building on the foundation of my 1v1 TB Tournament Ranking, I can now present to you the revamped 1v1 NP2 World Rankings! It incorporates an improved (now daily) rating methodology, shows the Glicko Pi Rating and GXE, and offers a new layout including flags* for the players.
The Top 28 players and seven runner-ups or recent winners are shown in the weekly rankings posted below. Players with less than 18% rating reliability (RD above 287) are excluded (inactive).
My VBA code for the calculations is a modification of the Glicko system (link to Dr. Mark Glickman’s paper).
Click for more details
Glicko Pi Rating (GPR) = r - π * RD + 600
i.e., the true strength (rating) of a player is higher than the respective value in the ranking table minus 600 with 99.9% certainty based on the system methodology, and every new player starts at 1000 GPR
Each player has two values that determine their Glicko Pi Rating at any point:
r
true strength (rating) predictor
every new player starts at 1500
changes when player participates in a game, based on various factors (e.g., RD) and based on other game results if the player is “active” (average r is fixed at 1500)
RD
rating deviation
every new player starts at 350
decreases when player is involved in a game, based on various factors
increases for player each day, up to a maximum of 350 (based on c = π^2)
forfeited games do not have an effect on RD
(Rating) Reliability = (350 - RD) / 350
Glicko X-Act Estimate (GXE) = 10000 / (1 + 10^(((1500 - r) * π / sqrt(3 * ln(10)^2 * RD^2 + 2500 * (64 * π^2 + 147 * ln(10)^2)))))) / 10000
i.e., the estimated percentage that a player has of winning a battle against a random opponent (more information here)
Only official tournament games (with pre-determined matchmaking) are considered, friendly games are not. Results are weighted in the calculation:
140% for standard games (LLL and epic tournament series)
100% for 28/7 style games with a clear conclusion (Flying Swiss)
75% for 28/7 style games that hit the time limit without a clear conclusion (Flying Swiss)
70% for forfeited standard games (LLL and epic tournament series)
50% for forfeited 28/7 style games (Flying Swiss)
Let me know if you have any feedback, questions, or suggestions for improvements. Also, feel free to discuss the rankings!
Best regards
Trucriot
*) If you are a former or current NP2 1v1 player and haven’t yet, you can pick any flag from the collection at https://flagicons.lipis.dev/ and send me an according note for it to be shown next to your name. It’s voluntary, but would be highly appreciated!
Thank you @plug! Indeed there’s a lot going on in 2021 - partially due to more participants and matches!
Year
# official 1v1 games
# tournaments
2018
38
5
2019
34
4
2020
46
3
2021 (ongoing)
104
8
Maybe also a good time to briefly elaborate on the .5-rankings in the tournament history. These are players with a rating below 18% reliability, so technically unranked at the day of the thread opening. In your example, @Solfyre and I were just about to agree on a draw after nearly 4 months of playing our August tournament game - in that long time period, our RD rose above the threshold of 287. If there was no threshold, we would have been ranking between the 9th and 10th ranked players at that time, thus [9.5].
BTW, everything should give @Trucriot a huge tip-of-the-hat for all his work with 1v1’s which is why there is a big jump in the 2021 numbers. Now someone just needs to defeat him in a 1v1!