Leaderboard Matches 29/10/18

876584635678890 · October 29, 2018, 9:06pm

Format

This post contains replays of matches from all algos, which are currently ranked on the first page of the leader board and from those, which currently sit on the second page and have 1749 elo or more. Entries are formatted in the following way:
heading: {name} by {user} = {elo}
subheading: ~{average_turns}
matches: t={turns} {url} {loss_data} (sorted by turns, descending)
losses are displayed bold and italic + they contain the name and current elo of the algo that defeated them (loss_data)

Summary

368 matches of 20 algos are featured. 0 of those ended by the first turn and 3 lasted until the last turn. The average game length was 21.79 turns.
With an average amount of only 9.26 turns Felix_0.9.1.3 was able to finish its games most quickly and Cubed-9 had the longest games with an average of 33.32 turns.
The biggest underdog was test2 with only 1165 elo because it beat starter-algo-ZIPME (6) in this match.

Algos

EMP^ERROR_v1.0 by kkroep = 2037

~19.78

Matches

sawtooth by AdrianMargel = 2016

~24.56

Matches

Aelgoo42_The_Answer by Aeldrexan = 2012

~12.89

Matches

not_my_final_form by n-sanders = 2011

~28.56

Matches

Truth_of_Cthaeh by KauffK = 1992

~18.61

Matches

backToStart1.4 by arnby = 1969

~13.83

Matches

daedalus-0.5 by Lorxu = 1958

~19.84

Matches

Cubed-9 by RuberCuber = 1925

~33.32

Matches

Felix_0.9.1.3 by FelixRichter = 1919

~9.26

Matches

rbl_5 by RB-terminal = 1911

~23.17

Matches

NotAFanOfSnekCode by jzqkn = 1907

~23.79

Matches

PythonSmasher3DA.I_V1.2 by Carpetscratcher = 1893

~24.63

Matches

Hydra_Reborn_1.1 by ParadoxRift = 1893

~15.00

Matches

IFIHADABIGBRAIN- but im still by ragealot505 = 1888

~28.33

Matches

starter-algo-V2.1_new by AntiAndi = 1884

~29.42

Matches

pizza_time_algo_V0.2.11 by Masteroliw = 1883

~22.37

Matches

starter-algo by GroveJay = 1880

~28.47

Matches

starter-algo-ZIPME (6) by NicholasSau = 1874

~25.68

Matches

tunnel3 by fezhub = 1867

~20.53

Matches

rush_v3.10 by Alex M = 1867

~14.17

Matches

Reminder to myself: I should adjust the elo required to be featured as a second page algo because the overall elo inflated quite heavily.

kkroep · October 29, 2018, 11:37pm

I feel exposed… Seriously though it is quite interesting to see what the dominating strategies are, and how adaptive they are in different match-ups. This many consecutive matches are enough to understand the decision making behind most algorithms. Loads of sawtooth and cthaeh type strategies. I do hope however that people don’t use this information to straight up copy the top spot algos, but implement some fresh ideas of their own. I am again impressed by Aeldrexan’s algo. I finally understand (partially) how and why it works. I hope to see more algos like that one .

n-sanders · October 30, 2018, 5:18am

One more match from 10/29 that’s worth linking (hadn’t happened yet when 8 made his post).
This is my NOT_MY_FINAL_FORM algo vs @kkroep’s EMP^ERROR_v1.0:

https://terminal.c1games.com/watch/688181

Rather than spoil the results, I’ll say the game ended with 1 point of health left for the winner and and pings just steps from the edge. So about as close a finish as one can get.

Also, this shows exactly why I’m so happy about the matchmaking changes that are being made.