Get Time
Search | Watch Thread  |  My Post History  |  My Watches  |  User Settings
View: Flat (newest first)  | Threaded  | Tree
Previous Thread  |  Next Thread
[ 1 2 3 4 ]    NEXT >
[stat] micromatch score | Reply
Here go the micromatch scores (number of times one member is placed above the other when they both participate) for the members with >=4 SRM wins.

Feel free to suggest other pairs/sets of members :)

handleSnapDragontomekPetrdmwrightNGBronsonradeyeZorbaTHutYarinmisofmalptEryxjonmacJohn Dethridgeplohreidkalinovnicka81bladerunnermarek.cygan
John Dethridge20/3812/3012/2418/135/944/2029/2227/1817/2121/528/2016/344/2222/1945/846/1223/1039/11
Re: [stat] micromatch score (response to post by Petr) | Reply
What about hosting a web application or a web service where you can make your own queries? :)
Re: [stat] micromatch score (response to post by Petr) | Reply
Where to download TC data feeds?

Can we use this mini-matches data (For all members) in order to make more accurate odds for TCCC or TC*? Any suggestions?

By the way, 0/0 is undetermined, and that means that mathematics works :)
Re: [stat] micromatch score (response to post by Yarin) | Reply
Why not? Let's hope it won't be abused...

EDIT. all your queries are logged (both for my fun and for safety reasons)
Re: [stat] micromatch score (response to post by mohamedafattah) | Reply
Re: [stat] micromatch score (response to post by Petr) | Reply
Does this hold across divisions
i.e. if applicable, who wins if A scores 300 in div2 while B scores 250 in div1?
Re: [stat] micromatch score (response to post by kurtrips) | Reply
Good point. Now it doesn't :)
Re: [stat] micromatch score (response to post by Petr) | Reply
The best time-killer I've seen in a few months :)
Re: [stat] micromatch score (response to post by Petr) | Reply
Someone should find the largest cycle (NP-complete? bah) in the "beats" graph. A beats B if A has a greater than 50% winning percentage. For example, tomek beats Petr beats SnapDragon beats tomek.
Re: [stat] micromatch score (response to post by rrenaud) | Reply
I think it will be very very large, because of people that have only one shared match and general volatility.
Re: [stat] micromatch score (response to post by Petr) | Reply
What if you restrict it to say people who have played more than 10 games matches against each other.
Re: [stat] micromatch score (response to post by Petr) | Reply
It would be cool if it enumerated the matches that you shared. It would be even cooler if it showed the times/scores/challenges during the match.
Re: [stat] micromatch score (response to post by rrenaud) | Reply
If there's enough support for that, I can do it. Let's wait for the '+'s and '-'s :)
Re: [stat] micromatch score (response to post by Petr) | Reply
Awesome. Its quite humbling to see just how 0wned I am by the top 10!

Tried searching down the ranklist to see how far I had to go down before being ahead of anyone- got to 13th for parity and 22nd for a lead.

One interesting thing would be to have the expected result displayed as well (where you calc that from curr ratings,vols, and assume same total number of matches.)

Other question is, what do you do about ties? If two people both score zero....

Anyway, great tool, thanks so much for putting it up.
Re: [stat] micromatch score (response to post by PaulJefferys) | Reply
Ties are not counted. That's why your score with yourself is 0/0.
[ 1 2 3 4 ]    NEXT >