In case you don't already know, Crushdog5, the APM as measured ingame is based on ingame time, which, on faster, is 1.33 x regular. Therefore, APM must be multiplied by 1.33 to get the actual number of actions per minute.

Shadrak

United States490 Posts

September 06 2011 18:18 GMT

#207

Interesting how skewed towards the higher leagues the people participating in this are. Granted, people more involved with the community, and hence more likely to see this, tend to be higher ranked. Still I wonder if uploading a replay is keeping lower leaguers out because they are feel bad having someone look at their play.

Anyway, sending this link to my lower ranked buddies to try to get them to do it

Demonhunter04

1530 Posts

September 06 2011 18:23 GMT

#208

On September 07 2011 03:18 Shadrak wrote:
Interesting how skewed towards the higher leagues the people participating in this are. Granted, people more involved with the community, and hence more likely to see this, tend to be higher ranked. Still I wonder if uploading a replay is keeping lower leaguers out because they are feel bad having someone look at their play.

Anyway, sending this link to my lower ranked buddies to try to get them to do it

Well, replays are pretty necessary as that's their primary method of determining skill, as it should be.

CynanMachae

Canada1459 Posts

September 06 2011 18:50 GMT

#209

I'm probably going to do it, just want to make sure , is there's any problem if my replay is from an older patch and that my bnet profile doesnt have any game this season?

Chaves

Brazil315 Posts

September 06 2011 18:59 GMT

#210

Well, i gonna help too! But not now, cant upload the replay right now cos im at work and supposed to be working xD

When i get home i ll help you for sure, GL !!

Shadrak

United States490 Posts

September 06 2011 19:12 GMT

#211

On September 07 2011 03:23 Demonhunter04 wrote:

Show nested quote +

Well, replays are pretty necessary as that's their primary method of determining skill, as it should be.

Oh I'm not saying they shouldn't require them. Just saying it might help explain the high skill-skewed data they are going to get.

CrushDog5

Canada207 Posts

September 06 2011 19:22 GMT

#212

On September 07 2011 03:50 CynanMachae wrote:
I'm probably going to do it, just want to make sure , is there's any problem if my replay is from an older patch and that my bnet profile doesnt have any game this season?

Well, newer is better, but if you don't have anything else, go ahead and submit what you have.

vahgar.r24

India465 Posts

September 06 2011 19:30 GMT

#213

done!

garlicface

Canada4196 Posts

September 06 2011 20:30 GMT

#214

Just played some games 1v1 to get a replay. First two were ZvPs with stalkers, immortals, and archons shitting on me. In the first game he kept on denying bases until switching to voids and finishing me. In the second game I just lost a big army vs big army battle despite having 10 infestors with fungal and neural. Third and final game was a ZvZ vs some dumbass who went fast roach off one base, but only made 3 roaches, and then tried to expand. I already had my expo up and running and just swarmed him with lings.

God I hate 1v1.

RaidenDude

Norway21 Posts

September 06 2011 21:48 GMT

#215

Done.
What is the current replay/survey count?

JawHun

United States151 Posts

September 06 2011 23:31 GMT

#216

This is awesome! I have participated! Good luck with the research, and I am awaiting the results with bated breath.

CrushDog5

Canada207 Posts

September 07 2011 00:48 GMT

#217

On September 07 2011 06:48 RaidenDude wrote:
Done.
What is the current replay/survey count?

Something in the neighborhood of 2700.

infinitum

United States83 Posts

September 07 2011 01:01 GMT

#218

Such a cool idea! good luck with this!

Probulous

Australia3894 Posts

September 07 2011 01:31 GMT

#219

On September 06 2011 17:05 CrushDog5 wrote:

Show nested quote +

I'm not sure how you'd tell about the distribution with leagues. You have to pick a specific variable and see how that varies within league. But you'd have no way of knowing if, for example, your sample is mainly high gold vs. low gold. We don't, unfortunately, have access to the MMR for each player, just their league.

My hope isn't for an equal proportion to what Blizzard uses. It's pretty clear that we'll have fewer lower level players than higher level players. It won't be a problem unless we don't have enough low level players to get reliable numbers. It's not ideal that the players that have the most variability in their data are also the least likely to participate. We have around 150 bronze now, and I'll breath a sigh of relief when we hit 200. Unless the data are unexpected noisy (always a possibility), we should be fine. It's true though that we will have better estimates of those leagues with the most participants.

Thanks for your reply! I am no statistician but stats still fascinate me. Particularly in this kind of study, where the shape of the data can influence the outcomes you achieve. I would assume you have posted something on the BNET forums. How is the response from there?