General Discussion - Page 4474

On December 06 2013 05:33 Klonere wrote:


its good for this as well

On December 06 2013 05:36 hariooo wrote:


Bulldog queues pubs with Sheever a lot. This particular example isn't too surprising.

On December 06 2013 05:39 SKC wrote:

The fact he queues with someone doesn't really change how good of a player he is.

On December 06 2013 05:39 hariooo wrote:
http://www.reddit.com/r/DotA2/comments/1s6ifu/so_the_author_of_mydotaskillcom_made_his_code/

I expect you all to turn into CS majors now.

On December 06 2013 05:45 hariooo wrote:


It means the average MMR of all players in those games are going to be lower than if he was solo q'ing. DSR is to some significant extent (from the creator's mouth) based off that value. I can't imagine Bulldog going EE-levels of tryhard in those kinds of games either, so it shouldn't shock anyone that he might have a lower DSR compared to pubstars.

On December 06 2013 05:47 SKC wrote:

2GD is a pubstar? Just because Admiral Bulldogs MMR is lower than it could be doesn't mean he should be that low. They also don't queue exclusivelly with each other.

On December 06 2013 05:47 SKC wrote:

2GD is a pubstar? Just because Admiral Bulldogs MMR is lower than it could be doesn't mean he should be that low. They also don't queue exclusivelly with each other.

2GD has even talked about how his games are very diferent when he doesn't stack with Draskyl. His MMR isn't that high at all.

On December 06 2013 05:53 hariooo wrote:


Right but MMR and DSR are different measures and not directly comparable as far as I know. If 2GD frequently stacks with people of higher MMR than him and Bulldog frequently stacks with people of lower MMR than him, this is an explanation for the difference in DSR's.

On December 06 2013 05:52 SKC wrote:
When he says it's better than Valve's system, you also have to account how that would affect matchmaking. Such as playing heroes in not standard ways and padding stats. Focusing too much on KDA is not a good idea.

Valve had that system that tried to compare your performance with other players with each heroes, and we all remember how accurate it was.

On December 06 2013 05:55 Sn0_Man wrote:
That might work in LoL where every hero is literally designed to be played exactly 1 way, but its completely irrelevant in Dota. Take mirana. Sing_Sing plays mirana middle as a dynamic playmaker and secondary farmer. Fluffnstuff plays Mirana as a secondary support, providing disable, damage, and a global invis ulti as well as some team mobility and ganks. XBOCT plays Mirana as a primary carry, gathering safe lane farm and looking to go into the lategame if he must.

The GPMs, KDAs and everything else are WILDLY different for all those players. Are you going to say that fluffnstuff is a 200DSR player because he has like 150gpm average on mirana when XBOCT has like 450 average?

Its literally retarded to try and compare people based on anything to do with KDA or GPM, especially when it is hero specific.

On December 06 2013 05:50 Milkis wrote:
I don't really want to go through all that PHP but it seems like the general idea was to train a bunch of data to get stuff like kda/xpm/gpm and such for every hero and judged you by the performance of those statistics per hero. That would be my guess. This is why he can get away with "20 games" for that reason.

I think such a methodology can work very well to rate players but I don't think how he's accomplishing is the best way ever. There's a lot of issues with such a system too that I wonder if it was addressed (I don't really want to read that code). But it's really not difficult to think which ones work or not.

One more thing is that every rating system has anomalies -- I mean, not everyone tryhard ladders every game and I think things like that affect your rating probably.

On December 06 2013 05:58 devilesk wrote:


No, he might train with all that data, but ultimately the question he's trying to solve with his model is whether a game is "pro-level" or "normal". From his training, he can "graph" a surface that represents the baseline between pro and normal. Anything below the surface is normal and anything above it is pro. He's using the distance from that surface as the metric for generating his rating, which is I believe a point of contention for people knowledgeable about SVM (I am not), because it seems like the model's simple purpose is just to make that binary decision.