» Sat Nov 17, 2012 1:11 pm
Hello again,
This is continuing from my last post ([Edit: So, im not allowed to post links as it turns out, my last link was pointing to a post on the last page of part 12 for this thread] ) where I posted data showing how long in days we have had to wait to see Bethesda DLC appear on all platforms after its initial release for xbox. The data set included the DLC's for Fallout 3, Fallout New Vegas and one DLC for Oblivion. Before moving on its important to note that the anolysis given on the previous post was a quick and dirty anolysis just meant to get the ball rolling for other people who are more studied in statistics, economics, and psychology than I. This post isn't much different but goes a little further with the quantitative anolysis. The point of showing the averages with the standard deviations was to loosely test the hypothesis that the 'time to wait' for all platform relase for each DLC was relatively the same to all other DLC's 'time to wait.' From the large std. dev relative to the average we can begin to suspect that this isn't the case. Furthermore, from plotting the data we can see a overall trend that the 'time to wait' has been decreasing for each subsequent DLC released by Bethesda (Yay, evidence that Bethesda may be 'listening' to its disgruntled fan base). So, for this anolysis I simply tried fitting the data set with the ommited data points (see link[now edit] above for the original post) to a linear model. The statistical summary of the fit:
Residuals:
1 2 3 4 5 6 7
-34.821 39.214 14.250 -3.714 -7.679 -17.643 10.393
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 269.857 22.186 12.164 6.64e-05 ***
x -31.036 4.961 -6.256 0.00153 **
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 26.25 on 5 degrees of freedom
Multiple R-squared: 0.8867, Adjusted R-squared: 0.8641
F-statistic: 39.14 on 1 and 5 DF, p-value: 0.00153
Sadly, the forums won't let me post a nice pretty graph (image) for everyone to see. So we get a linear model of ['time to wait' = -31.036*x +269.857] where 'time to wait' is the time in days we have to wait before the DLC is released on all platforms after the initial release on xbox and x is the iteration of DLC released that has had a Microsoft-funded delay.
so for the next DLC (Dawnguard) we would have a value of x = 8 and taking into account the residual std. error, giving us:
'time to wait' = approx. 47.822 days = -31.036*8+269.86+26.25
or, in other words a release date around 8/12/2012 (m/d/y) for Dawnguard DLC on all other platforms.
Again, I don't think a linear model is an accurate description of the data but this anolysis gives a different way of interpreting the data with a different possible prediction for the release date. This isn't meant to be a 'serious' attempt but to be something slightly better than guesswork and more quantitative and logical in nature.
Thank you for your time,
Yuki