How to test if chart patterns occure systematic or are just random artifacts in a price-time series? Given a time series with Events (chart patterns), I want to assess whether events (chart patterns) occur systematic or randomly in a time series.
My idea is to


*

*identify patterns in a given time series

*simulate multiple new time series

*identify patterns in the simulated time series     

*compare the results
The key question is how to compare the results. The events (chart patterns) have a specific length and can be overlapping (as illustrated in the image). 
I could easily just use the number of events in the original time series and compare them to the count/number of events in the simulated time series by simulating enough time series to construct a confidence interval. 
But this would not account for clustering, overlapping and the size of events.
The image is an example with orange lines being events of different size/length in time. They are overlapping but do not seem to be clustered. If they would be clustered they might only appear in the beginning and end of the time series. 

Hence, it is not enough to just compare the count/number of events (patterns) but also if there is clustering (margin between events) or differences in the size of the patterns.
I could perform the same test for different statistics (count/number, size, ?what is the best way to check and compare clustering?) but this does not appear to be very elegant to me (it might be the best approach, I do not know). 
Hence, my question is:
How do it test if events (with different size, overlapping, and maybe clustered) occur systematic in a time series or randomly?
Are there any similar problems in other disciplines?
edit:
I edited this question recently. It was a very old question but I tried to make it more comprehensible. Still, I am looking for some answers. The one proposed is one way but I am wondering if someone maybe know established setups to tests cases like this.
 A: This sounds something like posterior predictive checks in Bayesian statistics.
You can come up with several statistics that measure something about your actual data and then calculate those statistics on multiple (say 1,000) simulations and see where the actual value falls in the distribution.
For example, you could calculate the total number of events in your real data. Then calculate the total number of events in each of your 1,000 simulations. Then see if the actual number is contained in, say, the (0.025, 0.975) quantile of the simulations. If not, your simulations and actual data aren't consistent on this statistic. If you think your simulations are realistic, this is an indication that they're not as realistic as you might think. You might be able to flip this to make a statement about reality but that makes me nervous.
You can do the same thing for various aspects you want to check. You mention clustering at the beginning. Say your actual data has 10 events in the first six months and only two in the last six months: a ratio of 5:1. How about calculate that ratio for your 1,000 simulations and see where in that distribution your actual falls? Or maybe you could look at the maximum number of simultaneous events. Or the percentage of the year in which there were no events.
You can compare SD or mean of event lengths, the mean (or max) gaps between events, etc, etc. Find a way to express what would make the simulations realistic, in terms of  statistic and see whether the actual is consistent with the simulations or not.
I believe this approach might be helpful in your situation, and I believe it's called resampling statistics.
[Thinking about my 5:1 ratio remark, I'm not sure if ratios are tricky or not. If you're satisfied that you're getting a reasonable number of events from your simulation, you could perhaps use the difference in the number of events in the first half and the last half of the year instead of a ratio. Not sure.]
