No. Nice 38th post BTW. You must have the whole forum sussed after one week of membership. :lol: Nothing you think is statistically significant. Keyboard warrior much.
How many times are you going to rewrite this post ? The first one was telling me that nothing I said mattered because I was an America, what's this your third draft ? I notice after how ever many thousands of posts you've made here nobody is rushing to your defense, wonder why that is ? Maybe they're all afraid you'll rage on them ? You're like that guy who does something crazy and all his neighbors get on television and say "He seemed like such a quiet guy ..."
I once had a go (more like a challenge) at wrcmad over the Meier metal sample analysis because he is a metalurgylist, he did not have a go at me back. What you say there about him is not true.
I don't need people rushing to my defence. I can stand on my own two feet, and if what I say is wrong, then I'll admit it. I also give credit where credit is due, even on differing views to my own. However, your suggestion that no one is rushing to my defence is interesting. If you are wondering why, they are probably still trying to figure out what your attack is about. In fact, I bet you don't even know. :lol: All I did was point out an obvious fabrication regarding std deviation, and you decided to fling shit out of nowhere. It had nothing to do with you, and you call me a troll! :lol: What did I do crazy? I expect.... {crickets}
what will be the best way, I will only be buying 2016 koalas next year? suggestions most welcome thanks
In response to your offer to teach me a thing or two about stats here: http://forums.silverstackers.com/message-854830.html#p854830 Please review the following: I referred to your reference to percentage price being a red flag, because std dev of daily price movement is not measured in fixed %price. Why can't you use the normal bell-curve percentages? - because the normal distribution is composed of a population of data values, and a single price is not a population, but a mere single number. Std dev of daily price moves are dynamic and non-stationary, meaning that the underlying distribution from which prices are "drawn" shifts with time. Additionally, if you use all historical data as a data set, the effects of averaging and inflation over the years of data ensure your std dev calculation will not be representative of current prices or volatility. So it does not make any sense to reference std dev from a price unless that price is the mean or some other statistic for a given time period. Thus, std dev of price movements are not simply referenced to the normal bell-curve percentages. So, how is SD calculated for daily price moves? First, you have to nominate a time range for which to populate the data set. Standard practice is to then look at the standard deviation of price changes, instead of raw price. This is achieved by using the price change for each period in the nominated time-frame, which is merely the difference of the closing price for each period and the closing price of the previous period. In effect you are simply creating Bollinger Bands. Here is an example for a daily price chart of silver, using a 100-day timeframe. You can see the moving average price (dotted) with one standard deviation of price change above (MA + std dev) and below (MA - std dev) the line. The bar circled in blue is the day in question. It is evident from this chart that price did not move one std dev. It is also evident that std dev is not measured in % price. I look forward to your critique, and subsequent stats lesson.
Mmmmm.....??? Yes, I know..... that's what I said. :/ Rather than a cut'n'paste link that virtually repeats my erroneous approach to std dev, I was more hoping you could clarify the "really silly" bit that I falsely believed, and that cemented your opinion of me? Also, I still don't see the significance of a 5% price move, nor the one std dev move? :/ Any help would be much appreciated. Regards
mmm... shiney and Screaming Eagle. Learn for yourselves before thanking the guy. The link was for those on the forum, not wrcmad.
Miloman, I'm quite comfortable in my knowledge of statistics and also quite comfortable thanking wrcmad for his post. I greatly respect your knowledge when it comes to grading coins and about a number of other topics, however I appreciate the methodology that wrcmad presents when looking to calculate the SD of daily price moves, as I believe it to be 100% correct in accordance with the knowledge that I have on the subject, whilst being presented in simple terms. You could simply rebut mmm...shiney and wrcmad by presenting the time range you used and the subsequent calculations that you performed to come up with the assertion that the 5% price movement was greater than 1 SD. That's all I have to say on the matter.
