|
Register | Sign In |
|
QuickSearch
Thread ▼ Details |
|
Thread Info
|
|
|
Author | Topic: Discontinuing research about ID | |||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
RAZD writes: Curiously I note that the most common pattern is Case 15, covering 12 of the 24 episodes, yet it still results in relatively high error counts for the three models here. Yes, your models cause relatively high error counts. Is there a reason you want to replace the multi(4)-pattern model with a high predictive power and a low residual uncertainty with a new 10E, 11E or 15E model with a lower predictive power and a high residual uncertainty, probably 1:10? Curiously what I am doing is reviewing your "pattern" by breaking it down into a more understandable form and looking at the probabilities for each version.
RAZD writes: S1 (E1, E2 and E3), S2 ( E4, E5, E6, E7 and E8), S3 (E9), S4 (E10 and E11), S5 (E12), S6 (E13), S7 (E14) and S8 (E15) and the pattern would be: Then you would have for P.Wo for example: S1: *, +S2: *, +, - or P.Pi: S1: *, -S2: *, +, - That wouldn't be distinct. It appears that you are not understanding what the Sn's are -- or I am not understanding your comment. For clarity, Sn's are like your En's but with the elements grouped into fewer categories to make the uber pattern simpler. For example S1 = E1 elements + E2 elements + E3 elements and Event #1:Elements are observed, either singly or in combinations and all with possible repeated appearances -- P.Al, P.BW, P.Da, P.LF, P.Pi,P.Tr, P.WeC, P.Wo, P.WSA, M1, M2, M5, M6, M7, M13, P.Al-, P.BW+, P.Tr+, P.WeC-. Event #2: Elements are observed, either singly or in combinations and all with possible repeated appearances -- P.Al, P.BeC, P.LF, P.Ri, P.WeC, P.Ya, M4, M5, P.BW-, P.Da-. Event #3: Elements are observed, either singly or in combinations and all with possible repeated appearances -- P.Al, P.BW, P.Da, P.En, P.Pi, P.Ri, P.Tr, P.Wo, P.WSA, P.Ya, M1, M3, M5, M6, P.BW+, P.Pi-, P.Wo+. So S1: Elements are observed, either singly or in combinations and all with possible repeated appearances -- P.Al, P.BW, P.Da, P.LF, P.Pi, P.Tr, P.WeC, P.Wo, P.WSA, M1, M2, M5, M6, M7, M13, P.Al-, P.BW+, P.Tr+, P.WeC-,& & As you can see this adds P.BeC, P.Ri, P.Ya, M4, P.BW- and P.Da- from E2, with P.En, M3, P.Pi- and P.Wo+ additional from E3. removing duplicates and rearranging S1 becomes: S1: Elements are observed, either singly or in combinations and all with possible repeated appearances -- P.Al, P.BeC, P.BW, P.Da, P.En, P.LF, P.Pi, P.Ri, P.Tr, P.WeC, P.Wo, P.WSA, P.Ya, M1, M2, M3, M4, M5, M6, M7, M13, P.Al-, P.BW+, P.BW-, P.Da-, P.Pi-, P.Tr+, P.WeC-, P.Wo+.
Then you would have for P.Wo for example: S1: *, + S2: *, +, - or P.Pi:S1: *, - S2: *, +, - That wouldn't be distinct. If you are talking about the transition from S1 to S2, then it doesn't appear to me to be any different than the transition from E3 to E4 (or E9).
That's still more than 4 variations. Except that you have 24 variations, not 4, that are allowed by your rules.
There are 55 fixed transitions and only 4 transition which allow an additional subvariation. The average of possible variations is far below 8 as in your theoretical pattern. If you have additional rules than what has already been discussed then you need to spell them out. Simply. Every one. if x(1) then y(1)if x(2) then y(2) if x(3) then y(3) etc. What I currently understand is that if an element is observed that is not a member of the current event caste (including predecessors) then it triggers a transition to the next event that contains that element. There is no reason I can seen where they wouldn't apply to the 8 sequence 8 variation pattern and your 15 event 24 variation pattern. Enjoy Edited by RAZD, : corrected first E3 to E2per msg 265 comment by our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
RAZD writes: With this analysis we can see that none of the 7 sequence patterns are a good match to the episodes Seriously, why do you create your own patterns? Shall I review your selfmade patterns? Yes, your new patterns are not a good match. Do you want to hear more about your revised patterns? Oh please do. Here it is in more detail:
Do you refer to the E1-E15 pattern with "uber pattern"? If so, then maybe because of the predictive power? ... No, just that it is the "over" pattern that is made up from 24 sub-patterns and arbitrary rules.
