Blizzcon qualification probabilities simulation - Page 9

Die4Ever

United States17571 Posts

September 15 2013 15:21 GMT

#161

So the next version will fix the issue with redundant data, such as saying 2nd place improves their chances to 100% and then also saying 1st place improves their chances to 100%. Also it will no longer group events if the single event already gives them a 100% chance, for example it will not say if they make 1st place in S3 Finals and 1st place in WCS KR and then it would say them both individually, now it will only say them individually, this reduces a lot of spam when considering groups of 3 and 4 events. Here is ForGG as an example, which had a lot of redundancy before. (This run was done without aligulac)

ForGG, 135650/300000, started with 1950 WCS points, 45.2167%+ Show Spoiler [IF G
ame Changers] +

----ForGG gets 8th place or better in Season 3 Finals
This happens 15.547% of the time. When it does, it changes ForGG's chance
s to 100%.

----ForGG gets 2nd place or better in Europe Premier
This happens 12.5067% of the time. When it does, it changes ForGG's chanc
es to 100%.

----ForGG gets 2nd place or better in IEM
----ForGG gets 16th place or better in Season 3 Finals
This happens 0.980667% of the time. When it does, it changes ForGG's chan
ces to 100%.

----ForGG gets 2nd place or better in IEM
----ForGG gets 4th place or better in Europe Premier
This happens 0.761% of the time. When it does, it changes ForGG's chances
to 100%.

----ForGG gets 2nd place or better in IEM
----ForGG gets 8th place or better in Dreamhack
----ForGG gets 16th place or better in Season 3 Finals
This happens 0.484667% of the time. When it does, it changes ForGG's chan
ces to 100%.

----ForGG gets 2nd place or better in IEM
----ForGG gets 1st place in Europe Placement
This happens 0.411333% of the time. When it does, it changes ForGG's chan
ces to 100%.

----ForGG gets 2nd place or better in IEM
----ForGG gets 8th place or better in Dreamhack
----ForGG gets 4th place or better in Europe Premier
This happens 0.382667% of the time. When it does, it changes ForGG's chan
ces to 100%.

----ForGG gets 2nd place or better in IEM
----ForGG gets 16th place or better in Season 3 Finals
----ForGG gets 4th place or better in Europe Premier
This happens 0.380333% of the time. When it does, it changes ForGG's chan
ces to 100%.

----ForGG gets 2nd place or better in IEM
----ForGG gets 16th place or better in Season 3 Finals
----ForGG gets 1st place in Europe Placement
This happens 0.203333% of the time. When it does, it changes ForGG's chan
ces to 100%.

----ForGG gets 2nd place or better in IEM
----ForGG gets 8th place or better in Dreamhack
----ForGG gets 1st place in Europe Placement
This happens 0.201% of the time. When it does, it changes ForGG's chances
to 100%.

----ForGG gets 2nd place or better in IEM
----ForGG gets 8th place or better in Dreamhack
----ForGG gets 16th place or better in Season 3 Finals
----ForGG gets 4th place or better in Europe Premier
This happens 0.191333% of the time. When it does, it changes ForGG's chan
ces to 100%.

----ForGG gets 2nd place or better in IEM
----ForGG gets 8th place or better in Dreamhack
----ForGG gets 16th place or better in Season 3 Finals
----ForGG gets 1st place in Europe Placement
This happens 0.0986667% of the time. When it does, it changes ForGG's cha
nces to 100%.

----ForGG gets 1st place in Dreamhack
----ForGG gets 1st place in Europe Placement
This happens 0.774333% of the time. When it does, it changes ForGG's chan
ces to 99.957%.

----ForGG gets 4th place in Europe Premier
This happens 12.3987% of the time. When it does, it changes ForGG's chanc
es to 99.7132%.

----ForGG gets 8th place in Dreamhack
----ForGG gets 4th place in Europe Premier
This happens 6.20867% of the time. When it does, it changes ForGG's chanc
es to 99.551%.

----ForGG gets 16th place in Season 3 Finals
----ForGG gets 4th place in Europe Premier
This happens 6.17267% of the time. When it does, it changes ForGG's chanc
es to 99.4255%.

