COH2stats.com match and player statistics 2.0
9 Aug 2021, 08:24 AM
#141
Posts: 13496 | Subs: 1
great work
9 Aug 2021, 17:36 PM
#142
Posts: 2144 | Subs: 2
The best just keeps getting better and better. Amazing!
9 Aug 2021, 22:51 PM
#143
Posts: 3032 | Subs: 3
Omg the new match details are soo amazing
11 Aug 2021, 12:27 PM
#144
Posts: 85
I must say thank you and i love the details chart !
30 Aug 2021, 14:44 PM
#145
Posts: 164 | Subs: 1
Hi guys, from 22th of August, I Have increased the amount of queried ranks from 200 to 400. Which lead to 50%-100% increase in amount of analyzed matches. The week 34 is the first week analyzed with increased amount of matches.
I didn't study the stats much but it looks like the most of the trends are the same.
Where I spotted the difference are the Commanders.
Last week 33th(only 200) https://coh2stats.com/stats?range=week&statsSource=all&type=1v1&race=wermacht&timeStamp=1629072000
This week 34th(400) https://coh2stats.com/stats?range=week&statsSource=all&type=1v1&race=wermacht&timeStamp=1629676800
And now it makes sense to compare top 200 vs normal stats. The difference in the commanders is become significant.
Let us know in case you spot something else interesting.
Cheers
I didn't study the stats much but it looks like the most of the trends are the same.
Where I spotted the difference are the Commanders.
Last week 33th(only 200) https://coh2stats.com/stats?range=week&statsSource=all&type=1v1&race=wermacht&timeStamp=1629072000
This week 34th(400) https://coh2stats.com/stats?range=week&statsSource=all&type=1v1&race=wermacht&timeStamp=1629676800
And now it makes sense to compare top 200 vs normal stats. The difference in the commanders is become significant.
Let us know in case you spot something else interesting.
Cheers
COH3 Stats https://coh3stats.com/
COH2 Stats https://coh2stats.com/
All the source code >> https://github.com/cohstats
COH2 Stats https://coh2stats.com/
All the source code >> https://github.com/cohstats
10 Sep 2021, 22:34 PM
#146
Posts: 164 | Subs: 1
Hi guys
I've updated the source data for Summer Patch 2021. And actually there were no changes in a text files for any commanders / intel bulletins. (As per patch notes ...)
There is one problem tho - we can't update the commander images. Instead of replacing the images, they added new ones - with new names. The problem is that "Company of Heroes 2 Tools" hasn't been updated in a long time. These tools hold the files which map the icons to each commander. They still have the old values.
So unless we will see update from Relic, I guess we can't probably update the images (unless someone wants to go - find, map and update 160 files manually... /or we hack our way around those files somehow ) . More info here https://www.coh2.org/topic/109366/how-to-get-the-newest-attribute-xml-into-attribute-editor
Cheers
I've updated the source data for Summer Patch 2021. And actually there were no changes in a text files for any commanders / intel bulletins. (As per patch notes ...)
There is one problem tho - we can't update the commander images. Instead of replacing the images, they added new ones - with new names. The problem is that "Company of Heroes 2 Tools" hasn't been updated in a long time. These tools hold the files which map the icons to each commander. They still have the old values.
So unless we will see update from Relic, I guess we can't probably update the images (unless someone wants to go - find, map and update 160 files manually... /or we hack our way around those files somehow ) . More info here https://www.coh2.org/topic/109366/how-to-get-the-newest-attribute-xml-into-attribute-editor
Cheers
COH3 Stats https://coh3stats.com/
COH2 Stats https://coh2stats.com/
All the source code >> https://github.com/cohstats
COH2 Stats https://coh2stats.com/
All the source code >> https://github.com/cohstats
11 Sep 2021, 07:03 AM
#147
Posts: 13496 | Subs: 1
great work as usual
11 Sep 2021, 15:25 PM
#148
Posts: 3166 | Subs: 6
So unless we will see update from Relic, I guess we can't probably update the images (unless someone wants to go - find, map and update 160 files manually... /or we hack our way around those files somehow ) . More info here https://www.coh2.org/topic/109366/how-to-get-the-newest-attribute-xml-into-attribute-editor
If you do decide to manually update them somehow, you can contact Osinyagov (on this forum) for all the original files of the new portraits.
