LOL Pro League (LPL)

Spring 2021 – Grand Finals

 

 

 #5 FunPlus Phoenix -250 (-1.5 maps @ -116, -2.5 @ +276, +1.5 @ -625, +2.5 @ -5000)

vs

#1 Royal Never Give Up +183 (+1.5 maps @ -115, +2.5 @ -400, -1.5 @ +381, -2.5 @ +1051)

 

 

Map ML Price: FPX -192 / RNG +143

Total Maps Played: 3.5 maps (over -286 / under +205), 4.5 maps (over +179 / under -244)

Kill Total: 27.5 (over -115 / under -118)

Kill Spread: -4.5 @ -116 / +4.5 @ -116

Team Kill Totals: 15.5 / 11.5

Time Total: 31:00 (over -132 / under -103)

(Individual player model positional ranking is a “living” number comparing individual performance against other players at their position) (includes regular season)

FunPlus Tale of the Tape Royal Never GU
1966.7 Gold Diff @ 10 min -148.4
764.5 Gold Diff @ 15 min -259.8
185.6 Gold Diff @ 20 min -1588.3
112.6 GPM first 20 min vs Avg 2.9
106.5 Gold Diff / min first 20 26.4
659.8 Gold Diff / min Rest of Game 499.7
1936.1 Gold /min (GPM) 1871.5
142.6 Gold / min vs Avg 78.0
187.4 Gold Diff / min 131.6
2.5 Gold Percent Rating (GPR) 1.8
1692.5 Kill Agnostic GPM 1653.7
125.3 Kill Agnostic Gold Diff/min 80.5
2041.4 GPM in wins 1974.4
381.1 Gold Diff per min in wins 331.4
1683.4 GPM in losses 1638.2
-277.5 Gold Diff per min in losses -321.3
204.0 Adjusted Gold Diff / min 148.2
78.0 Win-Adjusted GPM 11.0
53.3 Win-Adjusted Gold Dif/min 3.6
53.0 Dragon Control % 54.3
59.8 Herald Control % 45.9
66.7 Baron Control % 69.9
17.0 Quality Wins? 22.0
47.2 % of wins as Quality 64.7

(These numbers use a composite blend of trending and season long performance)

 

Suggested Model Plays (in order of strength of edge):

none

 

Quantitative Analysis:

Series Outcomes
Wins Losses Probability
3 1 25.000%
3 2 20.695%
3 0 20.133%
2 3 14.614%
1 3 12.467%
0 3 7.091%
(Series Win): 65.827%

(FPX projected series win % via model)

 

 

 * FPX Left / RNG Right

 

  • Gold per minute in wins: 1st / 5th
  • Gold Differential per minute in wins: 4th / 9th
  • Gold per minute in losses: 2nd / 8th
  • Gold Differential per minute in losses: 3rd / 8th
  • Win-adjusted gold per minute: 1st / 5th
  • Win-adjusted gold differential per minute: 4th / 9th
  • Kill-agnostic gold per minute: 1st / 3rd
  • Kill-agnostic gold differential per minute: 2nd / 6th
  • Opponent kill-agnostic gold per minute: 1st / 3rd
  • Overall Objective Control: 3rd / 4th
  • Overall Team Rating: 1st / 4th
  • Regular season series 2-0 RNG, first game was a massive throw by FPX, second was decisive RNG win.
  • Playoff series was 3-0 for FPX on April 7th. No particularly close games.

 

By  the “traditional” numbers FPX are a better team across the board, including in RNG’s strongest asset which is overall objective control. In fact, RNG have performed below their overall season level substantially despite winning a few playoff series here and have actually moved down in the overall power ratings a decent amount with the playoff games included.

 

 

This more or less lines up with the film which I’ll discuss more below in the qualitative evalutaion.  Before we get to that, let’s take a look at how these two fare when you filter out their “easy” matches.