... You successfully managed to convert the E1-E15 pattern with 4 variations to other revised patterns with less predictive power. It's like to say: Look at the fossil record. There are simple lifeforms first and then more complex lifeforms. That means, that God created in 6 days first simple lifeforms and then more complex lifeforms. A theory of evolution might have a higher predictive power but we don't care about uber patterns. We will take a pattern with a lower predictive power and believe in creation. We don't care about uber patterns the fossil record might exhibit. I don't see any value in lumping an uber pattern into three 7 sequence patterns with a much less predictive power and a much higher residual uncertainty. Your opinion. Curiously I think it has better predictive power without all the extra exceptions of your "pattern" -- when every single episode in the first season doesn't match your "pattern" without invoking one or more exceptions to your pattern, most with different exceptions. Your "pattern" is like a deck of cards that you can arrange in many different ways, but only the jokers don't fit the pattern composed of all those different arrangements. Enjoy Edited by RAZD, : dup comment Edited by RAZD, : correct scene 1 here as well per comment Edited by RAZD, : bulletby our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
You accidentally added E3 twice. There is no M3, P.Pi- and P.Wo+ at E2. ... Corrected that above, thanks. M3 is in E3, so it is still part of S1
... You would have *P.Wo, P.Wo+ and P.Wo- at S2. In the "pattern" description in Message 235 *P.Wo is in E1 and the first P.Wo+ is in E3 (both now S1) and there is no P.Wo- until E5 (now S2).
RAZD writes: What I currently understand is that if an element is observed that is not a member of the current event caste (including predecessors) then it triggers a transition to the next event that contains that element. Yes. Any other rules? ie -- can you have both a P.ap+ and a P.ap- in the same event, or does the conversion cause a transition to the next event with the second conversion? Does *P.ap → P.ap+ → P.ap- mean that P.ap is back to where this started?
Yes, but there is only one variation for every start with E1, E3, E4 and E5. You can see that in the three examples in [Msg=261]. For every start with E1, E3, E4 and E5 there is only one possible variation. An exception is example 3 with E3: Let's look at your "pattern" as described in Message 235 again:
More options ... so NOW we have (updating the pattern in Message 166: Yes, but there is only one variation for every start with E1, E3, E4 and E5. You can see that in the three examples in Message 261. For every start with E1, E3, E4 and E5 there is only one possible variation. An exception is example 3 with E3: E1 is one of two options: E1 or not E1 which then each have other options per your Message 237:
E2 is not a possible start. The pattern has
for a total of 24 different pattern variations. To reduce this further you will need to have\show cause and a rule for it. Enjoy Edited by RAZD, : typoby our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
RAZD writes: Any other rules? ie -- can you have both a P.ap+ and a P.ap- in the same event, or does the conversion cause a transition to the next event with the second conversion? Until now there is always only P.ap+ OR P.ap- possible at the same event. *P.ap, P.ap+, P.ap- or *P.ap, P.ap+, P.ap- will therefore always cause at least one transition, if it doesn't break the pattern. Excellent, that gives purpose to observing +/- effects, so we now have two triggers to a new event:
So *P.ap or *P.ap & P.ap+ or *P.ap & P.ap- can occur in one event but not *P.ap & P.ap+ & P.ap- Can a second +/- effect in the same direction occur in one event (ie - *P.ap & P.ap+ & P.ap+ or *P.ap & P.ap- & P.ap-)? Then I can start building my sequences - using your data - and should be able to reproduce your results, yes?