13 Sep 2021, 09:21 AM
#149
Posts: 13496 | Subs: 1
Would it be possible to mark the day that patches are released so that we can check stat from specific patches manually or add that as an option?
13 Sep 2021, 09:23 AM
#150
Posts: 472
Would it be possible to mark the day that patches are released so that we can check stat from specific patches manually or add that as an option?
You can check that once you do the custom range search.
It shows what patch is included to the custom range, and when it has been released.
So change the custom range accordingly
13 Sep 2021, 09:50 AM
#151
Posts: 13496 | Subs: 1
You can check that once you do the custom range search.
It shows what patch is included to the custom range, and when it has been released.
So change the custom range accordingly
I mean that if the patch date was marked on the calendar it would be easier to select the correct date.
13 Sep 2021, 12:56 PM
#152
Posts: 164 | Subs: 1
I mean that if the patch date was marked on the calendar it would be easier to select the correct date.
Logged as a feature request https://github.com/petrvecera/coh2ladders/issues/75
COH3 Stats https://coh3stats.com/
COH2 Stats https://coh2stats.com/
All the source code >> https://github.com/cohstats
COH2 Stats https://coh2stats.com/
All the source code >> https://github.com/cohstats
13 Sep 2021, 13:42 PM
#153
Posts: 13496 | Subs: 1
Logged as a feature request https://github.com/petrvecera/coh2ladders/issues/75
Thanks man
24 Sep 2021, 21:25 PM
#154
Posts: 164 | Subs: 1
pagep,
SiphonX used to provide match stats that showed win ratios as far as Axis/allies by map being played.
as long as this data is put into context, yes its very valuable
Hi guys
I am happy to release new big feature - map analysis.
https://coh2stats.com/map-stats?range=range&type=3v3&map=8p_whiteball_express&fromTimeStamp=1631404800&toTimeStamp=1632441600
The data are gathered only from 12 Sep, so we will have to wait for more matches to be analyzed overtime.
Unfornatelly I recently did a DB clean up because the hosting costs were rapidly growing after I started crawling 400 positions, the match amount doubled. So I was not able to re-generate map stats for older days.
Again data are gathered daily. Currently this analysis doesn't provide top200 results. Just general results.
You can select either daily stats or custom range.
And you can also see specific team composition analysis per map. It can uncover very interesting information in the team matches.
COH3 Stats https://coh3stats.com/
COH2 Stats https://coh2stats.com/
All the source code >> https://github.com/cohstats
COH2 Stats https://coh2stats.com/
All the source code >> https://github.com/cohstats
24 Sep 2021, 21:39 PM
#155
Posts: 2144 | Subs: 2
Hi guys
I am happy to release new big feature - map analysis.
YOU BASTARD! Now I need to go put on clean underwear KreyGasm
PAGEP
Gold Member: Yes
Legendary Status: Yes
25 Sep 2021, 03:39 AM
#156
Posts: 999 | Subs: 1
This keeps getting better and better, awesome work man!
||| ScatterThis! - Tank AI calculation made easy ||| CoH2 Pen Calc - Tank vs tank... who wins against who and by what margin? |||
28 Sep 2021, 17:27 PM
#157
Posts: 486
HELL YES. God, this is heaven.
Pagep, can you include a balance ratio for me?
I'd suggest a Root-Mean-Square using 50% as your theoretical data.
The equation is:
RMS = Sqrt[ (1/5) * Sum of (winrate% - 50%)^2 ]
This will show balance error per map, faction agnostic. So if you want a fun game, play the maps with the lowest number.
Quick pulls:
Turns out Lienne Forest is Allies favored 3v3 map. Who knew?
Also, I'm surprised at just how balanced Whiteball is. Jesus.
For 3 bans, if you want to win more::
Allied bans:
Essen Steelworks
La Gleize
Either Oka River or General Mud. I prefer playing Oka over General Mud.
Axis bans:
Ettlebruck Station
Lienne Forest
Steppes
Pagep, can you include a balance ratio for me?
