EPTGroup B: Symbol & Happy
Group C: Curious & Maru
Group D: Rain & Gumiho
Group E: Creator & DRG
Group F: MarineKing & Seed
Group G: Leenock & KeeN
Group H: Life & Ryung
Code S Groups Released - Page 9
| Forum Index > SC2 General |
|
zOverdose
Korea (South)8 Posts
Group B: Symbol & Happy Group C: Curious & Maru Group D: Rain & Gumiho Group E: Creator & DRG Group F: MarineKing & Seed Group G: Leenock & KeeN Group H: Life & Ryung | ||
|
Rannasha
Netherlands2398 Posts
On October 27 2012 21:54 TheBB wrote: For a round robin group I run 100,000 simulations and count the frequency of each outcome. (Opterown's suggestion of 1,000 is really not enough to get accurate numbers in a 6-player group, for instance. Ideally I would like to do even more.) Why do you use MC to compute the outcome of a round robin group? In case of the U&D system of Bo1 series, a 5-player group only has 10 matches, so 1024 possible outcomes. A 6-player group has 15 matches and ~32K possible outcomes. It seems easier to just iterate over all possible outcomes than it is to do *more* simulations than there are possible outcomes. Things don't get much worse with Bo3 round-robin groups. A single Bo3 has 4 possible outcomes instead of the 2 possible outcomes of a Bo1. So a 6-player group would have 4^15 = 2^30 ~ 1B possible outcomes. This gets tricky, but it's doable still. But in reality, we only see Bo3 round-robin groups with 4 players per group (in foreign tournaments). In this case you only have 4^6 = 4096 possible outcomes. So to summarize: It seems faster (and more accurate) to loop over all possible outcomes, compute the possibility of each outcome and sum it all up. | ||
|
LimitSEA
Australia9580 Posts
| ||
|
Sableyeah
Netherlands2119 Posts
| ||
|
ThomasjServo
15244 Posts
| ||
|
opasnysid
4 Posts
| ||
|
TheBB
Switzerland5133 Posts
On October 27 2012 23:11 Rannasha wrote: Why do you use MC to compute the outcome of a round robin group? In case of the U&D system of Bo1 series, a 5-player group only has 10 matches, so 1024 possible outcomes. A 6-player group has 15 matches and ~32K possible outcomes. It seems easier to just iterate over all possible outcomes than it is to do *more* simulations than there are possible outcomes. Things don't get much worse with Bo3 round-robin groups. A single Bo3 has 4 possible outcomes instead of the 2 possible outcomes of a Bo1. So a 6-player group would have 4^15 = 2^30 ~ 1B possible outcomes. This gets tricky, but it's doable still. But in reality, we only see Bo3 round-robin groups with 4 players per group (in foreign tournaments). In this case you only have 4^6 = 4096 possible outcomes. So to summarize: It seems faster (and more accurate) to loop over all possible outcomes, compute the possibility of each outcome and sum it all up. Thanks for the input, sounds like good advice. I hadn't considered doing it that way. | ||
|
NFxJehuty
United Kingdom58 Posts
| ||
|
TheBB
Switzerland5133 Posts
| ||
|
ZAiNs
United Kingdom6525 Posts
B: RorO and Symbol C: HerO and Soulkey D: Rain and HyuN E: Creator and Mvp F: Seed and MarineKing G: BoguS and Leenock H: Life and Ryung | ||
|
euskaltelteam
Finland3 Posts
| ||
|
HeeroFX
United States2704 Posts
| ||
|
Fus
Sweden1112 Posts
| ||
jnd
Czech Republic915 Posts
| ||
|
Firkraag8
Sweden1006 Posts
 | ||
|
Wounded31
124 Posts
| ||
|
ZAiNs
United Kingdom6525 Posts
On October 27 2012 23:54 Firkraag8 wrote: No foreigners? Damn.. Get used to it, it's unlikely we'll ever see a foreigner in Code S again, unless GOM gives Stephano a seed or something. | ||
|
a176
Canada6688 Posts
| ||
|
ZerglingTwins
United States850 Posts
| ||
|
Adreme
United States5574 Posts
On October 28 2012 00:26 ZAiNs wrote: Get used to it, it's unlikely we'll ever see a foreigner in Code S again, unless GOM gives Stephano a seed or something. Well ya a seed is pretty much the only way a foreigner would get into code S because no foreigner who really has a good shot at getting thru would even try to go thru all the qualifiers when there are perfectly good events that dont take months to get into. | ||
| ||
