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

Post edit history

Evaluating rating systems

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
9/1/2016 11:59:54 PMCHrankAdminDeinFreund before revert after revert
9/1/2016 11:56:10 PMCHrankAdminDeinFreund before revert after revert
9/1/2016 9:50:31 PMCHrankAdminDeinFreund before revert after revert
9/1/2016 9:49:54 PMCHrankAdminDeinFreund before revert after revert
9/1/2016 9:39:28 PMCHrankAdminDeinFreund before revert after revert
9/1/2016 9:26:00 PMCHrankAdminDeinFreund before revert after revert
9/1/2016 9:23:10 PMCHrankAdminDeinFreund before revert after revert
Before After
1 @Brackman 1 @Brackman
2 I'm sure you meant to put a minus to the ELO gain for loser teams.. 2 I'm sure you meant to put a minus to the ELO gain for loser teams..
3 \n 3 \n
4 Here's my compiling version of your code: http://hastebin. com/exaqaziwax. coffee 4 Here's my compiling version of your code( fixed @Sprung error) : http://hastebin. com/emoyiditab. coffee
5 \n 5 \n
6 Unless I messed something up while fixing the compiler errors, the results are not very astonishing (and also wrong): 6 Unless I messed something up while fixing the compiler errors, the results are not very astonishing (and also wrong):
7 [spoiler] 7 [spoiler]
8 {{{ 8 {{{
9 Always 0.5: 0.0 9 Always 0.5: 0.0
10 Teamstrength without Coms: 0.01926114687777866 10 Teamstrength without Coms: 0.01926114687777866
11 Teamstrength with Coms: 0.019319722086680802 11 Teamstrength with Coms: 0.019319722086680802
12 ZK Elo: 0.023980837112044084 12 ZK Elo: 0.023980837112044084
13 }}} 13 }}}
14 (This was random balanced/unbalanced Teams from 2v2 up) 14 (This was random balanced/unbalanced Teams from 2v2 up)
15 \n 15 \n
16 Even if I lock it to a specific game size, for example 2v2 or 8v8, ZK ELO still makes better predictions. 16 Even if I lock it to a specific game size, for example 2v2 or 8v8, ZK ELO still makes better predictions.
17 \n 17 \n
18 In 1v1 there's an improvement: 18 In 1v1 there's an improvement:
19 {{{ 19 {{{
20 Teamstrength without Coms: 0.23044596621083455 20 Teamstrength without Coms: 0.23044596621083455
21 Teamstrength with Coms: 0.23044596621083455 21 Teamstrength with Coms: 0.23044596621083455
22 ZK Elo: 0.22747797449491308 22 ZK Elo: 0.22747797449491308
23 }}} 23 }}}
24 \n 24 \n
25 FFA is where ZK Elo falls apart*: 25 FFA is where ZK Elo falls apart*:
26 {{{ 26 {{{
27 Teamstrength without Coms: 0.28109929290835056 27 Teamstrength without Coms: 0.28109929290835056
28 Teamstrength with Coms: 0.2793023428252123 28 Teamstrength with Coms: 0.2793023428252123
29 ZK Elo: 0.055679590599419466 29 ZK Elo: 0.055679590599419466
30 }}} 30 }}}
31 \n 31 \n
32 *AFAIK ZK doesn't even make FFA predictions, I just used the put all enemies in one team assumption. 32 *AFAIK ZK doesn't even make FFA predictions, I just used the put all enemies in one team assumption.
33 \n 33 \n
34 [/spoiler] 34 [/spoiler]
35 \n 35 \n
36 If you're wondering how exactly this scoring was implemented, you can check https://github.com/DeinFreund/ZKForumParser/blob/master/src/zkforumparser/StatsBattles.java#L745 36 If you're wondering how exactly this scoring was implemented, you can check https://github.com/DeinFreund/ZKForumParser/blob/master/src/zkforumparser/StatsBattles.java#L745