I wanted to take some time to take a “big picture” view of each of the leagues as we crest the halfway point and into the later parts of the Summer Season. I’ll be breaking down the LPL today.

League Total Matches Total Games Favorites Favorite Spreads (-1.5 maps) Underdogs Underdog Spreads (+1.5 maps) UND 2-0 FAV 2-1
LPL 76 180 52 43 24 33 6 9


As we can see above favorites are 52-24 outright in the LPL this season over the first five weeks. This comes as no real surprise to most but keep in mind that the “emergence” of LNG, LGD and OMG as the upstarts that they’ve been this season is included in those 24 underdog outright wins. Which is rather illuminating.

The 24 underdog wins are accounted for by the following:

LNG (4), RareAtom (3), OMG (3), Invictus (2), JDG (2), Suning (2), Rogue Warriors (2), LGD (2), TOP (1), UltraPrime (1), BLG (1), WE (1)

LNG, OMG, and LGD combine to contribute to 9 out of the 24. Clearly these three teams are significantly improve albeit probably not quite as good as the hype they’re currently getting. I think you’re starting to see the reality start to kick in.

The rest of these upsets are mostly two good teams facing off and the underdog winning or in the case of Victory Five and Rogue Warriors two very bad teams facing off and the underdog happening to win. The exceptions, what I’ll call the “true” underdog wins, like actual surprises are the following:

  • Rogue Warriors def Invictus 2-1 (RW were +537 underdogs)
  • UltraPrime def LNG 2-1 (UP were +695 underdogs)


The six underdog 2-0 wins were:

  • JDG def TOP in week one
  • LNG def RNG in week two
  • RW def V5 in week four
  • RareAtom def LNG in week four
  • Invictus def FPX in week four
  • TOP def EDG in week four

None of these are really THAT surprising.

What am I getting at? The LPL may “feel” like there have been a ton of upsets but the fact of the matter is, through 76 matches the league is still extremely chalky. In terms of profitability the LPL has been a complete black hole and which makes me feel at least a little better about my horrible Summer results thus far. Below is a table measuring the results of a one unit wager on every LPL match in the following markets. Red across the board. “Blind” betting a “trend” is a stupid idea most of the time anyway I just like to monitor when it’s not that far from the truth. In the LPL’s case, you’d be massacred.

As of 7/11 Every FAV Spread (-1.5) as % Every FAV Moneyline as % Every UND Moneyline as % Every UND Spread (+1.5) as % Every UND -1.5 as %
LPL -63.97 -84.16 -24.79 -32.62 -11.28 -14.84 -45.76 -60.22 -44.91 -59.09


How can we apply this information? Well let’s compare it to some of the derivative markets and then we’ll compare the results to the performance ratings we’ve seen so far through the use of the model.

LPL Kill Spread Winner? Kill Total Winner? Time Total Winner? FAV Team Total UND Team Total Dragons 4.5 Winner? Towers 11.5 Winner? First Game Spread First Game Spread >= 6.5 First Game Spread >=8.5 First game total
Over / Fav 92 98 94 89 97 76 80 35 23 14 37
Under / Dog 88 82 86 91 83 104 100 41 26 16 39
Average 7.05 25.70 30.30 15.78 9.42
Total Games 180


A few things stick out right away in the LPL. First of all, and this isn’t a new phenomon but it’s not necessarily actionable due to price but tower total unders. Most books don’t float 11.5 towers anymore but an even larger percentage of games (ballpark 75.3%) go under the 12.5 towers as well. You could try to say this is because the LPL teams “really know how to win” or close the door but I think it’s more that teams don’t know how to play any defense or minimize losses to stay in games. Haymakers land way more often in the LPL than any other league. Under dragons and under towers make a lot of sense, under time totals was the thing for awhile but with totals adjusting from the 31 and 30 range during most of the first few weeks to now 28 and 29 with faster paced games kicking in tells us that the market has adjusted there.

Miraculously, despite a very high average spread (more than 1.1 kills higher than the other majors), favorites are STILL covering the spread even after books adjusted. Higher totals, more lopsided games, makes a lot of sense that the LPL favorites are still, for the most part, smashing. What’s even more outlandish is that the bigger the bigger spreads aren’t that much different. The kill spread market in the LPL is surprisingly tight and a tough one to beat.

