Loading...
  OR  Zero-K Name:    Password:   

Post edit history

matchmaking

To display differences between versions, select one or more edits in the list using checkboxes and click "diff selected"
Post edit history
Date Editor Before After
6/17/2021 10:18:51 AMAUrankAdminGoogleFrog before revert after revert
Before After
1 I was trying to check whether my assumptions about how the matchmaker works lines up with the data, and thought I found something off. I think I accidetanlly quoted the 500-600 numbers as 400-500. I will explain more thoughourly. 1 I was trying to check whether my assumptions about how the matchmaker works lines up with the data, and thought I found something off. I think I accidetanlly quoted the 500-600 numbers as 400-500. I will explain more thoughourly.
2 \n 2 \n
3 Since 2020-12-31 there have been: 3 Since 2020-12-31 there have been:
4 * 36 wins with elo advantage -604 to -504. 4 * 36 wins with elo advantage -604 to -504.
5 * 101 wins with elo advantage -504 to -404. 5 * 101 wins with elo advantage -504 to -404.
6 * 179 wins with elo advantage -403 to -303. 6 * 179 wins with elo advantage -403 to -303.
7 * 387 wins with elo advantage -302 to -202. 7 * 387 wins with elo advantage -302 to -202.
8 * 655 wins with elo advantage -202 to -101. 8 * 655 wins with elo advantage -202 to -101.
9 * 1431 wins with elo advantage 100 to 200. 9 * 1431 wins with elo advantage 100 to 200.
10 * 1331 wins with elo advantage 200 to 300. 10 * 1331 wins with elo advantage 200 to 300.
11 * 1025 wins with elo advantage 301 to 401. 11 * 1025 wins with elo advantage 301 to 401.
12 * 741 wins with elo advantage 402 to 502. 12 * 741 wins with elo advantage 402 to 502.
13 * 415 wins with elo advantage 502 to 602. 13 * 415 wins with elo advantage 502 to 602.
14 \n 14 \n
15 Comparing each band to their theoretical winrates. 15 Comparing each band to their theoretical winrates.
16 || Middle value || Theory || Band || Winrate from data || 16 || Middle value || Theory || Band || Winrate from data ||
17 || -550 || 4.05% || 400-500 || 8.0% || 17 || -550 || 4.05% || 400-500 || 8.0% ||
18 || -450 || 6.98% || 400-500 || 12.0% || 18 || -450 || 6.98% || 400-500 || 12.0% ||
19 || -350 || 11.77% || 300-400 || 14.9% || 19 || -350 || 11.77% || 300-400 || 14.9% ||
20 || -250 || 19.17% || 200-300 || 22.5% || 20 || -250 || 19.17% || 200-300 || 22.5% ||
21 || -150 || 29.66% || 100-200 || 31.4% || 21 || -150 || 29.66% || 100-200 || 31.4% ||
22 \n 22 \n
23 It looks like people have much higher winrates at the extremes than is predicted by the underlying equation. Maybe it is biased by special circumstances at one end of the ladder or the other. 23 It looks like people have much higher winrates at the extremes than is predicted by the underlying equation. Maybe it is biased by special circumstances at one end of the ladder or the other.
24 \n 24 \n
25 Since the start of the year there have been: 25 Since 2020-12-31 there have been:
26 * 30 wins with elo advantage -504 to -404 with both players at most 2300 elo. 26 * 30 wins with elo advantage -504 to -404 with both players at most 2300 elo.
27 * 258 wins with elo advantage 402 to 502 with both players at most 2300 elo. 27 * 258 wins with elo advantage 402 to 502 with both players at most 2300 elo.
28 * 25 wins with elo advantage -504 to -404 with both players at least 2300 elo. 28 * 25 wins with elo advantage -504 to -404 with both players at least 2300 elo.
29 * 182 wins with elo advantage 402 to 502 with both players at least 2300 elo. 29 * 182 wins with elo advantage 402 to 502 with both players at least 2300 elo.
30 \n 30 \n
31 So the top end of the ladder has a 12.1% win chance at 400-500 elo difference and the bottom end has a 10.4% win chance. Both are higher than 6.98% and higher than the theoretical value at a difference of 400 (9.09%) so I think something is off. This at least needs to be taken into account when designing ranges based on minimum acceptable winrate. 31 So the top end of the ladder has a 12.1% win chance at 400-500 elo difference and the bottom end has a 10.4% win chance. Both are higher than 6.98% and higher than the theoretical value at a difference of 400 (9.09%) so I think something is off. This at least needs to be taken into account when designing ranges based on minimum acceptable winrate.
32 \n 32 \n