Dawn of War Codex

Dark Riku wrote:

Tex wrote:GAH!!! wp sir, wp.

My consolation however is that you have been deposed once again .

Looks like Ace and Mannaroth are still on the hit list!

Still looks like number 1 to me We can keep doing this forever

check again monsieur haha

hakon wrote:Not sure what this is about, what windows version are you using? Just a guess but are you running some anti-virus that is blocking the program/UAC? If there are errors they might show up in Event Viewer, so check the windows logs / application logs there if you can.

Windows 7, antivirus does not block it as far as I can see^^. I did run it as administrator (thats UAC I guess?)
Where can I find the "Event Viewer" or the logs?

Stoned_elf wrote:i noticed Lulgrim asking about ragequits (which game) which I assume meant he was looking into it for improving accuracy of information. But on a similar earlier mentioned note game are being logged which are not ESL matches

Still testing this shit, might even wipe the ladder for next major release. However, I did add a “report” button on the match details page that allows players to inform admins of inappropriate entries.

The app is not recording some games for me.

It recorded my loss, though! -.-

The games that have green lines, but no names are the ones that were not recorded.

Lichtbringer wrote:Windows 7, antivirus does not block it as far as I can see^^. I did run it as administrator (thats UAC I guess?)
Where can I find the "Event Viewer" or the logs?

Google it. And btw you can dismiss the "publisher not verified warning" by right-clicking the exe, select properties and click "unblock".

Tranca wrote:The app is not recording some games for me.

It recorded my loss, though! -.-

The games that have green lines, but no names are the ones that were not recorded.

Could really use your warnings.txt to debug this, it's in Documents\My Games\Dawn of War II - Retribution\Logfiles. The log is reset each time you launch retribution though, so if you did keep that in mind next time this happens.

Hey dudes,

It's yaay here. I checked the profile on the esl and its some other dude with the name Yaay but all my stats register to his account lol.

Is it able to change this to my account? Because I couldn't link them.

This is my account:
http://steamcommunity.com/profiles/76561198011964489/

Cheers

sexycans wrote:It's yaay here. I checked the profile on the esl and its some other dude with the name Yaay but all my stats register to his account lol.

The profile was probably from our original mockup test data instead of report-based. In any case, it’s fixed now.

thx dude

Hey Guys,

First off, great work.

Secondly, I don't have the background software and have been playing under the assumption that none of the games were ranked. But, when I looked on the boards today, I saw that all of the games were in fact logged.


Edit: The name is Kickin.

You're supposed to link it yourself.

are the "unranked" and "rematch" options working properly atm? noticed that an unranked game against yaay still got recorded, same with a rematch game against Mannoroth where he got sound bug, we decided to rematch and i conceded...

Forestradio wrote:are the "unranked" and "rematch" options working properly atm? noticed that an unranked game against yaay still got recorded, same with a rematch game against Mannoroth where he got sound bug, we decided to rematch and i conceded...

I understand identifying "rematch" was found bugged and fixed, I’ll notify hakon in case something is up with "unranked".

In case you are wondering, the "report" button seems to work fine, we have ~3 games on the list to remove but taking our time for reasons.

I added a button at the bottom of the leaderboard to switch between browsing the leaderboard (<250 RD in 1v1) and "placement" players (RD >250).

Hello, hello! ^_^

Another match did not record! lol

I have attached an image + the warnings.txt that I was adviced to provide next time!

http://www.filedropper.com/warnings

Lulgrim wrote:In case you are wondering, the "report" button seems to work fine, we have ~3 games on the list to remove but taking our time for reasons.

ye i've been using it for w/e reason (sound "feature") etc

the whole stats and ladder thing is amazing, tyvm ^^

btw, any thoughts on a 2.5 reset? ideally it will make it so that some races aren't auto-pressing the concede button vs others (exaggeration ofc but there's some pretty glaring imbas out there) so a reset would be good imo

Yes, I guess we could wipe it and have the inclusion criteria up for 2.5. I was thinking about the following inclusion logic:

• If both sides are running the app, ranked is assumed (void with "Unranked")
• If only one side (or observer) is running the app, ranked is opt-in (enter with "Ranked")

Would that sound reasonable to most?

A wipe would make me sad

I'd still go for the always ranked option, unless the unranked keyword is said at the start.
This way people can look others up to see their approximate skill level.
To either have as criteria for tournaments or to get some well balanced games based on skill level.

Dark Riku wrote:I'd still go for the always ranked option, unless the unranked keyword is said at the start.

The problem with that is having people rated without their knowledge (possibly not even knowing that the rating system exists).

