| tomm1111 202 posts
 msg #67723
 - Ignore tomm1111
 modified
 | 9/25/2008 1:10:04 AM 
 CHUTES and LADDERS
 
 ...work in progress.  (Updated filter last post.)
 
 This filter calculates probability of price change % based on past performance of specified intervals.  I believe it can be applied to both short and longer term trading.
 
 To provide a normal distribution for the probability calculations (important), I used the daily price change percentage.  The calculation is basically a z-table value of a normal distribution (bell curve).
 
 Sort by "x-day" columns.  Negative "x-day" (or z-table) numbers reflect a positive change in close price.
 
 Please feel free to modify, tweak, and add your criteria.  Please share your work on this post.
 
 In the next few days, I will try to get a out a more detailed explanation.
 
 tomm
 
 
 
 
 
 
 
 
 
 | 
| chetron 2,817 posts
 msg #67724
 - Ignore chetron
 | 9/25/2008 6:30:30 AM 
 FOR THE CHEAP SEATS....
 
 
 YOU NEED TO OVERLAY THE PLOTS BY EYE.....
 
 
 
 
 
 
 | 
| chetron 2,817 posts
 msg #67725
 - Ignore chetron
 | 9/25/2008 6:47:09 AM 
 AND MAYBE....
 
 
 
 
 
 
 
 | 
| johnpaulca 12,036 posts
 msg #67727
 - Ignore johnpaulca
 | 9/25/2008 9:05:56 AM 
 You guys are amazing...thanks for the great work.
 
 Chet....thanks for the tickets to the nose bleed seats....lol.
 
 
 | 
| chetron 2,817 posts
 msg #67728
 - Ignore chetron
 | 9/25/2008 9:13:32 AM 
 
 MY PLEASURE.
 
 
 
 | 
| Sqwii 160 posts
 msg #67750
 - Ignore Sqwii
 | 9/25/2008 4:52:21 PM 
 PLease explain this filter in DETAILS... which stocks would be best to buy and hold ?
 
 The stocks on the top of the list ? what are YOU looking after?
 
 
 | 
| chetron 2,817 posts
 msg #67763
 - Ignore chetron
 | 9/25/2008 11:00:20 PM 
 AND EVEN .....
 
 I'M JUST CRANKIN' HERE, BOSS.
 WHAT DO YOU THINK?
 
 
 
 
 
 | 
| tomm1111 202 posts
 msg #67768
 - Ignore tomm1111
 modified
 | 9/26/2008 2:00:59 AM 
 I couldn't do exactly what I wanted to.  I think I ran out of available nested calculations, even with the advanced subscription.  However, this is a step in the direction I wanted to go.
 
 Here is what you want to look for:
 
 1.  clo in extreme low or extreme high
 2.  sort by the "5-day" column at both ends
 3.  Spike in at least z_5, z_10, or z-20.  All 3 is best. (2 better than 1)
 4.  Also, look for fake outs.  Pull backs from extreme top or bottom to quickly rebound.
 5.  All indicators are inverse.  Up = decrease in price, down = increase in price
 6.  This filter has great trend finding abilities.  Sort by x-day column. (neg = upward trend, pos = downward trend.)
 7.  The change in close price was the only metric used in this filter.
 
 tomm
 
 
 
 
 
 | 
| tomm1111 202 posts
 msg #67769
 - Ignore tomm1111
 | 9/26/2008 2:26:32 AM 
 Thanks chetron for crankin' away.  :o)
 
 I'm not sure if the filter in the last post will work up in the nose bleed section.
 
 
 
 | 
| chetron 2,817 posts
 msg #67772
 - Ignore chetron
 | 9/26/2008 8:28:57 AM 
 
 CONGRATULATIONS, TOMM.  YOU HAVE TRULY CROSSED THE THRESHOLD OF ACTUALLY WRITING AN ADVANCED SUBSCRIPTION FILTER.  THIS MAKES  6 SUCH FILTERS THAT HAVE NOW BEEN PUBLISHED ON THIS SITE.  THERE IS NO QUICK WORK AROUND FOR THIS NEW FILTER OF YOURS.  YOU ARE OFFICIALLY AN "EXCEEDER".
 
 
 
 
 |