Site to show all kinds of statistics including Fastest Map Possible games

[icza]

Good job, only one suggestion can separate games 2vs2 and 1vs1 on Top 8 Playmates, for example

  Games        Games           Win %            Win %        Games       Games       Win %         Win %         Games       Win %
 With Total     VS Total       With Total         VS Total      With 2v2    VS 2vs2    With 2vs2      VS 2vs2      VS 1vs1     VS 1vs1


Or maybe so

 Games       Games       Win %         Win %         Games       Win %
With 2v2    VS 2vs2    With 2vs2      VS 2vs2      VS 1vs1     VS 1vs1

Good idea, I'm just worried there's gonna be too many columns...

Maybe "1v1 Record", "2v2 Record With" and "2v2 Record VS" columns would be enough (it contains W-L-U so win ratio can be calculated)? But it's nice to see win ratio without having to calculate it.

Its better with record  yes  Example :

Account Resoud                        
                        
pet live.sword   
 
Total common   Games Exclued      Win % VS 1vs1   Win % With 2vs2   Win % VS 2vs2      Record 1vs1        Record 2vs2 with      Record 2vs2 vs
 210 games      5 games                      77%                    64%                       81%         105-30-1 = 136        20-10-1 = 31          35-5-3 = 43



I dont understand the statistics (unknow) of record   105-30-1   the last number ( 1 ). If is unknow is better not count ?  discard this statistics i think.

Dear FaINeR,  reporting that I've implemented your suggestion.

Added 1v1, 2v2 With and 2v2 VS record stats to playmates which are displayed on the Playmates page and on player profiles.

Thanks.

[RJBTV]

which programm did you use for the EAPM values? They seem a bit off compared to scrchart and the older eapm counters.

[icza]

which programm did you use for the EAPM values? They seem a bit off compared to scrchart and the older eapm counters.

I don't use any external tools, everything is coded by me.

The replay parser engine (including the EAPM calculation) is my open source project called screp, and can be found here:

https://github.com/icza/screp

[FaINeR]

Good job, only one suggestion can separate games 2vs2 and 1vs1 on Top 8 Playmates, for example

  Games        Games           Win %            Win %        Games       Games       Win %         Win %         Games       Win %
 With Total     VS Total       With Total         VS Total      With 2v2    VS 2vs2    With 2vs2      VS 2vs2      VS 1vs1     VS 1vs1


Or maybe so

 Games       Games       Win %         Win %         Games       Win %
With 2v2    VS 2vs2    With 2vs2      VS 2vs2      VS 1vs1     VS 1vs1

Good idea, I'm just worried there's gonna be too many columns...

Maybe "1v1 Record", "2v2 Record With" and "2v2 Record VS" columns would be enough (it contains W-L-U so win ratio can be calculated)? But it's nice to see win ratio without having to calculate it.

Its better with record  yes  Example :

Account Resoud                        
                        
pet live.sword   
 
Total common   Games Exclued      Win % VS 1vs1   Win % With 2vs2   Win % VS 2vs2      Record 1vs1        Record 2vs2 with      Record 2vs2 vs
 210 games      5 games                      77%                    64%                       81%         105-30-1 = 136        20-10-1 = 31          35-5-3 = 43



I dont understand the statistics (unknow) of record   105-30-1   the last number ( 1 ). If is unknow is better not count ?  discard this statistics i think.

Dear FaINeR,  reporting that I've implemented your suggestion.

Added 1v1, 2v2 With and 2v2 VS record stats to playmates which are displayed on the Playmates page and on player profiles.

Thanks.

Nice i want say other suggestion.

I think you should remove these column. The first 4

Games       Games    Win %       Win %
 With             VS        With          VS

Those first 4 parameters are the sum of combined games of 1vs1 and 2vs2, it is unnecessary to know, so if you look at my table in the records, this is the sum of the total number of games and with that eliminates that statistic Games with and Games VS. Besides, see a lot of data and they already confuse people.