Why should it be reduced? There are 24 variations in total and about 4 actual possible variations for every row of appearances. To describe all *, + and - in the first minutes of 76 episodes demands some complexity. The expected probability to fit was tested to 0.625 and calculated to <0.711 in Message 190. The residual uncertainty was then calculated to 1:10^7. You can verify the result here: Message 171. In other words, there are 24 possible variation in your overall (uber) pattern, but only 4 have been observed when you follow the rules above? Note that in my 8 possible Sequence variation overall (uber) pattern in Message 270 (corrected per your comment) there were only 5 of the 8 possible variation observed in the first season:
Sequence #1 observed 0 times (S1,S2,S3,S4,S5,S6,S7,S8) Sequence #2 observed 1 times (S1,S2,S3,,S5,S6,S7,S8) Sequence #3 observed 13 times (S1,,S3,S4,S5,S6,S7,S8) Sequence #4 observed 1 times (S1,,S3,,S5,S6,S7,S8) Sequence #5 observed 9 times (,S2,S3,S4,S5,S6,S7,S8) Sequence #6 observed 1 times (,S2,S3,,S5,S6,S7,S8) Sequence #7 observed 0 times (,,S3,S4,S5,S6,S7,S8) Sequence #8 observed 0 times (,,S3,,S5,S6,S7,S8) Logically you should have observed more than 5 variations in your pattern for the first season. I haven't checked that yet, but I think your claim of only observing 4 variations must be an error as I should end up with fewer sub-patterns observed than your system, due to the way I have combined some of your events. For instance if we observed:
E1, E2, E3, E4, E5, E6, E7, E8, E9, E10, , E12, E13, E14, E15 and then E1, E2, E3, E4, E5, E6, E7, E8, E9, , E11, E12, E13, E14, E15 That would count as two variations observed, while my system would only count 1 variety:
S1(E1, E2, E3), S2(E4, E5, E6, E7, E80, S3(E9), S4(E10, ), S5(E12), S6(E13), S7(E14), S8(E15) and then S1(E1, E2, E3), S2(E4, E5, E6, E7, E80, S3(E9), S4( , E11), S5(E12), S6(E13), S7(E14), S8(E15) And curiously when I do the same counting of your pattern variations for the first season I find 8 variations observed (16 possibilities not observe).
Event variation #1 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #2 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #3 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #4 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) Event variation #5 observed 0 times (E1,E2,E3,,,,,,E9,E10,E11,E12,E13,E14,E15) Event variation #6 observed 0 times (E1,E2,E3,,,,,,E9,E10,,E12,E13,E14,E15) Event variation #7 observed 1 times (E1,E2,E3,,,,,,E9,,E11,E12,E13,E14,E15) Event variation #8 observed 1 times (E1,E2,E3,,,,,,E9,,,E12,E13,E14,E15) Event variation #9 observed 0 times (,,E3,E4,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #10 observed 0 times (,,E3,E4,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #11 observed 0 times (,,E3,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #12 observed 1 times (,,E3,E4,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) Event variation #13 observed 0 times (,,E3,,,,,,E9,E10,E11,E12,E13,E14,E15) Event variation #14 observed 0 times (,,E3,,,,,,E9,E10,,E12,E13,E14,E15) Event variation #15 observed 12 times (,,E3,,,,,,E9,,E11,E12,E13,E14,E15) Event variation #16 observed 0 times (,,E3,,,,,,E9,,,E12,E13,E14,E15) Event variation #17 observed 1 times (,,,E4,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #18 observed 1 times (,,,E4,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #19 observed 7 times (,,,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #20 observed 0 times (,,,E4,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) Event variation #21 observed 0 times (,,,,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #22 observed 0 times (,,,,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #23 observed 1 times (,,,,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #24 observed 0 times (,,,,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) So your claim of only 4 viable variations is not correct.
Why should it be reduced? ... Why should it be made more complex than it needs to be? Enjoy Edited by RAZD, : .. Edited by RAZD, : ... Edited by RAZD, : ....by our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
RAZD writes: So *P.ap or *P.ap & P.ap+ or *P.ap & P.ap- can occur in one event but not *P.ap & P.ap+ & P.ap- Yes. Also P.ap+ or P.ap- can occur in one event. *P.ap or P.ap+ or P.ap- or *P.ap & P.ap+ or *P.ap & P.ap- can occur in one event but not *P.ap & P.ap+ & P.ap- or P.ap+ & P.ap-
RAZD writes: Can a second +/- effect in the same direction occur in one event (ie - *P.ap & P.ap+ & P.ap+ or *P.ap & P.ap- & P.ap-)? Yes. All with possible repeated appearances. So again, the rules as I understand them:
No. There are 24 possible variation in overall, but only 4 Sorry, repeating this claim after I have shown it to be false is rather disingenuous. As I said in Message 275:
And curiously when I do the same counting of your pattern variations for the first season I find 8 variations observed (16 possibilities not observe).