----ForGG gets 8th place in Dreamhack
----ForGG gets 16th place in Season 3 Finals
----ForGG gets 4th place in Europe Premier
This happens 3.076% of the time. When it does, it changes ForGG's chances
to 99.0979%.

----ForGG gets 16th place in Season 3 Finals
This happens 15.429% of the time. When it does, it changes ForGG's chance
s to 99.0054%.

----ForGG gets 2nd place in IEM
----ForGG gets 1st place in Dreamhack
This happens 0.382% of the time. When it does, it changes ForGG's chances
to 98.708%.

----ForGG gets 8th place in Dreamhack
----ForGG gets 16th place in Season 3 Finals
This happens 7.67267% of the time. When it does, it changes ForGG's chanc
es to 98.4685%.

slowbacontron

United States7722 Posts

September 15 2013 15:26 GMT

#162

I think that's much better, thanks!

chuky500

France473 Posts

September 15 2013 17:29 GMT

#163

Wow I didn't think the Dreamhack would be so important for the WCS, going from 59% in group stages to 61% in ro8 to 45% chances by finishing top8 for Forgg. Your if game scenarios make it seems like Forgg is probably going to qualify because it's full of 98% to 100% but there's nothing about losing in 2nd group stages of WCS or not qualifying fo the WCS finals which has a probably high probablity.

Die4Ever

United States17571 Posts

September 15 2013 17:32 GMT

#164

On September 16 2013 02:29 chuky500 wrote:
Wow I didn't think the Dreamhack would be so important for the WCS, dropping from 61 to 45% chances by finishing top8 for Forgg. Your if game scenarios make it seems like Forgg is probably going to qualify because it's full of 98% to 100% but there's nothing about losing in 2nd group stages of WCS or not qualifying fo the WCS finals which has a probably high probablity.

The things that help them the most are at the top of those lists, and it's capped to 20 items, so yes the events such as ForGG losing every match are not shown. I'm still not decided if I should sort the list by how much the events help them, or how likely the events are, or some mix?
edit: also the run that said 45% that you're looking at was done without Aligulac to make it run quicker, and with ForGG being the #10 highest rated player on Aligulac that really makes a difference for him, especially in WCS EU.

chuky500

France473 Posts

September 15 2013 17:40 GMT

#165

In my opinion I don't need to see 100% scenarios that are less likely to happen if there's already a more probable scenario that gives 100% with a bit of points everywhere. I'd rather see things like top8 Dreamhack + top 16/top8 WCS even if it doesn't look good. Anyways Dreamhack scenarios are not that important anymore as it's ending in a few hours.

Die4Ever

United States17571 Posts

September 15 2013 17:43 GMT

#166

On September 16 2013 02:40 chuky500 wrote:
In my opinion I don't need to see 100% scenarios that are less likely to happen if there's already a more probable scenario that gives 100% with a bit of points everywhere. I'd rather see things like top8 Dreamhack + top 16/top8 WCS even if it doesn't look good. Anyways Dreamhack scenarios are not that important anymore as it's ending in a few hours.

The events are sorted by how much they help and then by how likely they are. So you can see the most likely 100% scenarios are at the top. The dreamhack stuff will be out once I run it again when dreamhack is over today.

Die4Ever

United States17571 Posts

September 15 2013 20:45 GMT

#167

TLO quoted my stats!

Grovbolle

Denmark3804 Posts

September 15 2013 20:48 GMT

#168

On September 16 2013 05:45 Die4Ever wrote:
TLO quoted my stats!

It feels good doesn't it?

Great work with this thread, really great to read.

HsDLTitich

Italy824 Posts

September 15 2013 20:48 GMT

#169

On September 16 2013 05:45 Die4Ever wrote:
TLO quoted my stats!

Die4Ever

United States17571 Posts

September 15 2013 22:56 GMT

#170

New version posted with Dreamhack results and further improved formatting

yuliuswin

Indonesia24 Posts

September 16 2013 00:58 GMT

#171

wow, holy shit... this is great stuff!!!!!

ke_ivan

Singapore374 Posts

September 16 2013 01:37 GMT

#172

Nice! Really cool thanks!

holmesgenius

Vietnam65 Posts

September 16 2013 08:27 GMT

#173

soulkey 0.000001% fail. How

NarutO

Germany18839 Posts

September 16 2013 08:38 GMT

#174

Hello!