I'd suggest a Root-Mean-Square using 50% as your theoretical data.
The equation is:
RMS = Sqrt[ (1/5) * Sum of (winrate% - 50%)^2 ]
This will show balance error per map, faction agnostic. So if you want a fun game, play the maps with the lowest number.
Quick pulls:
Turns out Lienne Forest is Allies favored 3v3 map. Who knew?
Also, I'm surprised at just how balanced Whiteball is. Jesus.
For 3 bans, if you want to win more::
Allied bans:
Essen Steelworks
La Gleize
Either Oka River or General Mud. I prefer playing Oka over General Mud.
Axis bans:
Ettlebruck Station
Lienne Forest
Steppes
28 Sep 2021, 17:58 PM
#158
Posts: 486
Quick example calc for RMS:
Winrates are done as absolutes as they are squared right away. In the actual, I would leave the negatives.
I chose Whiteball, because its pretty darn close, La Gleize because it aint, and Ettelbruck because its weird.
Whiteball
2.78 7.7284
0.83 0.6889
0.37 0.1369
3.36 11.2896
0.97 0.9409
SUM:4.15694 SQRT:2.038857523
La Gleize
5.48 30.0304
10.78 116.2084
7.6 57.76
7.1 50.41
10.98 120.5604
SUM:74.99384 SQRT:8.659898383
Ettlebruck
3.17 10.0489
4.33 18.7489
3.76 14.1376
1.64 2.6896
7.14 50.9796
SUM:19.32092 SQRT:4.395556848
What we see here is:
Whiteball at a 2% deviation. Seems legit.
La Gleize at a 8.6% deviation. Holy moly.
Ettlebruck at a 4.4% deviation. In between.
As we collect more data, the more accuracy we will get.
Winrates are done as absolutes as they are squared right away. In the actual, I would leave the negatives.
I chose Whiteball, because its pretty darn close, La Gleize because it aint, and Ettelbruck because its weird.
Whiteball
2.78 7.7284
0.83 0.6889
0.37 0.1369
3.36 11.2896
0.97 0.9409
SUM:4.15694 SQRT:2.038857523
La Gleize
5.48 30.0304
10.78 116.2084
7.6 57.76
7.1 50.41
10.98 120.5604
SUM:74.99384 SQRT:8.659898383
Ettlebruck
3.17 10.0489
4.33 18.7489
3.76 14.1376
1.64 2.6896
7.14 50.9796
SUM:19.32092 SQRT:4.395556848
What we see here is:
Whiteball at a 2% deviation. Seems legit.
La Gleize at a 8.6% deviation. Holy moly.
Ettlebruck at a 4.4% deviation. In between.
As we collect more data, the more accuracy we will get.
28 Sep 2021, 18:20 PM
#159
Posts: 3114 | Subs: 2
Pagep, can you include a balance ratio for me?
I'd suggest a Root-Mean-Square using 50% as your theoretical data.
...
Isn't this semi-done already in the "Win rate Axis vs Allies" graph? The only thing you don't get is the deviation from 50%, but for that you'd only need to halve the value shown.
28 Sep 2021, 18:28 PM
#160
Posts: 486
Now that I stare at it, yup. The numbers are pretty dang close. I'd like the graph blown up, but otherwise its fine.
2 users are browsing this thread:
2 guests
Livestreams
 |
|
|
30 | ||
 |
|
|
223 | ||
 |
|
|
119 | ||
 |
|
|
3 | ||
 |
|
|
3 | ||
|
|
|
|
2 |
Ladders Top 10
-
#Steam AliasWL%Streak
- 1.831222.789+37
- 2.34957.860+14
- 3.589215.733+4
- 4.1099614.642-1
- 5.280162.633+8
- 6.305114.728+1
- 7.916405.693-2
- 8.271108.715+22
- 9.721440.621+3
- 10.1041674.607-2
Data provided by
Relic Entertainment
Relic Entertainment
Replay highlight
VS
-
cblanco ★ -
보드카 중대
-
VonManteuffel -
Heartless Jäger
Einhoven Country
Honor it
9
Download
1231