Event variation #1 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #2 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #3 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #4 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) Event variation #5 observed 0 times (E1,E2,E3,,,,,,E9,E10,E11,E12,E13,E14,E15) Event variation #6 observed 0 times (E1,E2,E3,,,,,,E9,E10,,E12,E13,E14,E15) Event variation #7 observed 1 times (E1,E2,E3,,,,,,E9,,E11,E12,E13,E14,E15) Event variation #8 observed 1 times (E1,E2,E3,,,,,,E9,,,E12,E13,E14,E15) Event variation #9 observed 0 times (,,E3,E4,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #10 observed 0 times (,,E3,E4,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #11 observed 0 times (,,E3,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #12 observed 1 times (,,E3,E4,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) Event variation #13 observed 0 times (,,E3,,,,,,E9,E10,E11,E12,E13,E14,E15) Event variation #14 observed 0 times (,,E3,,,,,,E9,E10,,E12,E13,E14,E15) Event variation #15 observed 12 times (,,E3,,,,,,E9,,E11,E12,E13,E14,E15) Event variation #16 observed 0 times (,,E3,,,,,,E9,,,E12,E13,E14,E15) Event variation #17 observed 1 times (,,,E4,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #18 observed 1 times (,,,E4,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #19 observed 7 times (,,,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #20 observed 0 times (,,,E4,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) Event variation #21 observed 0 times (,,,,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #22 observed 0 times (,,,,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #23 observed 1 times (,,,,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #24 observed 0 times (,,,,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) So your claim of only 4 viable variations is not correct. And I have now expanded this analysis for the third season (the second season is skipped) data you provided in appendix A:
Event variation #1 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #2 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #3 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #4 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) Event variation #5 observed 2 times (E1,E2,E3,,,,,,E9,E10,E11,E12,E13,E14,E15) Event variation #6 observed 0 times (E1,E2,E3,,,,,,E9,E10,,E12,E13,E14,E15) Event variation #7 observed 4 times (E1,E2,E3,,,,,,E9,,E11,E12,E13,E14,E15) Event variation #8 observed 2 times (E1,E2,E3,,,,,,E9,,,E12,E13,E14,E15) Event variation #9 observed 0 times (,,E3,E4,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #10 observed 0 times (,,E3,E4,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #11 observed 2 times (,,E3,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #12 observed 1 times (,,E3,E4,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) Event variation #13 observed 0 times (,,E3,,,,,,E9,E10,E11,E12,E13,E14,E15) Event variation #14 observed 0 times (,,E3,,,,,,E9,E10,,E12,E13,E14,E15) Event variation #15 observed 19 times (,,E3,,,,,,E9,,E11,E12,E13,E14,E15) Event variation #16 observed 2 times (,,E3,,,,,,E9,,,E12,E13,E14,E15) Event variation #17 observed 0 times (,,,E4,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #18 observed 1 times (,,,E4,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #19 observed 11 times (,,,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #20 observed 1 times (,,,E4,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) Event variation #21 observed 0 times (,,,,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #22 observed 2 times (,,,,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #23 observed 4 times (,,,,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #24 observed 0 times (,,,,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) out of 51 total episodes (not counting episode 1x02 as it is a continuation of 1x01) 12 of the 24 possible cases were not observed 12 of the 24 possible cases were observed in two seasons of data This is your information from your data in your Appendix A for seasons 1 and 3 (season 2 was skipped by you). And I would expect to pick up more of the variation cases with the next season of episodes -- because I don't expect them to fit the pattern of use built up in the first two seasons ... just as season 3 did not fit the pattern of use seen in the first season. Enjoy ps -- I note that you have two errors in your appendix A found so far: (1) 3x08 The Price - recorded elements: *P.Al, P.Tr+,*P.Ya, *P.Tr, P.Tr-, *P.BW,P.Tr-, *P.Pi, P.Tr-... P.Tr+ not preceded by *P.Tr -- should be *P.Al, *P.Tr, P.Tr+, *P.Ya, *P.Tr, P.Tr-, *P.BW, P.Tr-, *P.Pi, P.Tr- (2) 3x23 Sarek - recorded elements: *P.Pi, P.Al+, *P.BW, *P.WeC, *P.BW, P.Pi-... P.AI+ not preceded by *P,AI -- should be *P.Pi, *P,AI, P.Al+, *P.BW, *P.WeC, *P.BW, P.Pi- (I have deleted your /E and time data from these strings for clarity) Edited by RAZD, : psby our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined:
|