Btw, does the system right now take into account how highly your opponent is rated? So if someone with 1100 wins against someone with 1700 he gets lets say 200 points. But if he loses, he only loses 5 points? I would think so because thats the whole point isn't it?^^

But from what I could see on the matchlog I am not sure. It rather seems as if that has no influence and the amount you get only depends on how sure the system is of your rating.

For example King Kay played a ton of games (good job btw!) and losing against Browser (1860) https://www.dawnofwar.info/esl/log?page=matches&id=1149 he lost 26 points. The same amount he lost/gained when playing against someone completly new (1500) and when playing against -insert Coin- (1100) https://www.dawnofwar.info/esl/match/1137


Second thing:
I don't think the system actually uses the Credibility rating. I would have expected that you would get/lose less points when playing against someome with a high variance rating. But this wouldn't matter anyway when you get/lose the same amount of points no matter how high the enemy is rated.


Wouldn't you easily rise to the top of the ladder if you only played against weaker players with the way it is right now?
Thanks!

The result depends on both skill rating and ratings deviance, of both players. You can see the equations here: https://en.m.wikipedia.org/wiki/Glicko_ ... New_Rating

Lulgrim wrote:The result depends on both skill rating and ratings deviance, of both players. You can see the equations here: https://en.m.wikipedia.org/wiki/Glicko_ ... New_Rating

Yes thats what I thought. But I am not sure if it is applied correctly in our App.
Maybe I am just missinterpreting the display?

Could it be that the +- number displayed in the log is not correct? Maybe a number before some calculations are applied?
So for example: With this game you would gain 124 points, but because of the enemys Rating you actually only get 18 points?

Yeah, I think thats it. Because with that in mind the rest looks correct.

Edit: Oh, now I get it. I totally misinterpreted that number Its not how much you gain/lose for a game, but what your RD rating is at that moment^^
But I would like to see the number I first thought it would be^^

I could add gain/loss numbers if there is interest.

This might be a bit of a chore...

But now that the data is flooding in, I was wondering if we can finally compile some detailed matchup stats such as, for example, Tech marine win% vs Chaos Lord, and that sort of thing.

It’s certainly doable as database queries and Highcharts options go, but is the data sufficient? For example, we have 18 matchups with TM vs. CL:

518: 1794±184 vs 1400±263
596: 1786±108 vs 1769±188
597: 1696±167 vs 1811±104
598: 1652±154 vs 1830±101
637: 1568±50 vs 1852±97
645: 1738±73 vs 1574±94
646: 1632±140 vs 1607±91
647: 1629±89 vs 1581±131
649: 1736±128 vs 1621±84
650: 1682±122 vs 1644±82
651: 1715±115 vs 1629±80
652: 1673±110 vs 1649±79
653: 1702±105 vs 1634±77
655: 1725±101 vs 1621±75
941: 1796±105 vs 2044±88
1079: 1737±49 vs 1764±82
1110: 1767±78 vs 1731±49
1112: 1749±77 vs 1738±49
1113: 1764±75 vs 1732±48

and 4 matchups with Apo vs. FS:

306: 1280±217 vs 1551±127
362: 1365±249 vs 1876±164
402: 1353±205 vs 1500±350
819: 1804±83 vs 1873±52

which is not a whole lot to build statistics from, even without filtering them down to rating-balanced matchups.

That kind of statistic would be something that is built up over time. You could consider separating those win ratios by patch, but other than that yeah. Just wait :/

I'd be in favor of implementing that so its configured and available, and leave it up to the playerbase to populate it with enough data that it becomes meaningful soonish

Either way I think it would be really cool to have that kind of data out there to look at.

Yeah you're right, the specifics of the data really break it down into small packets, and thus make it less reliable. It is nice to know and be able to see it for sure.

Going further, I really like how you laid out the skill levels of each match as well. Helps to decide which of the match stats can be seen as credible.

Perhaps at some stage it can be worked into menu options for games whether it's ranked or unranked.

Most people aren't disclosing whether they're running it or not.

I don't do ranked games but I had a look since I was in the tournament and there are two ranked games against crewfinity which shouldn't be counted.

http://www.filedropper.com/warnings

This time it recorded the match on ESL, but not on the site.

Note: The first match was played before I opened ESL, so that one shouldn't have recorded.

I play like 15 games a day. Out of those 15, like 3 are usually blank games. But, have never had this happen before, though.

Thanks tranca, I've updated the reporter with a fix. Found no issue in the second log though, I think it might just have been a connection problem. The new version will retry any failed uploads instead of giving up after 1 try. You can also trigger an update manually with F5 now.

https://dawnofwar.info/index.php?page=download

Niiiice. Thank you thank you

Gonna put this to use right now. lol