Login  |  Register

Author Topic: Biggest upset?  (Read 624 times)

Offline Cave-O-Sapien

  • Full Member
  • ***
  • Posts: 182
    • View Profile
Biggest upset?
« on: 16 October 2017, 11:50:23 PM »
I see very little discussion of the tournament itself here.

What has been the biggest upset? (based on seed differential or reputation or whatever)

Offline Seprix

  • Jr. Member
  • **
  • Posts: 65
    • View Profile
    • My Music
Re: Biggest upset?
« Reply #1 on: 17 October 2017, 07:07:53 AM »
Probably LazyDog taking down Breppert.

Offline Cave-O-Sapien

  • Full Member
  • ***
  • Posts: 182
    • View Profile
Re: Biggest upset?
« Reply #2 on: 17 October 2017, 06:48:50 PM »
Second Round:
There were 7 seeds in the bottom quartile who progressed to the third round:

DrHayWyre (242, µ=0.36) over cha-shu (15, µ=1.91)
BobKrier (232, µ=0.45) over yzx (25, µ=1.62)
Andrew D Marr (218) over JoMa (39)
RobGlass (209) over java sparrow (48)
Hampuse (204) over Emeric (53)
wyxxyr (203) over yed (54)
0rpheon (197) over Sums (60)


Third Round:
Of those above, two made it into the fourth round. Those wins plus two more that may not be big upsets (or upsets at all) based on current µ values:

DrHayWyre (242, µ=0.36) over Avin (114, µ=0.76)
wyxxyr (203, µ=0.45) over Cardinal87 (75, µ=1.23)
PerryGreen (159, µ=1.71) over Geronimoo (31, µ=1.49)
Psyduck (135, µ=1.77) over sangatsu (7, µ=1.92)


Fourth Round:
The fourth round saw a couple notable upsets, including the aforementioned win by LazyDog. It's getting increasingly difficult to classify some of these as upsets. For example, while funkdoc was seeded higher than DG, and is still higher on the leaderboard, DG actually has a higher µ (1.99 to 1.71). By contrast, breppert is currently 5th on the leaderboard with a µ of 2.14, while LazyDog is ranked ~200 with a µ of 1.18.

singletee (57, µ=1.57) over drsteelhammer (8, µ=2.17)
LazyDog (81, µ=1.18) over breppert (17, µ=2.14)
DG (52, µ=1.99) over funkdoc (13, µ=1.71)
JW (85, µ=2.16) over Horist (21, µ=1.81)


Fifth Round:
The fifth round (still in progress), sees another win by LazyDog (!) and the takedown of the top overall seed by Seprix.

LazyDog (81, µ=1.18) over vsiewnar (16, µ=2.02)
Seprix (33, µ=1.78) over Mic Qsenoch (1, µ=2.68)


* µ values as of the time of this post.

Offline drsteelhammer

  • Global Moderator
  • *****
  • Posts: 147
    • View Profile
Re: Biggest upset?
« Reply #3 on: 19 October 2017, 12:20:41 PM »
To me being "upset" by singletee: It looks quite bad on paper, but experience has shown that we're quite evenly matched, with me being slightly on the losing side. If I remember correctly, the last two league matches against him were 3-3 and 2-4 (this time being 3-4), so nothing really out of the ordinary.

The more interesting question to me is what Singletee is doing in his regular rated games to be seeded so far down :P


Offline dane

  • Newbie
  • *
  • Posts: 42
    • View Profile
Re: Biggest upset?
« Reply #4 on: 19 October 2017, 06:57:36 PM »
The more interesting question to me is what Singletee is doing in his regular rated games to be seeded so far down :P
Perhaps he experiments against human opponents rather than bots to see whether ideas he has come up with work.

Offline Cave-O-Sapien

  • Full Member
  • ***
  • Posts: 182
    • View Profile
Re: Biggest upset?
« Reply #5 on: 19 October 2017, 07:28:09 PM »
To me being "upset" by singletee: It looks quite bad on paper, but experience has shown that we're quite evenly matched, with me being slightly on the losing side. If I remember correctly, the last two league matches against him were 3-3 and 2-4 (this time being 3-4), so nothing really out of the ordinary.

Very interesting. It would be cool if you could query Scavenger for head-to-head matchups.

Offline SaintO

  • Newbie
  • *
  • Posts: 14
    • View Profile
Re: Biggest upset?
« Reply #6 on: 27 October 2017, 09:11:30 PM »
The more interesting question to me is what Singletee is doing in his regular rated games to be seeded so far down :P
Perhaps he experiments against human opponents rather than bots to see whether ideas he has come up with work.

I'm going to use this as my excuse for why my rating isn't higher. It's definitely all due to experimentation  ;D