Until you separate the design elements from the chance elements you have no chance of calculating a probability of any occurrence. I have yet to see you make any attempt to identify any elements of direct human choice. All you do is deny that humans have any affect on the establishing a pattern that covers a mere three episodes. There are actually several levels of ambiguity that I can see:
Which aspect of an element is recorded? (We don't know from the data: for instance, when *P.AI is recorded it can be any one of 13 aspects):
quote: Romulans are counted as P.BW instead of P.Wo when the definition of P.BW is "colour black/white, silver, ice, cold, invisible" and the definition of P.Wo includes "hostile aliens" ... ??? Does *P.Wo as Lt Worf shooting a *P.Wo as a "hostile alien" count as P.Wo-? Does the sequence {*P.AI, P.Wo+} mean that P.Wo was one of "more than five" people but only he was affected or even that P.Wo was a "hostile alien" in the group of "more than five" people. When an appearance of an element that is not part of the cast in E(n) it triggers a transition to the next event that contains that element E(n+j), which may be the next event or one further down the line, but the only invalidation of the "pattern" is an element appearing out of sequence. The "hop-scotch" between events hides sequences that don't conform to the strict step by step chronology of the events (which is not recorded for any of the observed episodes, they are all sub-patterns of the full length "pattern"). What it appears to be is an arbitrary lumping\conflating of things into groups with no rhyme or reason other than being able to create a "pattern" from the mish-mash by conflating several things arbitrarily as equally of value to the "pattern" ... one that is so flexible that it covers thousands of different sequences of diverse things ... and thus the "pattern" appears to be an artifact of multiple options at every level by hiding what is actually digitized. A more rigorous approach would have each variety of every element defined and recorded as a separate element, Enjoy Edited by RAZD, : typoby our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
Sigh... Oh it just gets worse ... I went back to the "pattern" event descriptions in Message 235 and listed all the options for the elements in each event ... That's 67 E1 options, 34 E2 options, 58 E3 options, 35 E4 options, 44 E5 options, 26 E6 options, 60 E7 options, 59 E8 options, 83 E9 options, 15 E10 options, 24 E11 options, 62 E12 options, 34 E13 options, 84 E14 options and 83 E15 options ... If I do that for all the different events and then run those through all the 24 possible variations of combining events into sequences it becomes rather obvious that the number of variations of elements and their combination into sequences would result in an astronomical number of fits. And I'll bet his "probability calculation" treats them as single entities. Looks like CatsEye was onto something with replacing the elements with their descriptions. Enjoy Edited by RAZD, : corrected, added up all the event options for each eventby our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
Have I got this right - are you saying that Dubreuil is defining as a qualifying event a wide range of possible scenarios (ie you hit the side of the barn somewhere), but then calculates the probability of the specific scenario that does occur (ie the probability that you would hit the specific bullet hole you actually hit) ? In other words, a very convoluted version of the sharpshooter fallacy ? Well he also paints bulls-eyes on the house, the car, the tractor and the cat to make sure he doesn't miss one. And some of those elements are just colors ... something you never see on color tv shows, right? Enjoyby our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined:
|
I have a good idea. I will ... ... declare victory and run away from the comments that raise serious questions about basic honest and rigorous approaches to doing such "studies" ...
... even RAZD still becomes surprised about what he reads in the paper. ... Curiously, I would say I am amazed at the deception you have created -- apparently it fooled you.
... then I can block every insulting person I want. No one is blocking here anything. And that's the difference between having an open discussion on the merits of the paper and having a closed discussion where you only admit those who don't criticize your fantasy document. Calling you paper bogus is not insulting, it is accurate. Frankly I care more about all the ant frass in antarctica than I do about your responses, especially when you cling to clear fallacies than admit to error. Enjoyby our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
You misunderstood this. All 24 possible variation can be observed, but only 4 can be observed for one row of appearances. There are 4 possible variations for one quantisation. If there are 24 quantisations, then all 24 variations can be possibly observed. This doesn't make any sense to me -- can anyone explain what he means?
For season 3, 4, 5 and 6 there are only appearances and affected persons noted that trigger the next event. *P.Al and *P.Tr were not mentioned therefore. In other words the data presented is incomplete, and only the data that supports the presumed "pattern" is provided, the rest is ignored.
