Match Making Rating Tool - Page 113

On January 04 2013 11:11 InfCereal wrote:



There're some posts to read regarding this tool and the mmr system, do you remember the names of the posts so I could read them?

If you're giving the project away, I'd be more than willing to upkeep it. Lots of free time here.

On January 04 2013 13:18 NoFool wrote:
2:16:12 DATA Alg.B: failed, No Bonus Win 2:16:12
DATA Alg. A : failed, No Ramscanner 2:16:12 ERROR
AFTERCHECK Change: 0 or 1

What dat means ?

and another game:
2:34:10 ERROR PRECHECK PLAYER: 0 Ladderpoints - No Data

On January 04 2013 14:05 Khorik wrote:
What exactly are the criteria for a 100% game? I've played over 20 games so far and it still lacks sufficient data. They've all been 50 or 75% games so far.

On January 04 2013 15:41 Chillyo wrote:
All games being played now are being marked as bad games, so you won't get good data until the new db comes out, from what I understand.

Does that mean you won't be re-calculating them?

EDIT: Nvm didn't see your new post. Can I delete posts here?

you cant.
Normaly you edit text out if you want to delete.
Last post is very optimistic. i will calculate them if i have the time and the the feeling...
at the moment i dont have it,
Will give this project away soon,

There're some posts to read regarding this tool and the mmr system, do you remember the names of the posts so I could read them?

If you're giving the project away, I'd be more than willing to upkeep it. Lots of free time here.

thought about you.
But you are not a programmer are you?
I thought about a team or something with MMR guys and program guys.
Nts old threads is linked in the op under theories.

On January 04 2013 21:01 skeldark wrote:
The problem with the team is:
1) diffrent timezones and free time.
2) personal problem
3) people quit fast...

There are so many people who want to do it and there is no way for me to tell who will really do it.
I thought about letting you guys handle it yourself, instead of hard pick people.

Perhaps make a list and meet on TS and figure it out...

On January 04 2013 21:12 leopardb wrote:


I'm afraid it's not gonna work this way. Too chaotic. As the creator of the programm you must choose a son :D but firstly, i think you should give some guiding lines.

1) Which programming environment do you use for this project (if you use one) ?
2) Could you publish the source, so that i or another could already compile the bytecodes once (to set up a working environment)
3) Somebody should know something about the MMR, i've got the feeling some people here know already about it
4) When it begins to move (when people engage into it) we should make a separate thread for development and keep this one only for questions/bug reports.

Open to questions, but we need you to begin skeldark

First thing first who can/wants to help ? let's begin this list...

EDIT: also we have to think/know how the things are gonna work with HOTS, after all and if Blizzard respects its own timelines, it's due in March2013, so in 3 months, meaning this tool is gonna be obsolete quite soon (i don't think that so many players are going to stay in WoL when HOTS is out).

On January 04 2013 22:56 Mavvie wrote:
Does the memory scanner act similar to an external maphack?

Obviously it doesn't read map data while in game since it only looks for bonus pool/other profile data, but is it the same concept? I always wondered how these work...how do you know what memory address to read?

On January 04 2013 20:38 skeldark wrote:

1) you have memory scanner disabled and one of the players had no bonuspoints = No calculataion
Turn on memory scanner

2) No ladderpoints = no data , a mistake that is often for obvious reasons at the start of a season




To complex to say in a quick post.
There is a big diffrence between 50 and 75.
75 are uncontrollable mistakes. (the data is collected but dont hit the criteria to calculate mmr)
50 means a mistake in the collection process.
Turn on memory reader to reduce the 50% mistakes.



Season 11 DB is up.
Only china data for 11 is still missing,.

On January 04 2013 23:21 skeldark wrote:

It reads the nickname of your opponent from the sc2 memory.
I dont change anything, so sc2 is unable to detect it in the frist place.
Even if they detect it none of my 40k users good banned for using it.
IS it against the TOS? yes but installing sc2 is against the sc2 TOS, because its written so vague.


I search the memory for the oppoenent name after path. remember the adress and thats it.
Address relative to the program ram offsets dont change.

On January 05 2013 02:51 Mavvie wrote:

Oh yes I know it doesn't change anything, hence "external maphack"; the type that just reads data and doesn't change any of it.

I wasn't questioning whether it's good to use, I was just curious. I do a lot of programming but was unaware; I never knew that it would be so constant, even if graphics settings are different and stuff.

Is there any way for blizzard to prevent this? I thought ASLR or something similar would prevent it; it would completely deny any maphacking, but I guess it has its cons.

On January 05 2013 02:56 skeldark wrote:



You can not detect any memory reader. Even not maphack.
the only way to do it, is to search on the computer for an process that reads memory of other processes.
And this would be against the law...

You can never prevent memory reading. When you can read it , windows can read it, all other process can read it.
Not the program decide how the memory is handled the main system does that.
And if the main system ( here windows) allowes me to read the memory of program A, there is no way for A to deny it or to detect it.
Settings are a memory value not a memory structure. The program is the structure. So only a change in the program ( patch) change the address. (you could change the structure in runtime to make it harder to read but they dont do it)
Would not be worth it anyways because i could counter it, and would slow the program down.