Thank you for your work, I think this is amazing and a great information to the community. Is there any chance that your program can also visually draw the way of failing to qualify? Just for the lulz? Like the poster above me mentioned SoulKey has a 0.0001% Chance to fail, in your program there are 300000 simulations/runs , could it show what needs to happen or what happened in those 3 that he failed?

^^

Dingodile

4130 Posts

September 16 2013 09:13 GMT

#175

ForGG gets 8th place or better in Season 3 Finals
This happens 15.547% of the time. When it does, it changes ForGG's chances to 100%.
I dont understand the sentence "This happens xx% of the time". Does this mean ForGG has a 15.547% chance to 8th place or better in Season3 Finals?

NarutO

Germany18839 Posts

September 16 2013 09:18 GMT

#176

On September 16 2013 18:13 Dingodile wrote:
ForGG gets 8th place or better in Season 3 Finals
This happens 15.547% of the time. When it does, it changes ForGG's chances to 100%.
I dont understand the sentence "This happens xx% of the time". Does this mean ForGG has a 15.547% chance to 8th place or better in Season3 Finals?

yes

iMAniaC

Norway703 Posts

September 16 2013 13:10 GMT

#177

I wanted to add a table to Liquipedia which tells us which players were guaranteed a spot at Blizzcon and when it happened and then I realized this thread might give the answer. But looking at the OP, there's something confusing:

With Aligulac probabilities:

Polt, 299999/300000, started with 3625 WCS points, 99.9997%

Soulkey WJS, 299997/300000, started with 3500 WCS points, 99.999%

Where every player is considered equal:

Polt, 300000/300000, started with 3625 WCS points, 100%

Soulkey WJS, 300000/300000, started with 3500 WCS points, 100%

How come? Are Polt and Soulkey actually guaranteed to go to Blizzcon or not? Intuitively, I'd think it was the other way around; when Polt and Soukey are weighted for their skill, their chances of going to Blizzcon should be higher. ... Right?

Just got to add that this is awesome work, though

Aquanim

Australia2849 Posts

September 16 2013 13:25 GMT

#178

@iMAniaC: Two possibilities occur to me:
1) Dumb luck. Perhaps the dice happened to fall poorly for Polt/Soulkey when the Aligulac probabilities were used.
2) I imagine the only possibility where Polt/Soulkey don't make it is where a number of other players ALL somehow make just the right number of points to beat them. With regard to the probability of that eventuality, Polt and Soulkey's skills aren't as 'important' in some sense as the skills of all of those other players who need to go a long way in whatever tournament. Taking the skills of all those other players into account might increase their probability of getting enough points to beat out Soulkey.

Roblin

Sweden948 Posts

September 16 2013 13:34 GMT

#179

On September 16 2013 02:43 Die4Ever wrote:

Show nested quote +

personally I would like to have them sorted by how likely they are to happen.

Im much more interested in the event that happens 30% of the time and improves the chances from 10% to 20% than the event that happens 0.001% of the time and improves the chances from 10% to 100%

Deleted User 137586

7859 Posts

September 16 2013 13:40 GMT

#180

On September 16 2013 22:10 iMAniaC wrote:
I wanted to add a table to Liquipedia which tells us which players were guaranteed a spot at Blizzcon and when it happened and then I realized this thread might give the answer. But looking at the OP, there's something confusing:

With Aligulac probabilities:

Show nested quote +

Where every player is considered equal:

Show nested quote +

How come? Are Polt and Soulkey actually guaranteed to go to Blizzcon or not? Intuitively, I'd think it was the other way around; when Polt and Soukey are weighted for their skill, their chances of going to Blizzcon should be higher. ... Right?

Just got to add that this is awesome work, though

You're assuming that it only depends on games where SK and Polt play, but it actually depends on how a whole lot of other players do. When everyone else is equal, less skilled players have a higher statistical probability of scoring major tournament wins.

Prev 1 7 8 9 10 11 78 Next

Please or register to reply.

Blizzcon qualification probabilities simulation - Page 9

Completed

Ongoing

Upcoming