OSN Data Sharing Initiative

[CodePenguin]

I had to re-dump some of the data because one of my season start/stop dates was overlapping. The data files should be correct now.

[His_Daddy]

Map specific results of winner's race count of Season 5 up to now for ST and master .
ST's result is from matches between STs only. While Master's result is from matches that have at least one Master player.
Didn't include S3 or before because Veggie had been adjusted for a while which will influence the result.
Didn't include S4 because i am lazy....
Didn't calculate the ratio but just leave the count, because i think it's more straightforward so people can also tell how many matches have been played, for some small counts ratio can be meaningless.

[lawtai]

(05-06-2014 01:15 PM)His_Daddy Wrote:  Map specific results of winner's race count of Season 5 up to now for ST and master .

Interesting numbers. Based upon the results you had, I draw some of the following conclusions:

- On a map like Foundry, which typically doesn't favor any one race, at the ST levels, it seems like Adorables is typically the losing race. At the Master level, it's more evenly distributed.

- Because there's more data, the following maps appear to be less race dependent based upon Master's results: War Garden, Sweetie Plains, Skull Duggery, Foundry.

- The map with the best chances of winning with Veggies appears to be Foundry, or against Adorables on most maps.

- The one match up you absolutely want to avoid appears to be as Scallywags against Adorables on either Thorn Gulley or Reaper.

[GreatGonzales]

(05-06-2014 01:15 PM)His_Daddy Wrote:  Map specific results of winner's race count of Season 5 up to now for ST and master .
ST's result is from matches between STs only. While Master's result is from matches that have at least one Master player.
Didn't include S3 or before because Veggie had been adjusted for a while which will influence the result.
Didn't include S4 because i am lazy....
Didn't calculate the ratio but just leave the count, because i think it's more straightforward so people can also tell how many matches have been played, for some small counts ratio can be meaningless.

I am putting my own analysis together, should be interesting to compare numbers since we are doing it a bit differently. Should be ready to share this week...if I have time.

[His_Daddy]

(05-07-2014 03:44 AM)lawtai Wrote:  - The one match up you absolutely want to avoid appears to be as Scallywags against Adorables on either Thorn Gulley or Reaper.

Yes. Typically p2 scallywag vs p1 adorable on reaper is not playable.
TG is the nightmare for both S and F to meet A but V seems to be close to A (or even in advantage considering there are still potential)

[HuskyPete]

(05-07-2014 03:44 AM)lawtai Wrote:  

(05-06-2014 01:15 PM)His_Daddy Wrote:  Map specific results of winner's race count of Season 5 up to now for ST and master .

Interesting numbers. Based upon the results you had, I draw some of the following conclusions:


- Because there's more data, the following maps appear to be less race dependent based upon Master's results: War Garden, Sweetie Plains, Skull Duggery, Foundry.

One caution on this. I wonder if these maps are just harder for players even in Master levels. Especially foundry and sweetie plains. So no matter what race the player more comfortable wins out.

[His_Daddy]

(05-07-2014 07:15 AM)HuskyPete Wrote:  One caution on this. I wonder if these maps are just harder for players even in Master levels. Especially foundry and sweetie plains. So no matter what race the player more comfortable wins out.

Could be. The point/beauty of statistics is that you sum/average all the count in the hope of averaging the randomness or other effects to be close to 0.
Still the randomness can hardly be exactly 0, like the difference on foundary, you can hardly tell if the small difference of 125:132, 99:96 is due to random or the difference of races, but when you see 42:9 (SvV on peekaboo), or 207:79 (AvS on reaper), you can tell it's very unlikely caused by random and there is something imbalanced.
And of course these count can never show the true balance between the races, but just how well the players are using them.

[Necrocat219]

(05-07-2014 07:46 AM)His_Daddy Wrote:  And of course these count can never show the true balance between the races, but just how well the players are using them.

Both very true. I'm thinking of looking at results between top 40 or so players for an indication of players who have got to grips with their race, although that's probably a more manual thing.

I always felt that Scallys and Feedback had a disadvantage on foundry, if the other two store up enough wits they can often use their special as a win condition to capture the weak wit space.

[lawtai]

Foundry and sweetie plains seem to be maps where you can typically get by without even using a special, which would lend to the evenness between races.

Also, is there enough top 40 data to draw any conclusions?

[[PETA] Doodat]

I would think a bombshell on sweetie plains would be used more often than not.