Smogon Shoddy Server Statistics - July 2008

[BAD WOLF]

Doug, if you don't mind me asking, when you're parsing the logs do they contain information about moves, EVs, hold items etc? Because you show for example how many Infernapes run max speed EVs. Was this information readily available in the logs, or were calculations required? Same thing applies to Pokemon who didn't use all four of their moves. Can it be shown that these move lists for pokemon used in July are accurate just by looking through the logs?

Don't mind me asking these questions, I'd just love to know how you got all this information from battle logs because I'm really into statistics, I'm a kind of stat nerd if you will.

 

[animenagai]

i agree with the specs luke stuff. imo it is a bit like mix mence back when DDmence was the standard. specsluke i believe is the best luke set there is because there are so little things that can counter it (spiritomb is the only counter i see so far) much like how there are only maybe 2 counters to mixmence. even if everybody starts using specsluke now and gets rid of the surprise factor, i'll stand by that comment.

 

[Age of Kings]

It's interesting how Froslass, the most used UU in standard, has more usages on the standard ladder than in UU (despite being ranked a lot lower).
60 | Froslass | 4864
11 | Froslass | 3951
So perhaps this means that the smaller number of UU ladder players equals a far, far smaller amount of team variety (although I have no problem what Steelix at all). =/
Hang tight, I'm not the only one saying that next month's stats for UU will be even worse will all the shit they unbanned...

 

[Tracker]

| Heatran | Nature | Modest | 63.7 |
| Heatran | Nature | Timid | 11.1 |

| Heatran | Item | Choice Scarf | 52.2 |

I think we are about to see a rise in HP Ground Jolteon (and Aerodactyl) once people realize how good it is at countering Modest ScarfTran

 

[JabbaTheGriffin]

It's interesting how Froslass, the most used UU in standard, has more usages on the standard ladder than in UU (despite being ranked a lot lower).
60 | Froslass | 4864
11 | Froslass | 3951
So perhaps this means that the smaller number of UU ladder players equals a far, far smaller amount of team variety (although I have no problem what Steelix at all). =/
Hang tight, I'm not the only one saying that next month's stats for UU will be even worse will all the shit they unbanned...

1 | Garchomp | 83003
1 | Steelix | 8658

as you can see it's not really a reasonable comparison. if UU was played on the same scale that OU was then Froslass's usage would be closer to 40,000

 

[DougJustDoug]

Doug, if you don't mind me asking, when you're parsing the logs do they contain information about moves, EVs, hold items etc? Because you show for example how many Infernapes run max speed EVs. Was this information readily available in the logs, or were calculations required? Same thing applies to Pokemon who didn't use all four of their moves. Can it be shown that these move lists for pokemon used in July are accurate just by looking through the logs?

Don't mind me asking these questions, I'd just love to know how you got all this information from battle logs because I'm really into statistics, I'm a kind of stat nerd if you will.

The information about every pokemon in the battle is output to the log file (not the screen log) at the beginning of every battle. Colin made this change to the base Shoddy code about a month ago. Prior to that, the information was not readily available. With the text log, it is not terribly complicated programming to parse the top of the log and get all the information contained in the statistics I am reporting. Collating and reporting the stats, and handling file errors is more work than finding the information in the log. The server puts the information in the log for the express purpose of using it for stat collection.

I will be uploading my stat collection programs to the Shoddy SourceForge repository for other server admins to use, if they would like. If you want to inspect the source code and see how I am gathering the stats, you will be able to look at it there in the near future.

 

[Age of Kings]

1 | Garchomp | 83003
1 | Steelix | 8658

as you can see it's not really a reasonable comparison. if UU was played on the same scale that OU was then Froslass's usage would be closer to 40,000

I was going for definite proof of the statement "There is a far smaller number of people playing UU, a smaller variety in team makeup, and therefore what looks to be overcentralization". Hence why I didn't use a ratio so much as what is actually there.
Although I agree Froslass was a poor example....

 

[Stallion]

This is remarkable, thankyou for all your time and effort.

 

[Cathy]

In case anybody was curious, the stats from the official server are available for July as well. Although I haven't done it myself, a comparison might be interesting. We've also added a fancy trend grapher that allows you to see changes in usage over the last ten months, since we first launched the ladder.

[Edit: The size of OU is 48 on the Official Server for July, by the way.]

 

[the_artic_one]

I really like the neat exponential curve of Deoxys-E usage.

 

[dekzeh]

Surprises me how Jynx is almost not used on UU when it's near uncounterable there.

 

[Forsety]

I'm actually rather surprised Electrode has found a niche and is now one of the most used UU leads. At the same time I expect it to rise in popularity with Aerodactyl now joining UU as it makes a great Aero counter.

 

[phalanx]

Doesn't surprise me, its rediculous to counter as it sets up taunt and rain.

 

[bolt.]

Kinda figured it would look like this, in ou. I really want to see the suspect ladders statistics. I can't wait.

I think bronzong usage went down.