estimating the next time rank will increase

djrbk · Post by **djrbk** » Sun Aug 23, 2009 7:42 am

what's a gatt?

is it like an internceptor?

omegahero · Post by **omegahero** » Sun Aug 23, 2009 8:12 am

djrbk wrote:QUOTE (djrbk @ Aug 23 2009, 05:42 PM) what's a gatt?

is it like an internceptor?

hmm...

i see what you're trying to do there...

badpazzword · Post by **badpazzword** » Sun Aug 23, 2009 10:58 am

The blue curve in my graph is not actually a fitting. I could find no satisfying fit for that plot (I didn't try a lot tbh). The blue line is just a 'smoothed' line of the sample, some sort of moving average.

Remember, game size plays a big part in gaining ranks. The smaller the game, the more the game outcome depends on you (statistically), the larger the updates you get.

MrChaos · Post by **MrChaos** » Sun Aug 23, 2009 1:47 pm

Badp wrote:QUOTE (Badp @ Aug 23 2009, 05:58 AM) The blue curve in my graph is not actually a fitting. I could find no satisfying fit for that plot (I didn't try a lot tbh). The blue line is just a 'smoothed' line of the sample, some sort of moving average.

Remember, game size plays a big part in gaining ranks. The smaller the game, the more the game outcome depends on you (statistically), the larger the updates you get.

Try y=Zx^n as a potential curve fit. Try scatter plot and add trendline (right click) using Excel. There are any number of possible ways to fit the curve and it also computes the error and equation for you.

MrChaos

notjarvis · Post by **notjarvis** » Mon Aug 24, 2009 8:31 am

fwiffo wrote:QUOTE (fwiffo @ Aug 23 2009, 04:00 AM) instead, we can keep track of things like "this month's leading drone killer" which is not only easier to do, but a heck of a lot more accurate!

I like this idea and would find it quite fun. You could post the Leading Drone Killer per month and other stats every so often

Other suggestions I thought of...

Best win/lose ratio of the past month?Best Commander of the Past Month?Biggest Whore?Podrider of the month (may have to filter out new players to not think we're picking on them)?

could be fun to see on the forums....

Gandalf2 · Post by **Gandalf2** » Tue Aug 25, 2009 12:59 am

I wonder if there a correlation between sigma and stack rating (for players who have played, say, >500 hours)?

MrChaos · Post by **MrChaos** » Tue Aug 25, 2009 2:27 am

Gandalf2 wrote:QUOTE (Gandalf2 @ Aug 24 2009, 07:59 PM) I wonder if there a correlation between sigma and stack rating (for players who have played, say, >500 hours)?

none

Necrotic · Post by **Necrotic** » Wed Aug 26, 2009 12:04 am

I graduated from engineering....I dont need to worry about linear regression ever again...its just annoying...

MrChaos · Post by **MrChaos** » Wed Aug 26, 2009 1:35 am

Necrotic wrote:QUOTE (Necrotic @ Aug 25 2009, 07:04 PM) I graduated from engineering....I dont need to worry about linear regression ever again...its just annoying...

oh yes you will

MrChaos <--- engineer too

edit: maybe

(black belts use it for example, and Ive used it as a development engineer when doing DOEs to find the root cause of a particualr issue a number of times)

Necrotic · Post by **Necrotic** » Wed Aug 26, 2009 2:36 am

MrChaos wrote:QUOTE (MrChaos @ Aug 25 2009, 09:35 PM) oh yes you will

MrChaos <--- engineer too

edit: maybe (black belts use it for example, and Ive used it as a development engineer when doing DOEs to find the root cause of a particualr issue a number of times)

Ya well...im making special point of not choosing a profession that requires me to use that.
<< Mechanical Engineer, Georgia Tech...you?