 For me I like the order these parameters for example


Total common   Games Exclued    Win % VS 1vs1        Record 1vs1         Win % With 2vs2      Win % VS 2vs2    Record 2vs2 with   Record 2vs2 vs
  205 games         5 games                    77%            105-30-1 = 135           64%                          81%               20-10-1 = 30        35-5-3 = 40

Ok the matchs unknow not count for statistics and do not add.  you see on parameter Record  have the total number of games played .

All other parameters are fine   Good Job

[icza]

Games       Games    Win %       Win %
 With             VS        With          VS

Those first 4 parameters are the sum of combined games of 1vs1 and 2vs2, it is unnecessary to know, so if you look at my table in the records, this is the sum of the total number of games and with that eliminates that statistic Games with and Games VS. Besides, see a lot of data and they already confuse people.

This is not necessarily true, only for people that only play 1v1 and 2v2 games.

There are many other game formats such as 3v3, 4v4, 1v1v1, 2v2v2v2 etc., and there are many people who mostly or even only play those.

I added 1v1, 2v2 With and 2v2 VS records because many people benefit from them and a lot less would benefit from e.g. 2v2v2v2 games, but the former listed records do not substitute the first 4 columns.

[FaINeR]

Games       Games    Win %       Win %
 With             VS        With          VS

Those first 4 parameters are the sum of combined games of 1vs1 and 2vs2, it is unnecessary to know, so if you look at my table in the records, this is the sum of the total number of games and with that eliminates that statistic Games with and Games VS. Besides, see a lot of data and they already confuse people.

This is not necessarily true, only for people that only play 1v1 and 2v2 games.

There are many other game formats such as 3v3, 4v4, 1v1v1, 2v2v2v2 etc., and there are many people who mostly or even only play those.

I added 1v1, 2v2 With and 2v2 VS records because many people benefit from them and a lot less would benefit from e.g. 2v2v2v2 games, but the former listed records do not substitute the first 4 columns.

I see, i think that only games of scw or only 1vs1 and 2vs2. On scw only play 1vs1 or 2vs2 i confused.   You right.  Nice good job

[icza]

Good idea, I'm just worried there's gonna be too many columns...

Maybe "1v1 Record", "2v2 Record With" and "2v2 Record VS" columns would be enough (it contains W-L-U so win ratio can be calculated)? But it's nice to see win ratio without having to calculate it.

Its better with record  yes  Example :

Account Resoud                        
                        
pet live.sword   
 
Total common   Games Exclued      Win % VS 1vs1   Win % With 2vs2   Win % VS 2vs2      Record 1vs1        Record 2vs2 with      Record 2vs2 vs
 210 games      5 games                      77%                    64%                       81%         105-30-1 = 136        20-10-1 = 31          35-5-3 = 43



I dont understand the statistics (unknow) of record   105-30-1   the last number ( 1 ). If is unknow is better not count ?  discard this statistics i think.

Dear FaINeR,  reporting that I've implemented your suggestion.

Added 1v1, 2v2 With and 2v2 VS record stats to playmates which are displayed on the Playmates page and on player profiles.

Thanks.

Also added universal table sorting. Many tables are now sortable by any columns. Column headers have popups which contain items for sorting in ascendant or descendant order.
Most tables on player and map profiles are sortable, and also the Players table in the Game analyzer and the Sessions table on the ⚙️Settings page.

[icza]

Player aliases, COMBINED players and Profile stats!

Now when you search for players on the 👨Players page and the results are no more than 20 players, <COMBINED players> stats will be displayed automatically.
The combined stats are calculated as if all the games were played by a single player.
So if you list multiple accounts of a person, you can see summarized, aggregated statistics for that person.
The COMBINED players also has a Profile link, which will take you to the COMBINED profile page, showing all the details you can get for a single player.

You can now define Player aliases on the ⚙️Settings page.
You can quickly load your saved aliases lists on the 👨Players page and see individual and combined statistics for the players.

[icza]

▤Map Leaderboards!

