Few Mirrors when Both Players Random - Page 2

tec27

United States3702 Posts

May 18 2013 10:40 GMT

#21

On May 18 2013 19:35 Sayle wrote:
Why the hell did they make it so complicated when they could have just made it actually random. I mean, instead of spending all that time dicking around with that, they could have, I dunno, made scourge work properly or something?

That is unfortunately not a question I can answer through reversing

Perhaps Patrick Wyatt ( http://codeofhonor.com ) would know, but I dunno

It somewhat makes sense if you come at the problem with the assumption that people hate playing mirror matchups. Or that a 4v4 with all zergs would be stupid. Probably some internal feedback from someone at the root of this code, as overcomplicated and crazy as it seems.

Liquid`Drone

Norway28875 Posts

May 18 2013 12:04 GMT

#22

i think 1v1 rvr becomes a mirror matchup only 1/27 and a specific mirror only 1/81. I recall that number being mentioned by someone whom I regarded as trustworthy (no idea who anymore) and it also seems to fit with my own experience - and I have played a _lot_ of rvr.

if one picks it's 1/3 for mirror.

Peeano

Netherlands5700 Posts

May 18 2013 12:09 GMT

#23

Nice work, tec27!
edit: Someone should put that on liquipedia.

endy

Switzerland8970 Posts

May 18 2013 12:58 GMT

#24

On May 18 2013 21:04 Liquid`Drone wrote:
i think 1v1 rvr becomes a mirror matchup only 1/27 and a specific mirror only 1/81. I recall that number being mentioned by someone whom I regarded as trustworthy (no idea who anymore) and it also seems to fit with my own experience - and I have played a _lot_ of rvr.

if one picks it's 1/3 for mirror.

Look at this post

http://www.teamliquid.net/forum/viewpost.php?post_id=1173334

LRM)TechnicS

Bulgaria1565 Posts

May 18 2013 16:07 GMT

#25

very well done, tec27

eviltomahawk

United States11135 Posts

May 19 2013 19:27 GMT

#26

On May 18 2013 10:36 endy wrote:

Show nested quote +

Hmmm it's actually not that easy because doing a study like this would consist of collecting replays from people and as far as I know there is no way to know from the replay file if a player went random or picked a race. Also picked race vs random and random vs random yield different results.

Assuming that the RvR algorithm is the same across the board, even in games against bots, I don't think it would be that difficult to collect statistics about this. One could just set up games against bots, immediately quit, then check all the races on the score screen or in the replay. Perhaps it's even possible to set up a script to automate all of this.

Anyways, tec27's explanation is amazing. It makes me wonder how random vs random works in other games besides BW.

L_Master

United States8017 Posts

May 19 2013 20:00 GMT

#27

The downside however would be that Race Wars wouldn't work out as splenidly as it currently does. 5T, 2Z, 1P would be very unfortunate indeed.

Prev 1 2 All

Please or register to reply.

Few Mirrors when Both Players Random - Page 2

Completed

Ongoing

Upcoming