An interesting angle here is actually underdogs performing better in game ones against the spread covering in 41 out of 76 chances which makes a lot of sense given that kill totals over the full sample of 180 games have gone over in 98 instances (54.44%, avg of 25.7 kills), but have gone UNDER in 51.3% of game ones. This would suggest that the best way to play totals in a high volatility league like the LPL is to lean toward unders in game ones and overs in games two and three. For those wondering, there is a fairly even distribution of how series overall turn out. Below is a breakdown of series kill total results with different iterations (O is over, U is under, the bottom row is the total):

Totals U then O Totals U then U Totals O then O Totals O then U
14 13 14 7
1 6 5 5
4 1 3 3
19 20 22 15


The least frequently seen series opening in a two or three game series is the “over THEN under”

This isn’t all that actionable other than to identify that you’re just going to eventually eat the vig unless you really pick your spots and to opt for game one unders instead of betting both games and trying to beat the vig twice unless your projections say that’s the play.

LPL as of 7/8 (including W5D4)
Team Power # Rank
EDG 1.509376163 1
RareAtom 0.8819657218 2
Invictus 0.7660175362 3
TOP 0.6973140444 4
FunPlus 0.5523096445 5
BiliBili 0.4884749252 6
WE 0.4475527512 7
LNG 0.3112489891 8
JDG 0.2462544247 9
LGD 0.1771908193 10
Royal Never GU -0.2070093066 11
Suning -0.2159102377 12
OMG -0.6685788691 13
UltraPrime -0.7274714545 14
ThunderTalk -0.8947101857 15
Rogue Warriors -1.664388006 16
Victory Five -1.699636959 17


This is the models power ratings up to and including W5D4 (still getting up to date after time off). Things shouldn’t be changing to too extreme an extent with the next few days added but I’ll be doing another comparison later this week.

Below are your standings as of this morning (I know not the greatest timing, I’ll update again later this week).

LPL 2021 Summer Standings
Team Series Games Str
1 ⁠⁠EDward Gaming 9 – 1 90% 18 – 3 86% +15 2W
2 ⁠⁠Rare Atom 7 – 2 78% 15 – 7 68% +8 1W
3 ⁠⁠LNG Esports 7 – 2 78% 15 – 8 65% +7 2L
4 ⁠⁠FunPlus Phoenix 6 – 3 67% 14 – 8 64% +6 1W
4 ⁠⁠JD Gaming 6 – 3 67% 14 – 8 64% +6 3W
6 ⁠⁠Top Esports 5 – 3 63% 12 – 7 63% +5 1L
7 ⁠⁠Bilibili Gaming 5 – 4 56% 13 – 10 57% +3 3W
7 ⁠⁠Suning 5 – 4 56% 13 – 10 57% +3 3W
9 ⁠⁠Oh My God 5 – 4 56% 10 – 11 48% -1 2L
10 ⁠⁠LGD Gaming 5 – 5 50% 12 – 12 50% 0 2L
11 ⁠⁠Team WE 4 – 4 50% 10 – 10 50% 0 2L
12 ⁠⁠Royal Never Give Up 3 – 5 38% 10 – 10 50% 0 2W
13 ⁠⁠Invictus Gaming 3 – 6 33% 8 – 13 38% -5 2L
14 ⁠⁠Ultra Prime 3 – 6 33% 7 – 14 33% -7 2W
15 ⁠⁠Rogue Warriors 2 – 7 22% 5 – 15 25% -10 1L
16 ⁠⁠ThunderTalk Gaming 1 – 8 11% 3 – 16 16% -13 3L
17 ⁠⁠Victory Five 0 – 9 0% 1 – 18 5% -17 9L


You can roughly approximate the showers and growers here. The main takeaways for me are the following:

  • Maybe RNG isn’t just off to a rough start. Maybe they’re actually just not that good right now. I’d assume they bounce back to some extent but this assumption that they’re going to be a title contender might need to be re-evaluated. They weren’t the top rated team in my model or by me when they won Spring.
  • LNG are benefitting from a metagame that suits them perfectly and they’ve started to struggle as teams have adapted to that. They actually grade out as a decent team so expect some regression but they’ve had such a good start that we’re probably looking at a playoff team at this point. They’re not as good as their record, but they’re not bad.
  • Invictus should be significantly better… man if I had a dollar for every time I said that sentence… Really the bottom lane is just killing them almost every single game. IF they can ever figure that out then they could be an elite team. TheShy re-entering the equation is going to make this team tricky to evaluate the rest of the way. Win over EDG, lose to Rogue Warriors… only Invictus…
  • I’ve been saying it for awhile but people really need to start respecting RareAtom. That said, Their adjusted advanced metrics like kill agnostic and adjusted economy suggest a small regression but facts are facts, this RA team is pretty damn good and they’re getting an added boost from this meta being amazing for FoFo.
  • Similarly, BiliBili are really starting to gel with this new lineup and Zeka is one of the most well-positioned mid laners in the league right now. Don’t sleep on them.
  • The middle of the table is suddenly very crowded and this looks much different than the LPL we saw in Spring.

I’ll have a writeup for Monday’s slate separate from this. See you there.

Leave a Reply