Talk:Deadeye

I know there's a lot of silly stuff going around relating to deadeye, and I, like Kylet, am a Denizen, I get most of my information from the Den Kirson Proboards. The recoil plots for Ghosts on Symthic are generated by Probaddie, and the gun stats on Symthic's Ghosts portion are similarily taken from the game code and decyphered by Marvel4. Drift0r, as well, commonly comes into our IRC chat asking Marvel4 and Probaddie and Mousey and Corpsecreate questions, and cites us in his videos as where he gets the information, I can't say the same about Tmartn, but. . . Anyways, looking through the edits, I want to clarify something.

Crazy sam10 stated "Game code overtakes guidebook which is commonly flawed," which is absolutely true, and I completely agree with that assessment, however, unless I am mistaken, the source that Crazy sam10 was referring to was not citing the game code, but rather an empirical test. See, in Drift0r's video description, he cites http://denkirson.proboards.com/thread/6584/deadeye-damage-graph, and if you actually read the thread, you see that it is not based on game code, but instead based on these empirical tests: "Building on the work done by Mousey in  this thread, I performed an additional analysis on the Deadeye perk and its tangible benefit on bullets to kill.  From what I've gathered so far: Now, it seems consistent that if the ultimate reference which cannot be disputed is game code, then such a test would be insufficient basis to cast sufficient doubt on the earlier reference provided by Marvel4, who is the guy who goes through the game files and deciphers the mess to provide the vast majority of us with the gun stats, who I have personal contact with via this Den Kirson IRC.
 * Deadeye adds 1.4 multiplier to base weapon damage
 * Each bullet is an independent event (i.e. independent random variable)
 * Probability of benefit increases with number of kills (12.5%, 16.67%, 33.33%, 40% and 50%)
 * Probability of benefit reception follows binomial distribution"

Notice, further, the differences in the percentages. 12.5, 16.67, 33.3, vs 10, 20, 30. Now, these percentages are exceedingly close, and here is the concern that we have not resolved yet; Broman55, the publisher of the damage graph, and Mousey, who did the work that Broman55's is based on, did not post trial data, that is, they did not show the number of samples they did, the amount of testing so to say, and as we all know, testing of probablistic nature often has a large margin of error with a low enough amount of tests, and even rarely when there are a large amount of tests. Talking with these people via the IRC, I have drawn the conclusion that all the data thus far is inconclusive. Marvel4 edited the Deadeye page with the official strategy guide information: http://denkirson.proboards.com/thread/6734/deadeye-info-official-strategy-guide, because he felt that the numbers were close enough to account for the margin of error, and the peculiar nature of the guidebook this time around being far more in-depth and demonstratably accurate than previous entries in the guidebook series (I know this is anecdotal evidence, but I pray by now I have your trust)

Don't get me wrong, I'm not advocating that Marvel4 is always right, and in fact in this case I would venture so far as to say he is not speaking from sufficient ground to dispute the claims made above. Yet the claims made above themselves are in question, and no information we have thus far is based on the game code. Therefore, I propose, the wiki should not publish either of these potentially flawed sources, if anything we should publish both with a section that enumerates their disagreements and establishes the ethos between the two. Further, we should engage in a very, very large scale test, with upwards of 10,000 trials at each Deadeye level to get a more reliable statistic, until the actual game files regarding the particular effects of Deadeye are found and decyphered. Jaedrik (talk) 21:33, November 21, 2013 (UTC)