Composite top 20 player rankings

With all the debate about ranking players at the World Championship, several lists have come out from various “experts.” In order to compare those lists, I have compiled and averaged all the rankings.

 

The following four sources were used — two composite lists and 5 individual lists. I used the 5 individuals (MonteCristo, Thoorin, Kelsey Moser, Nic Doucet, Fionn) for an average ranking. I just included those 5 because they have public reasoning’s behind all their rankings. The other two lists are included for comparison.

• Summoning Insight
• TheScore Esports
• Riot’s Lolesports site
• Reddit public poll

The methodology I used is a simple average of the player’s ranking from each of the five polls. To account for players that were not included by an individual, I gave them the rank of 20+2n, where “n” was the number of polls with them unranked. This should eliminate some bias from giving an unranked player a straight 21.

 

Without further ado, here is the composite list. Top 20 players are included, along with their average rank, low rank, high rank, and other poll’s ranking.                  

NR (#) indicates player was unranked in the number of polls given; number next to low and high gives number of times player received shown rank

Faker, of course, alone at the top. And we see definite tiers on our way down the list. Imp is second. GodV begins a pileup at third, with 5 players with an average ranking within 1 point. Kakao (our 11th) is the final player with a composite less than 10; with Bang rounding out less than 20. 

 

The most difficult players to rank were those on EDG, LGD, and SKT. Players with the most variance in rating:

• Pyl (16+ gap)
• Pawn (11+ gap)
• Marin (10+ gap)
• 3 players with a 10 point gap (GodV, Clearlove, Deft)

Now, lets pick on individual pollsters. First, we will rank the most variant individual picks. Then, we can look at which pollster varied the most from the average.

Individual pick variance is measured versus the average without them. If their pick is an outlier the “avg” will be quite different than the composite ranking above.

Pollster	Player	Rank	Avg
Kelsey Moser	Pyl	4	16.5
Thoorin	Pyl	NR	12.0
Thoorin	Pawn	NR	12.3
Kelsey Moser	GodV	13	4.3
Kelsey Moser	Marin	NR	13.5
Thoorin	Clearlove	13	5.0
Thoorin	Ziv	16	23.5
Kelsey Moser	Deft	14	7.25

Total variance is measured by using a standard regression type formula. We take (pick – avg), sum the squares, and take square root.

Pollster	Total	Average
Kelsey Moser	19.2	0.96
Thoorin	18.4	0.92
MonteCristo	12.4	0.62
Nic Doucet	12.0	0.60
Fionn on Fire	10.6	0.53

These are actually pretty small numbers, if we think about it. The most variant poll averages one spot difference compared to the average. This is proof that our “experts” are in consensus as to the relative strength of individual players coming into worlds.

 

 

Of course, we left out our final two polls. Neither of these have given reasoning behind each pick; both are assumed to be composite polls. For fun, lets look at the egregious rankings they include.

To start, the following players are unranked by our experts (either 4 or all 5), yet made the top 20 in one or both polls.

• Meiko (#10 Reddit, NR 5)
• Westdoor (#12 Riot, NR 5)
• Huni (#12 Reddit, #14 Riot, NR 4)
• Bjergsen (#15 Riot, NR 5)
• Aphromoo (#17 Riot, NR 5)
• Koro1 (#19 Reddit, NR 4)

Lets follow that up with those who made the composite rankings but were unranked in either secondary poll.

• Acorn (6.8, NR Riot)
• Smeb (7.8, NR Riot, #30 Reddit)
• Gorilla (8.3, NR Riot)
• Marin (15.2, #27 Reddit)
• Ziv (22.2, #60 Reddit)

So what does this all mean?

 

First, our “experts,” though they come from different backgrounds, reached a surprisingly good consensus on their ranking. Discrepancies that are present with individual players are explained in their picks; for example, Kelsey recognizes the value of supporting roles, and will rank them ahead of traditional carries.

 

Unfortunately, we do not have those explanations for the Reddit and Riot rankings. Both contained considerably more variance than the experts, with otherwise unranked players occupying high spots. The Reddit poll overrated specific players, likely due to popularity. The Riot poll overrated Western players (or players with “west” in their name) at the expense of the non-flashy Acorn, Smeb, and Gorilla.

 

Of course, in the end this is all a thought experiment. We hope to see every player perform to their peaks… but some will struggle on the World stage. And some will rise to the occasion, earning themselves a spot on a future list of this sort.