 

Below are the “only vs top nine teams” measures (all playoff teams besides LNG):

 

FunPlus Tale of the Tape Royal Never GU
1832.3 Gold Diff @ 10 min -362.0
447.8 Gold Diff @ 15 min -675.8
185.6 Gold Diff @ 20 min -1588.3
91.1 GPM first 20 min vs Avg -44.1
72.3 Gold Diff / min first 20 -30.8
487.6 Gold Diff / min Rest of Game 438.0
1879.3 Gold /min (GPM) 1827.4
91.7 Gold / min vs Avg 39.9
111.5 Gold Diff / min 49.5
1.5 Gold Percent Rating (GPR) 0.7
1661.5 Kill Agnostic GPM 1618.0
71.7 Kill Agnostic Gold Diff/min 21.3
1992.5 GPM in wins 1950.3
320.9 Gold Diff per min in wins 282.3
1671.7 GPM in losses 1632.9
-272.4 Gold Diff per min in losses -319.3
137.4 Adjusted Gold Diff / min 75.4
33.4 Win-Adjusted GPM -8.8
-0.5 Win-Adjusted Gold Dif/min -39.0
51.6 Dragon Control % 49.7
57.4 Herald Control % 43.5
58.3 Baron Control % 68.0
8.0 Quality Wins? 11.0
36.4 % of wins as Quality 57.9

 

 

(Tale of the Tape both teams filtered vs top nine only)

 

Series Outcomes
Wins Losses Probability
3 1 25.534%
3 0 20.969%
3 2 20.729%
2 3 14.162%
1 3 11.919%
0 3 6.687%
(Series Win): 67.232%

 

 

(FPX series win % filtered to both teams vs top nine)

 

* FPX Left / RNG Right (both vs top nine teams)

 

  • Gold per minute in wins: 3rd / 10th
  • Gold Differential per minute in wins: 8th / 14th
  • Gold per minute in losses: 2nd / 9th
  • Gold Differential per minute in losses: 3rd / 7th
  • Win-adjusted gold per minute: 3rd / 10th
  • Win-adjusted gold differential per minute: 8th / 14th
  • Kill-agnostic gold per minute: 2nd / 8th
  • Kill-agnostic gold differential per minute: 2nd / 8th
  • Opponent kill-agnostic gold per minute: 3rd / 9th
  • Overall Objective Control: 6th / 4th
  • Overall Team Rating: 2nd / 7th

 

Suggested Model Plays (in order of strength of edge) vs top nine teams only:

none

FunPlus’ overall performance level doesn’t change much when they’re facing the elite teams in the LPL. RNG’s do take a noticeable dip suggesting that they perform worse against the good teams while FPX’s performance level isn’t affected as much by opponent quality.

The model doesn’t factor in these normalized and adjusted statistics as much as a composite blend of trending and season long raw numbers vs league level but the TL:DR here is that FPX are a more “legitimately” good team “by the numbers.”

Conclusion: Model makes this right about market. No play. Advanced metrics more heavily favor FPX though.

 

Qualitative Analysis:

The last time these two faced off it was a swift and merciless 3-0 by FPX who exposed RNG’s tendency to tunnel on playing around Xiaohu. They laid the blueprint for future teams to attack that predictability. It’s not that RNG are incapable of doing other things but they did lean quite a bit on that “pitch” so to speak. That said, they showed some adjustments in their last series against EDG showing some more bottom lane centric looks to mix things up. To some extent it worked.

I’m just going to be forthright since you’re here for my opinion. I think if EDG showed up for either of those first two games OR didn’t make the moronic over-committal on the dragon in a game five where they were in control I think that we’re not talking about RNG here. Not to mention the silly backdoor it took in game five against TOP’s ocean soul and fed Draven buffed setup (dumb to not commit at least one recall). At some point you have to give credit to RNG for pulling a rabbit out of a hat, and I do. That kind of creativity, playing to your outs, etc is something that should be admired but it’s way WAY overvalued because people seem to forget that 99% of the time where you lose that exact game because you were in that situation to begin with. RNG are very good but they have not played to their standards during the playoffs.

