Since Brass Man made it sound like unbreaking the old site will take a while and Kevin asked, here is basic data from the January and Februrary P9 challenges. Matt is working on doing an archetype vs. archetype breakdown. In the meantime I thought I'd post what we have readily available. I think the only reclassifying we did was to move gush storm from gush to combo, but I believe that was only played by one person so the effect should not matter much.

January Power 9 Challenge:
Total Number of Players 104
Top 16 Decklists: http://magic.wizards.com/en/articles/archive/mtgo-standings/vintage-premier-2016-01-31

Shops (ravager, aggro, tiny robots, uba stax, terra nova, martello)
Number of Players 21
Win Percentage 49.6%

Gush (mentor, delver, gush storm)
Number of Players 15
Win Percentage 58.2%
Gush Win Percentage w/o Mirrors: 59.1%

Big Blue (painter, tezz, BUG control, etc.)
Number of Players 8
Win Percentage 53%

Dredge (traditional, pitch, etc.)
Number of Players 16
Win Percentage 52.1%

Oath (odd, Brian Kelly, Fenton, DP)
Number of Players 10
Win Percentage 42.6%

Combo (DPS)
Number of Players 16
Win Percentage 51.1%

Februrary Power 9 Challenge:
Total Number of Players 83
Top 16 Decklists: http://magic.wizards.com/en/articles/archive/mtgo-standings/vintage-premier-2016-02-29

Shops (ravager, aggro, tiny robots, stax, terra nova)
Number of Players 19
Win Percentage 61.2%
Shops Win Percentage w/o Mirrors: 66.3%

Gush (mentor, delver)
Number of Players 17
Win Percentage 51.4%

TFK Blue (painter, tezz, 3C control, academy)
Number of Players 5
Win Percentage 40%

Dredge (traditional, pitch)
Number of Players 12
Win Percentage 48.4%

Oath (odd, Brian Kelly, Fenton, Oathstill)
Number of Players 7
Win Percentage 37.1%

Combo (DPS, belcher, rector, doomsday)
Number of Players 18
Win Percentage 44%

Thank you!

Thanks @diophan ! Podcasting tonight and this data is invaluable.

Shops had a 49.6% win rate? Thank God we got that taken care of.

@The-Atog-Lord said:

Shops had a 49.6% win rate? Thank God we got that taken care of.

Vintage, so coin flippy

@The-Atog-Lord That was in January. The only relevant data is February, the month in which modo programmers begin writing their quadruple nested for loops to ensure that the restricted list will be implemented correctly.

As though they understand loops...

Here is the matchup breakdown for January (Thank you for the help, Ryan. These things are a pain).

When drawing conclusions from this, be sure to look at the sample size: (# of matches = # of wins for deck X against Y + # of wins for deck Y against X) (Match win % = # of wins for deck x / # of matches).

Well it seems like if combo-summer comes, i'll just start playing "other"

Here is the matchup vs matchup data for February. Same caveats apply.