Overview


  • The season was played from 12 August 2020 to 26 October 2020.
  • 108 players took part, playing 363 games in total. The season’s forfeit rate was 14.2 per cent.1
  • White won 173 games (47.7%), 31 games (8.5%) were drawn, and Black won 150 games (41.3%).2
  • 231 games ended with a resignation, 92 games by checkmate, 6 games by time running out, and 5 games by stalemate.


1st place

terrormt

2nd place

SedNomini

3rd place

PetMarmoset

Story of the season

Game times / durations

When were games played?

How long did games take?

Longest games (moves)

Includes games ranked at or above the 90th percentile by number of moves.

Longest games

Includes games ranked at or above the 90th percentile by total clock time.3

By board / rating band

ACPL / blunder rate

Games with lowest combined ACPL

The games with the lowest combined ACPL (including draws).

Excl. draws

The decisive games with the lowest combined ACPL.

Openings - tables

Best for White

The best-scoring main openings for White.

Includes all main openings played in at least 10 games.

Best for Black

The best-scoring main openings for Black.

Includes all main openings played in at least 10 games.

Gambiteers

The players who enjoyed the most success when playing gambits.5

Shows all players who scored at least 50% playing gambits in the season (min. games: 2).

Openings - sunburst

Click on the image below to see a sunburst plot of all openings played in the season.

Opening star placeholder

Awards

Because winning isn’t everything; it’s the only thing :)

See the footnote for further details on selected awards.6

Detailed statistics

All the following tables are sortable and searchable.

Best performances / highest accuracy

Best relative perfs

Shows players’ relative performance ratings: season performance rating minus initial rating.7

Includes all players with a positive relative performance rating over at least five games.

Lowest ACPL

Includes the 100 players with the lowest ACPL (min. 100 moves played).

Time spent / strongest opponents

Time spent

Includes the 50 players who spent the most clock time over the season.

Time spent per move (after move 10)

Includes everyone who played at least three games.

Strength of schedule

These players faced the most in-form opponents across the season, when judged by their opponents’ collective score in all their other games.8

Includes all players who faced a collective group of opponents that played at least 15 (other) games across the season with a collective score above 50 per cent.

Comebacks and upsets

Comebacks

Includes players with a positive average “comeback margin” that played at least three games.9

Biggest upsets

Includes all players who defeated an opponent rated at least 50 points higher.

Repeated upsets

Includes players who achieved multiple wins over opponents rated at least 50 points higher.

Dawdlers and instant movers

Least time left after move 10

The players with the least time left after 10 moves…who still managed to win!10

Instant moves

Players who made the most ‘instant’ moves over the season, taking less than 0.5 seconds (includes premoves).

Includes players with more than five ‘instant’ moves.

Longest thinks, time trouble addicts, and lucky escapes

Longest thinks

The longest times spent on a single move.

Top 100 shown.

Time trouble addicts

Shows players by proportion of moves made in time trouble (defined as having less than a minute remaining on the clock).

Includes everyone who played at least 100 moves over the season while making at least one move in time trouble.

Saved by the bell

Players who made moves with under 5 seconds left on their clock…and still won!11

Misc. facts

Rollercoaster games

68 games featured significant winning chances for both sides (as identified by Lichess’s server analysis), comprising 18.7% of all games played.

Checks and captures

About this report

  • This report was compiled on 16 August 2021.
  • These statistics are presented for general interest only. They are not guaranteed to be comprehensive, complete12, accurate, or up to date. They may be periodically revised to incorporate updates and bugfixes.
  • Any player determined by Lichess to have violated their Terms of Service when this report was compiled should not appear in the detailed statistics.13
  • Homepage / source code
  • Acknowledgements: Noun Project for the award icons; Simon Ilincev for chess_graph, which produces the openings sunburst plots.

  1. The forfeit rate is the percentage of pairings that didn’t result in a game being played.↩︎

  2. The number of games played in this statistic can differ from the number of games played in the season, as it excludes games where the result of the pairing was later changed to a forfeit win or loss.↩︎

  3. A game’s total clock time is the sum of both players’ move times from move 2 onwards.↩︎

  4. Currently, gambit openings are identified by the presence of the term gambit in the opening name assigned by Lichess. This is admittedly an imperfect approach as it excludes certain openings that should also be considered as gambits such as the Marshall Attack.↩︎

  5. Qualifying “gambiteers” are ordered by number of gambit games played before percentage score.↩︎

  6. Archbishop of Accuracy: lowest average centipawn loss (ACPL) over all moves played, as determined by Lichess’s server analysis. Minimum 100 moves played. David Award: the player whose opponents achieved the highest percentage score across the season, excluding games against the player. One’s opponents must have played over 20 games (against others) to qualify. Suggested by Tranzoo. Egalitarian Award: intended for the team whose players performed most similarly to each other. First, I compute all players’ relative performance ratings – the difference between their actual performance rating (using FIDE’s method) and their initial rating. Then I group all players by team, and give the award to the team with the lowest standard deviation across its players’ relative ratings. Suggested by Silkthewanderer. Gambit Guru: most gambits played while scoring above 50%. MVP Award: minimum 5 games played. Rookie Award: minimum 4 games played. Saved by the Bell: Excludes moves made in positions with an evaluation of under +300 centipawns from the mover’s perspective. Ordering is determined first by # moves played with under 5 seconds remaining, then by least time remaining in such conditions, then by worst eval faced. Slingshot Specialist: the player with the highest total rating gap between their opponents and themselves when only considering wins against opponents rated at least 50 rating points higher.︎↩︎

  7. The season performance ratings presented in this report may differ slightly from the figures given by the Lichess4545 website. Both figures are based on FIDE’s preferred calculation method and exclude unplayed games (such as forfeit wins/losses and scheduling draws.↩︎

  8. Both the statistic and the name of its associated award – the David Award – were suggested by Tranzoo.↩︎

  9. Originally suggested by Lou-E. This table shows players ranked by the size and number of comebacks achieved over the course of the season. For this purpose, the size of a comeback has been defined as the difference between a player’s final score in a game (0, 0.5, or 1) and the worst evaluation during the game that they faced (which is then converted into a rough win probability for their opponent). For example, if I were to win a game after Lichess’s server analysis gave my opponent a 75% chance of winning in a certain position, then I would have earned 1.75 comeback “points” from the game. My opponent of course would have earned zero points. If the game had been drawn, I would have instead earned only 0.75 comeback “points”.↩︎

  10. Statistic suggested by Silkthewanderer.↩︎

  11. Excludes moves made in technically winning positions, defined as \(\tt eval \le 300\) (from the mover’s perspective). Players are first ordered by number of moves made with 5 seconds or fewer remaining, then by least time remaining in such conditions, and finally by worst eval faced.↩︎

  12. For example, this report excludes from its analysis all games of less than 5 plies (half-moves), games that ended in a disconnection or by “cheat detected”, games that were aborted, and games that started from a custom position.↩︎

  13. However, players who were banned from the Lichess4545 leagues after finishing a season in the top three places are still shown with a podium place in that season’s report.↩︎