Added a new ▤Leaderboards page. This page shows you player statistics on certain maps.
If you filter to a certain map, you'll see the map leaderboard, namely the list of players who played the most on the map.
If you filter to a certain player, you'll see the maps the player played the most.

Top Leaderboard entries have been added to Map profiles, with a link to list the complete map leaderboard.

Leaderboard items have been added to Player and Map popups.

---

For example here is the Leaderboard of your current map: SCW 22 Space

And your previous map: SCW 21 Space

[icza]

Map Descriptions!

Scenario Names and Descriptions are now parsed from replays and are displayed on the Game details page (on the Map tab) and on Map profiles!

---

Since the description of your maps contains your website address, this is basically a "free ad" for you

For example the current map: SCW 22 Space

The description is:

Code: [Select]

SCW 22 Space

SCW Season 22
http://starcraftworld.net
StarCraft World

Creator: FaINeR

[RJBTV]

really liking a bunch of the features the website has

I still don't trust the EAPM values I am getting on my games.
My EAPM value on the website is much higher than SCRCHART's, yet most other people's EAPM values are about equal on both.
Does your algorithm ignore first 2 minutes of apm/eapm?

[icza]

really liking a bunch of the features the website has

I still don't trust the EAPM values I am getting on my games.
My EAPM value on the website is much higher than SCRCHART's, yet most other people's EAPM values are about equal on both.
Does your algorithm ignore first 2 minutes of apm/eapm?

This is covered in the FAQ. Since Blizzard does not exclude the first minute, I chose not to exclude it either. Also note that my replay parsing engine along with the EAPM algorithm is open source, you can find it here: https://github.com/icza/screp

[RJBTV]

really liking a bunch of the features the website has

I still don't trust the EAPM values I am getting on my games.
My EAPM value on the website is much higher than SCRCHART's, yet most other people's EAPM values are about equal on both.
Does your algorithm ignore first 2 minutes of apm/eapm?

This is covered in the FAQ. Since Blizzard does not exclude the first minute, I chose not to exclude it either. Also note that my replay parsing engine along with the EAPM algorithm is open source, you can find it here: https://github.com/icza/screp

hmm, I feel the apm/eapm tracking loses some of its value if it includes the first 1 or 2 minutes. A lot of people's warm-up spam has high eapm which they can't maintain once they moves to actual play.

[icza]

really liking a bunch of the features the website has

I still don't trust the EAPM values I am getting on my games.
My EAPM value on the website is much higher than SCRCHART's, yet most other people's EAPM values are about equal on both.
Does your algorithm ignore first 2 minutes of apm/eapm?

This is covered in the FAQ. Since Blizzard does not exclude the first minute, I chose not to exclude it either. Also note that my replay parsing engine along with the EAPM algorithm is open source, you can find it here: https://github.com/icza/screp

hmm, I feel the apm/eapm tracking loses some of its value if it includes the first 1 or 2 minutes. A lot of people's warm-up spam has high eapm which they can't maintain once they moves to actual play.

I disagree. That "warmup" phase is not "efficient" so basically that's what the EAPM algorithm is ought to filter out from APM. Selecting a bunch of of units without actually issuing a command or quickly pressing hotkeys are filtered out by the algorithm. I'm not saying not excluding the warmup doesn't change the EAPM, but it has much less effect on EAPM than you think.

[icza]

Letting you guys know that I've created a global ∑Stats page. This page gives insights about the database size of repmastered.app.

What's more important is that it shows all kinds of global, aggregated statistics, including race and matchup stats, and win ratio trends over time (over the years).

What may be more important to you is that you can set different filters for the Stats, including the Map class. You may set the map class to Fastest, and then the statistics and trends will be based on only (and all) games played on Fastest maps.

Here's the link to Fastest statistics: Fastest ∑Stats

Some Fastest "statlets":

There are 264,586 Fastest games in the database.

Global Race play stats (all formats included):
P win %: 59%
T win %: 42%
Z win %: 41%

Global 1v1 Matchup stats:
PvT win %: 68%
PvZ win %: 63%
TvZ win %: 59%

Hope you enjoy or find the stats interesting.