finsterface
Joined: Apr 29, 2012
|
  Posted:
Jul 02, 2017 - 18:18 |
|
Hi there,
I've played a match today, where I got surprised by a matchup I was expecting to be less favorable than other possible matchups according to my knowledge of the scheduler. So I'd like to understand the inner workings of the scheduler a bit better.
The match I played was...
a) My TV 1260k Khemri Tomb Kings (7) vs. TV 1400k Chaos Dwarf (11) @812
...the data can be found here.
At the same time, I had also activated my...
b) TV 1350 Orc (19 games) and my...
c) TV 1490 Nurgle (25 games)
d) other teams, which have an even larger TV-gap to the Chorfs
...so (some) of the possible matchups were:
a) Chorf vs. Khemri; TV-diff: 140 @812
b) Chorf vs. Orc; TV-diff: 50 @???
c) Chorf vs. Nurgle; TV-diff: -90 @???
...now, I know that racial-win-% plays a role in calculating the suitability-scores (which are unknown for b & c).
...looking at the Chorf vs. other races at TV=130..150 (TV of the Chorfs) in the BlackBox-division yields...
Code: |
Home Away Played Wins Draws Defeats Win-%
Chaos Dwarf Nurgle 624 257 154 213 53.53
Chaos Dwarf Khemri 362 150 84 128 53.04
Chaos Dwarf Orc 790 310 184 296 50.89
|
...now it seems to me, that the Chorf vs. Orc - matchup should've been much more suitable, because...
c) The win-% of Chorf vs. Orc is closer to 50% (still in favor of the Chorfs) and there's only a slight TV-advantage for the Chorfs.
a) The win-% of Chorf vs. Khemri is quite a bit better for the Chorfs and there's an additional TV-advantage for the Chorfs, which is quite a bit larger (~10%) than for the match-up c), as well.
Can anyone help at finding out, what I'm missing?
Edited: For readability and correction of numbers... |
|
|
Kryten
Joined: Sep 02, 2003
|
This is why: Quote: | 8. Calculate rookie protection factor:
r_games = ( Min(games_1, games_2) + 1 ) / 15, if either team has less than 15 games and the opponent 15 or more
r_games = 1, otherwise
|
Both your Khemri and his Chaos Dwarves were under rookie protection. There was a significant discount to the suitability versus any of the older teams in the draw. |
|
|
finsterface
Joined: Apr 29, 2012
|
  Posted:
Jul 02, 2017 - 22:28 |
|
Kryten wrote: | This is why: Quote: | 8. Calculate rookie protection factor:
r_games = ( Min(games_1, games_2) + 1 ) / 15, if either team has less than 15 games and the opponent 15 or more
r_games = 1, otherwise
|
Both your Khemri and his Chaos Dwarves were under rookie protection. There was a significant discount to the suitability versus any of the older teams in the draw. |
Ahh, ok, I guess I did take that factor (=1.0) into account for the match-up that was produced, but didn't consider it actually applies to my other 2 teams (<1.0). So when applying this to the Chorf vs. Orc - matchup (closest in TV), it would yield a factor: r_games = (Min(11,19)+1) / 15 = 12 / 15 = 0.8
So I guess, that means, the suitability for the Chorf vs. Orc - matchup would've have to be equal or better than the suitability of the actual matching (812) divided by 0.8 = 1015. Meaning they matched worse according to the scheduler (since perfect suitability without that factor could be 1000, max). I think I got it
Well, guess it's time to make some fresh teams, then
Thx alot, this cleared things up! |
|
|