Like “life” in a truly Jeff Goldblum sense, RNG…. find a way. I’ll bet against these teams every single time because 9 times out of 10 they’re at least a little bit fraudulent. I’m not saying RNG are bad but they are quite lucky to even be here despite being the #1 seed.

I think we’re more likely than not to see something close to what we saw last time than a “competitive” five games series. This is a stylistic mismatch for RNG. FPX’s opening twenty is ridiculously strong and they don’t make nearly as many macro mistakes as the other LPL teams do so RNG’s “hodl and scale” approach isn’t a reliable way to approach things. Of course, I assume that they learned that last time around and they’ll show us some different looks here.

This should be a good final but I think we’re looking at a 3-1 for FPX, likely two close games, two blowouts. The number for this series is right on the money according to my model and the numbers but I’m giving a slight bump to FPX for both the stylistic matchup and, to me, better solo laners. I’d even argue that the FPX bottom lane is playing better right now as well, especially if they take GALA off of Kaisa.

FPX roll.

Conclusion: I like FPX more than the model does.

A quick note on my futures positions:

I’m holding an RNG +3300 to win the split for 1 unit (to net +33). Obviously I’ll be hedging here since I think FPX are the side. For disclosure, I got FPX at -233 on the opener for this hedge anticipating that it’d move in their direction over the course of the week, which is has. For record keeping purposes I’ll be using the numbers in this post for my public record (as always).

I mentioned in one of the matches last week that different people have different thoughts on hedging and futures/outrights in general but I typically will “take profits” in spots like this. Part of why I’m willing to bet into high vig markets like outrights is that I’ll have opportunities like this.

This is a pretty straight up “dutch” or even hedge. The number is 24.45 units. 24.45 at -256 gives us a +8.55 profit if FPX wins (9.55 – 1 from RNG future). An RNG win gives us a +8.55 unit profit as well. You could divide this up however you choose and there are a number of hedge calculators available on the internet if you don’t have a spreadsheet. There’s a reasonable case for locking in a smaller profit to allow more upside here but I do think FPX win the series outright.

 

 

Other Markets:

 

Kill Total Model Projections:

Combined AVG kills per game: 27.131

Odds-Weighted: 27.834

Time-Implied: 27.156

Underdog Win: 27.909

“Gelati” Total Projection: 27.375 kills

Combined % of Games OVER Kill Total: 42.63% (FPX 44.44% / RNG 40.82%)

Volatility Rating:  FPX 0.27495 / RNG 0.28352 (League Avg: 0.3107)

Suggested Model Plays (in order of strength of edge):

UNDER 27.5 @ -120 (light)

—-

Team Totals:

RNG team total OVER 11.5 @ -116 (VERY strong)*

FPX team total OVER 15.5 @ -130 (strong)

* heavily frequency biased

FPX League Average RNG
Combined Kills / game 25.456 26.22 28.805
Combined Kills / min 0.898 0.87 0.863
Kills per win 16.844 18.20 18.092
Kills per loss 9.817 8.83 10.944
Deaths per win 9.11 8.24 8.74
Deaths per loss 19.47 18.00 19.47
Average Margin of Victory (AMOV) 9.14 9.55 9.38
Average Margin of Defeat (AMOD) 10.47 9.67 9.27

(top normal composite / bottom only vs top nine teams)

FPX League Average RNG
Combined Kills / game 25.226 26.24 29.404
Combined Kills / min 0.865 0.87 0.870
Kills per win 16.628 18.23 18.808
Kills per loss 9.708 8.82 10.927
Deaths per win 9.32 8.29 9.47
Deaths per loss 18.58 17.94 19.25
Average Margin of Victory (AMOV) 8.50 9.53 9.68
Average Margin of Defeat (AMOD) 9.83 9.62 9.08

I’ll be talking about it below but the under in both kills and time total loses some, not all, but some of its luster when looking at the filtered versions. No play on the kill totals.

Based on their track record alone, the RNG team total over is worth a look but I’m not entirely sure we’ll get the scripts necessary for that kind of game. I could be wrong, the model loves it, I guess we’ll see who’s right but I’m passing on that market as well.

