Data Official Smogon University Usage Statistics Discussion Thread, mk.2

Status
Not open for further replies.

Antar

is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
Official Data Miner
But shouldn't everything be a whole number for OU-0?
Believe it or not, there are some players with Glicko ratings so low, they don't have a 1.0 weight, even with the 0 cutoff. 0 is "roughly" unweighted, but maybe in the future I'll do *actually* unweighted. That being said, where do you see non-integers?

Regarding the discrepancy between the usage stats and detailed stats, this is due to the fact that the usage stats remove all battles with less than six turns while the detailed stats do not. This is an oversight, and I'll correct it for the next month.
 
Imran, the values are the sum of the weights.
Thanks, additionally, are the lead stats generated from something other than the chaos file? I can't see anything in it that suggests that you can derive the lead stats.

I would suggest the same is true of Mega forms and your metagame analysis, but am less confident asserting those. I assume there's no further public data source at this stage that gives access to information like that?
 
Last edited:

Antar

is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
Official Data Miner
Why is Battle Spot, 1v1 and GBU weighted above 1?
I'll have to see what's going on there.

Thanks, additionally, are the lead stats generated from something other than the chaos file? I can't see anything in it that suggests that you can derive the lead stats.

I would suggest the same is true of Mega forms and your metagame analysis, but am less confident asserting those. I assume there's no further public data source at this stage that gives access to information like that?
Mega analysis is easily derived from the chaos files (mega stone incidence percentage x usage). The rest are not.
 
Thanks for the stats Antar!
I was wondering, will we ever get back the "changes folder" like we used to before Showdown had problems back in november? On OU this month there are some incredible changes, like for example Garchomp and Bisharp, it would be very useful to analyze precisely how much certain 'mons gained or lost usage and better understand metagame's trends.
 
Last edited:
I was looking through OU's April 1825 stats and found this EV spread for Aggron:

Adamant:96/0/0/252/0/20 3.064%

Somehow that spread seems unlikely. Is there an issue with how usage stats are counted that would explain this? Did a high level player put that in by accident and laddered with it?
 

Antar

is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
Official Data Miner
Toljik,

April OU-1825 stats said:
| 141 | Aggron | 0.15630% |
April OU-1825 moveset stats said:
+----------------------------------------+
| Aggron |
+----------------------------------------+
| Raw count: 29116 |
| Avg. weight: 0.00055950187863 |
+----------------------------------------+
So we're dealing with really small numbers here. But I did check--it was one player with Glicko rating in the high 1700s who used that Aggron build 15 times.
 

Antar

is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
Official Data Miner
Too Bad hackmons is still to be missed.
What do you mean? BH stats are up, anf Hackmons itself hasn't had a ladder in months.
And om of the month are shared very late.
These stats went up at the same time as all the others...
Why did pg little cup get special treatment?
The PGL Little Cup ladder went up largely to facilitate teambuilding for the event. I posted interrim stats early so that players could use them to prepare, considering the competition ended before the end of the month.
 
What do you mean? BH stats are up, anf Hackmons itself hasn't had a ladder in months.

These stats went up at the same time as all the others...

The PGL Little Cup ladder went up largely to facilitate teambuilding for the event. I posted interrim stats early so that players could use them to prepare, considering the competition ended before the end of the month.
Yeah I was referring to Hackmons not Balanced hackmons. I love the statistics given in Balanced hackmons and greatly appreciate it. Everything I posted is not really a monthly problem but moreso of a yearly problem to me. OMotM are usually given in every month after it's completely over, which (for myself at least) ruins the use of it since the the tier's time is already over. It'll be great if statistics of these temporary tiers can be given before it's completely over. the PGL timing was great, it was used for an event that seemingly ends, same can be applied to OMotM's too.

Back to hackmons, I suppose you're right, I haven't noticed it was completely gone now, just always wanted to be prepared to view when the time came around. But during those times it was around, hackmons was completely dismissed along with Balanced hackmons during these statistics at one point, was a little unfair in my book. I apologize for expecting statistics on Hackmons when it's currently not even around. I haven't played much of the laddering but I always hang around in tournament chat to participate and greatly make use of each tier by using these statistics to make a good time, unfortunately when OMotM's are hosted I would get hammered with little knowledge to refer to (like just now with Linked). It's also a bit of a disappointment that statistics from past years suddenly just vanished, would have been a decent reference to go back to.
 

Antar

is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
Official Data Miner
But during those times it was around, hackmons was completely dismissed along with Balanced hackmons during these statistics at one point, was a little unfair in my book.
Hackmons/BH didn't get usage stats during that period because the tiers weren't compatible with my analysis scripts--the fact that you could use Megas without stones broke a bunch of things. By the time I had time to write a workaround (keep in mind: I have a full-time job, a wife and kid, so it's not like I have an overabundance of free time), Hackmons had been discontinued.
 
Theoretically, would a Pokemon that gains over ~3.41% usage over the three months in a tier two spots over its current tiers (such as UU is to NU), but not enough usage in the tier directly above it, jump to the tier it gained the ~3.41%, or the tier directly above? Or would it not move at all?
 
Status
Not open for further replies.

Users Who Are Viewing This Thread (Users: 1, Guests: 1)