|
apetersson
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 08:31:50 AM |
|
don't be so harsh ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif) drawing trendlines is an art or do you complain to your barrista when he hands you his cafe latte with - last week the pattern was different! ![](https://ip.bitcointalk.org/?u=http%3A%2F%2Fwww.barista-show.de%2Ffiles%2Fdrache_latte.jpg&t=665&c=uMH_Y9vBIZFM_A)
|
|
|
|
fivebells
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 10:42:33 AM |
|
What do "MACD (26,122): -0.753", "EXP(9): -0.851", "Divergence: 0.098", "RSI(14): 34.72" mean? What is bullish about this divergence measurement?
|
|
|
|
Jixtreme
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 01:51:11 PM |
|
Trogdor says, "It's time to buy Bitcoins, kids!"
|
|
|
|
S3052 (OP)
Legendary
Offline
Activity: 2100
Merit: 1000
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 03:44:24 PM |
|
This is a good point. There are always slight variations on how to draw the lines, but the key point is that however you draw it, there has not been a substantial violation of the major uptrend.
Because if the major trend is broken, you see a rapid, crash-type decline with increasing volume. we clearly have not seen that and this is why the longterm uptrend has held.
|
|
|
|
zby
Legendary
Offline
Activity: 1592
Merit: 1001
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 03:45:11 PM |
|
don't be so harsh ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif) drawing trendlines is an art or do you complain to your barrista when he hands you his cafe latte with - last week the pattern was different! ![](https://ip.bitcointalk.org/?u=http%3A%2F%2Fwww.barista-show.de%2Ffiles%2Fdrache_latte.jpg&t=665&c=uMH_Y9vBIZFM_A) Are you saying that next time that trend line will be drown even lower?
|
|
|
|
S3052 (OP)
Legendary
Offline
Activity: 2100
Merit: 1000
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 03:56:48 PM |
|
Thanks for your feedback. Feedback is always a gift.
See attached the two most extreme trendlines
1) based on closing prices (pink)
2) based on intraday prices (red)
Both have not been broken clearly.
![](https://ip.bitcointalk.org/?u=http%3A%2F%2Fbit.ly%2FprvZtD&t=665&c=sgdHmyynC7aqxg)
|
|
|
|
fivebells
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 05:14:36 PM |
|
S3052, just repeating because my questions might have been obscured by the fact that they are at the bottom of the last page, below some images:
What do "MACD (26,122): -0.753", "EXP(9): -0.851", "Divergence: 0.098", "RSI(14): 34.72" mean? What is bullish about this divergence measurement?
|
|
|
|
|
fivebells
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 05:59:02 PM |
|
Thanks, netrin. If I understand correctly, in mid to late August there was an even stronger "bullish indication" by this reasoning, because the MAC(26,12) was above its 9-day EMA, and the price steadily dropped. It appears that S3052 did indeed read a bullish signal at that time, which must have been somewhat painful. All the MACD framework seems to indicate is that there was a recent reversal in a short-term trend (in this case, the bounce from ~$4.8 around Sep 17.) This does not seem very convincing to me. I haven't read the RSI wiki page, yet, but appreciate the pointer. I will look at it later.
|
|
|
|
speeder
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 06:37:07 PM |
|
August we had a sort of correction upwards, after a large dip.
|
|
|
|
osmosis
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 06:54:04 PM |
|
Thanks, netrin. If I understand correctly, in mid to late August there was an even stronger "bullish indication" by this reasoning, because the MAC(26,12) was above its 9-day EMA, and the price steadily dropped. It appears that S3052 did indeed read a bullish signal at that time, which must have been somewhat painful. All the MACD framework seems to indicate is that there was a recent reversal in a short-term trend (in this case, the bounce from ~$4.8 around Sep 17.) This does not seem very convincing to me. I haven't read the RSI wiki page, yet, but appreciate the pointer. I will look at it later. Any one signal can always give false positives, so its important to look for multiple signals. Additionally, its important to remember that there is always enough room left for reversal of the current signals. TA is a bit like listening to a sports commentator describing a live game. Never know when there is going to be an interception and trend reversal.
|
|
|
|
speeder
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 07:53:50 PM |
|
The signals, cannot preview for example if some crazy guy will dump 100k BTC or 1 million USD in BTC...
|
|
|
|
fivebells
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 07:58:50 PM |
|
Then what's it good for? Are there any objective, systematic analyses of the predictive powers of these indicators that I could read?
|
|
|
|
grod
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 08:00:08 PM |
|
The signals, cannot preview for example if some crazy guy will dump 100k BTC or 1 million USD in BTC...
No, but they help you estimate whether such an event is likely. And if so, in which direction. Crazy guys follow herds same as everyone else, and TA is a way of getting a sense of which way the herd is migrating. In other words, it won't help you predict when a kid that mined a few thousand coins in 2009 got a driver's license today and wants a car, but it should help you determine when a large portion of the current community has had enough and wants out. Also when they've run out of coins or money.
|
|
|
|
grod
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 08:02:33 PM |
|
Then what's it good for? Are there any objective, systematic analyses of the predictive powers of these indicators that I could read?
That's one of the problems of TA. It's more of an art than a science. Human judgement and "gut feel" is far more important than rigorous automation and documentation of the process. If you look long enough at different types of charts you'll find anything you're looking for. The key is being consistent, picking *multiple* indicators which work for you at the moment and going with your gut. And this part blows people's minds: it's just as key to know when your approach is no longer working so you can find a different one.
|
|
|
|
S3052 (OP)
Legendary
Offline
Activity: 2100
Merit: 1000
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 08:10:02 PM |
|
You are right, it is the combination of numerous chart tools and indicators. (Un)fortunately it is not automated and works with only one tool. And even worse, the market follows certain "tools" for a couple of months or even years and then not any more. The key is to accept when this is the case and not automatically follow things.
|
|
|
|
fivebells
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 08:17:13 PM |
|
It's more of an art than a science. Human judgement and "gut feel" is far more important than rigorous automation and documentation of the process. And that's fine, but has anyone established that this is actually a better predictor than other methods? I mean, human judgement was (maybe still is, haven't been following), the most accurate method for protein structure prediction, but that didn't stop people from testing the accuracy of those predictions in competitions with machine prediction methods... You are right, it is the combination of numerous chart tools and indicators. (Un)fortunately it is not automated and works with only one tool. And even worse, the market follows certain "tools" for a couple of months or even years and then not any more. The key is to accept when this is the case and not automatically follow things.
Do you know what went wrong with your August prediction?
|
|
|
|
S3052 (OP)
Legendary
Offline
Activity: 2100
Merit: 1000
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 29, 2011, 08:22:18 PM |
|
|
|
|
|
|
|