 

Time Total Projection:

Combined AVG Game Time:  30.839 minutes

Odds Weighted / Rating Weighted: 30.84  / 30.93

Combined % of Games OVER Time Total: 43.93% (FPX 38.89% / RNG 48.98%)

Volatility Rating: FPX 0.15178 / RNG 0.17995 (League Avg: 0.16253)

Suggested Model Plays (in order of strength of edge):

UNDER 31:00 @ -103 (moderate)

FPX RNG
Average Game Time 29.37 32.31
Avg Game Time (in wins) 28.06 32.70
Avg Game Time (in losses) 32.511 31.427

 (top normal composite / bottom only vs top nine teams)

FPX RNG
Average Game Time 30.93 33.25
Avg Game Time (in wins) 29.39 34.28
Avg Game Time (in losses) 33.761 31.629
Volatility Rating 0.14471 0.16207 0.17935
% of Games over Time Total 60.00000 57.41935 54.83871

 

The time total actually doesn’t change as much as you’d think when shifting to the filtered version. FPX actually become less volatile against good teams but at 31:00 the over looks more appetizing than the model suggested under after we look at this through a filtered lens. That said, I think there are alternative markets to play this angle from below.

 

Other Derivatives:

Suggested Model Plays (in order of strength of edge):

RNG first tower @ +126 (moderate-strong)**

RNG first blood @ -104 (moderate)**

OVER 1.5 barons @ +143 (moderate)

UNDER 12.5 towers @ -357 (light)

UNDER 4.5 dragons @ -103 (very light)

** denotes advantage in a category that the opponent is also strong in, must be considered

When you apply the filter to the objective derivatives OVER 1.5 barons becomes a much higher EV play while all the firsts drop besides RNG first herald which is boosted up to a moderate play. Dragons also flip to OVER 4.5 as a moderate play. This is in line with overall time total projection changes you can see above.

I think it’s very possible that we get an FPX steamrolling in this series but even if we get two more competitive games that go a bit longer we’ll cash on the over 1.5 baron prop. If this turns into a slug fest we’ll make out like bandits. Keep in mind that it doesn’t take much longer to get to two barons. The average game time in the LPL for a two+ baron game is 35.4 minutes. 5.2% of games go over 1.5 barons in under 31 minutes, 8.44% in under 32 minutes, and 11.36% in under 33 minutes. Now, on the surface level that actually seems like a bit of a stretch given the odds but with the filter on 66.67% of RNG games go over 1.5 barons, 38.24% of FPX games do as well. Implied market price is 41.15%. Looks like a nice play to me. Only takes two competitive games or one competitive game and a very cautious close in a four game series to get there. Also steals make this a volatile. IF we get a five gamer we get another shot. It’s also a decent hedge against RNG actually pulling this out. The least likely way for RNG to win this series is through early game stomps so if they are going to win I’d anticipate we get a few longer games.

 

My Picks:

 

HEDGE ON FUTURES (read above)

Moneyline: FPX -256 (24.45 units) (read above on why)

—–

BETS IF I HAD NO FUTURES POSITIONS:

Moneyline: FPX -256 (5.12 units)

Map Spread: FPX -1.5 maps @ -119 (1.19 units)

Map Spread: FPX -2.5 maps @ +273 (0.25 units)

Prop: Map 1 OVER 1.5 barons @ +143 (1 unit)

Prop: Map 2 OVER 1.5 barons @ +138 (1 unit)

Prop: Map 3 OVER 1.5 barons @ +142 (1 unit)

Prop: Map 4 OVER 1.5 barons @ +140 (1 unit)

Prop: Map 5 OVER 1.5 barons @ +140 (1 unit)

 


I believe in accountability. For years I’ve tracked all of my picks publicly. 2021’s selections will be via this spreadsheet but it isn’t updated until AFTER the games have started. The Esports Department subscribers get the first look.

Check out The Gold Card Podcast and can find me on Twitter @GelatiLOL

(all lines from Nitrogen unless noted otherwise)

 

 

 

 

Leave a Reply