0 : 25.0
cloud9 : team solomid 1.72 : 2.0
cloud9  ev: 1.29    <------------------- High EV
team solomid    ev: 0.5

A snapshot of the odds, and calculated EV at any point during the split show C9 as a strong case for a FT bet. Let’s look at first baron next.

TSM has a ever so slightly positive EV over the latter half of the split regarding first baron, but given the strong correlation between first baron and winning the game, which TSM is not favored to do, I’m not sure it’s a good bet. FB and FT have a very minimal correlation to the actual result of the game (~0.1), so I’m happy to bet on either regardless of what I expect to be the final result.

Since first baron has a high correlation with victory in NA (around 0.65 for the top teams), I would like to weigh first baron more heavily in the future. Things I would like to include in the weighting are the matchup and recent game history.

Team Liquid vs 100 Thieves

What looks like two of the top teams this split meet in the semifinals. Their overall wins put them at a 49% : 51% weighted average win percentage, for 100 and TL respectively. That changes to 43% : 57% considering the last 8 games, still a close contest, but factors in the recent win stream for TL.

Looking at the running victories for both teams:

TL is on fire, with 5 straight wins. 100 had a similar streak earlier, although I haven’t checked which team it was against.

Let’s take a look at first blood:

100 fell of pretty heavily in their last few games, but other than that, maintained a healthy 60-70% throughout the split. TL, on the other hand, hovered around 50%, and shot up recently. With the odds at 1.72 : 2 for LT and 100 respectively, 100 maintains a positive EV at all points of the split.

=== team liquid vs 100 thieves - first_blood last -1 games ===
Relative %   47.34 : 52.66
team liquid : 100 thieves 1.72 : 2
team liquid     ev: 0.81
100 thieves     ev: 1.05

=== team liquid vs 100 thieves - first_blood last 8 games ===
Relative %   31.25 : 68.75
team liquid : 100 thieves 1.72 : 2
team liquid     ev: 0.54
100 thieves     ev: 1.38    <------------------- High EV

=== team liquid vs 100 thieves - first_blood last 4 games ===
Relative %   37.5 : 62.5
team liquid : 100 thieves 1.72 : 2
team liquid     ev: 0.64
100 thieves     ev: 1.25    <------------------- High EV

Their recent downtrend, along with TL’s recent uptrend makes me a little less confident, but since FB is impacted by a long of factors, like champion selection, I’d say 100 is still a solid choice. My personal win rate with FB is around 45%, so as long as the EV is sufficiently positive, it’s a long term winner.

First turret tells a very different story:

Until 3/4 of the way through the split, the teams were fairly close. TL, however, has pulled away to become to FT kings, and again we observe a downward trend for 100. The lack of consistency is shown in the EV:

=== team liquid vs 100 thieves - first_turret last -1 games ===
Relative %   48.31 : 51.69
team liquid : 100 thieves 1.66 : 2.1
team liquid     ev: 0.8
100 thieves     ev: 1.09

=== team liquid vs 100 thieves - first_turret last 8 games ===
Relative %   50.0 : 50.0
team liquid : 100 thieves 1.66 : 2.1
team liquid     ev: 0.83
100 thieves     ev: 1.05

=== team liquid vs 100 thieves - first_turret last 4 games ===
Relative %   50.0 : 50.0
team liquid : 100 thieves 1.66 : 2.1
team liquid     ev: 0.83
100 thieves     ev: 1.05

Positive EV on 10. I prefer to bet on a minimum EV of 1.1, so this is really on the fence for me. I’ll think about it a little more.

How about first dragon?

TL are the first dragon KINGS. Mainly because of Xmithie, arguably the best jungler in the league, sitting at a cool 70% for most of the split. 100 have a very low percentage - their botlane is solid, especially mid and late game, but they certainly don’t play for the first 10 minutes. TL, on the other hand, basically have Doublelift carry every game. This dominant bottom lane, in combination with Xmithie’s dominance in the jungle, leads to a ludicrously high FD percentage.

The EV also reflects this:

=== team liquid vs 100 thieves - first_dragon last -1 games ===
Relative %   64.37 : 35.63
team liquid : 100 thieves 1.72 : 2.0
team liquid     ev: 1.11    <------------------- High EV
100 thieves     ev: 0.71

=== team liquid vs 100 thieves - first_dragon last 8 games ===
Relative %   75.0 : 25.0
team liquid : 100 thieves 1.72 : 2.0
team liquid     ev: 1.29    <------------------- High EV
100 thieves     ev: 0.5

=== team liquid vs 100 thieves - first_dragon last 4 games ===
Relative %   87.5 : 12.5
team liquid : 100 thieves 1.72 : 2.0
team liquid     ev: 1.5     <------------------- High EV
100 thieves     ev: 0.25

Looks like a good bet.

First baron is interesting. 100’s ongoing downtrend is revealed here, too:

Until recently, it’s about 50% for each team. How about some EVs?

=== team liquid vs 100 thieves - first_baron last -1 games ===
Relative %   48.91 : 51.09
team liquid : 100 thieves 1.61 : 2.2
team liquid     ev: 0.79
100 thieves     ev: 1.12    <------------------- High EV

=== team liquid vs 100 thieves - first_baron last 8 games ===
Relative %   62.5 : 37.5
team liquid : 100 thieves 1.61 : 2.2
team liquid     ev: 1.01
100 thieves     ev: 0.82

=== team liquid vs 100 thieves - first_baron last 4 games ===
Relative %   50.0 : 50.0
team liquid : 100 thieves 1.61 : 2.2
team liquid     ev: 0.8
100 thieves     ev: 1.1

Pretty inconsistent, no clear winner here.

Conclusion

Cloud9 has a strong history of taking first turret all split, compared to TSM and their struggling bottom lane. FT is a good bet for C9 this weekend. 100 Thieves also look to be a solid choice for first blood against TL. TL’s strong bottom lane and excellent jungle presence puts them as the main contender for first dragon. C9 and TL as the favourites for the respective games, with the postive EV are particularly good picks.