RAZD writes: Does *P.Wo as Lt Worf shooting a *P.Wo as a "hostile alien" count as P.Wo-? If there are more than 13 persons, then a 14th person can appear as an already present person. For example P.Wo. This demonstrates the primary problem with the way the elements are defined with so much ambiguity: you can have P.Wo (Worf) in a battle with P.Wo(Klingons) and one wins (P.Wo+) and the other loses (P.Wo-), then P.Wo(hostile aliens) come to the aid of the losing side (P.Wo+). And again we see that the data in the appendix is not a record of what was observed (Lt Worf, hostile aliens or Klingons) but the category that arbitrarily combines them (P.Wo)
RAZD writes: one that is so flexible that it covers thousands of different sequences of diverse things Yes, and it also doesn't fit with thousands of different sequences of diverse things. For two sequences that fit, there is about one sequence that doesn't fit. Curiously I don't think you realize the problem inherent in having 24 to 84 different items that make up your events causing such an open ended system that matches to your events is almost a foregone conclusion -- and that even when that fails to fit the pattern you adjust the pattern so that events can be skipped ... and you have no list of elements that would invalidate the "pattern" and no record of anything that isn't part of one of your element. Things like the color blue are not recorded in the data (as you haven't tagged that to an element) so we don't know from your record of the data whether blue occurred or not, even though it would be an element that is not part of the sequence and would therefore invalidate it. You have "cherry-picked" what you record as data so that it supports your "pattern" ... this is neither an honest nor a rigorous table of data. When you look at the number of options for each element and compare that to them being recorded from 76 episodes it is easy to see why you always get E9 (83 options), E14 (84 options) and E15 (83 options) because of the way they are packed and stacked.
Message 287: That's 67 E1 options, 34 E2 options, 58 E3 options, 35 E4 options, 44 E5 options, 26 E6 options, 60 E7 options, 59 E8 options, 83 E9 options, 15 E10 options, 24 E11 options, 62 E12 options, 34 E13 options, 84 E14 options and 83 E15 options ... Message 281: ... when I do the same counting of your pattern variations for the first season I find 8 variations observed (16 possibilities not observe). And I have now expanded this analysis for the third season (the second season is skipped) data you provided in appendix A: and now I have added season 4, the third one used to make\create\manufacture the "pattern" Event variation #1 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15)Event variation #2 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #3 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #4 observed 0 times (E1,E2,E3,E4,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) Event variation #5 observed 2 times (E1,E2,E3,,,,,,E9,E10,E11,E12,E13,E14,E15) Event variation #6 observed 0 times (E1,E2,E3,,,,,,E9,E10,,E12,E13,E14,E15) Event variation #7 observed 6 times (E1,E2,E3,,,,,,E9,,E11,E12,E13,E14,E15) Event variation #8 observed 3 times (E1,E2,E3,,,,,,E9,,,E12,E13,E14,E15) Event variation #9 observed 0 times (,,E3,E4,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #10 observed 0 times (,,E3,E4,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #11 observed 3 times (,,E3,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #12 observed 1 times (,,E3,E4,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) Event variation #13 observed 0 times (,,E3,,,,,,E9,E10,E11,E12,E13,E14,E15) Event variation #14 observed 2 times (,,E3,,,,,,E9,E10,,E12,E13,E14,E15) Event variation #15 observed 23 times (,,E3,,,,,,E9,,E11,E12,E13,E14,E15) Event variation #16 observed 5 times (,,E3,,,,,,E9,,,E12,E13,E14,E15) Event variation #17 observed 0 times (,,,E4,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #18 observed 1 times (,,,E4,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #19 observed 12 times (,,,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #20 observed 2 times (,,,E4,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) Event variation #21 observed 0 times (,,,,E5,E6,E7,E8,E9,E10,E11,E12,E13,E14,E15) Event variation #22 observed 2 times (,,,,E5,E6,E7,E8,E9,E10,,E12,E13,E14,E15) Event variation #23 observed 12 times (,,,,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15) Event variation #24 observed 2 times (,,,,E5,E6,E7,E8,E9,,,E12,E13,E14,E15) out of 76 total episodes (episode 1x02 is a continuation of 1x01 and episode 4x01 is a continuation of 3x26) → 10 of the 24 possible cases were not observed → 14 of the 24 possible cases were observed in two seasons of data Note that this is two more cases observed than before (and I predicted there would be an increase ... in order to force a pattern onto the data) ... and that the base pattern (#1) has the most overall options for being found (67!*34!*58!*35!*44!*26!*60!*59!*83!*15!*24!*62!*34!*84!*83! = 1.2610^1030 optional "fit" variations ..), and yet it wasn't found once in the 76 episodes ... this should be your first clue that there is no real pattern. What makes it seem like a pattern is the incorporation of skips between events when an element, el(x) does not fit the current event, E(n) or the next chronological event, E(n+1) ... because the way the data is recorded and the way items are incorporated into elements and the way the elements are grouped in the events is guaranteed to find a following Event, E(q) that "fits", so the only way you get a failure is if that skip has not been built into the 24 variations. The second clue should be that the most common variation found (#15 occurs 23 out of 76 episodes, or 30%) uses one of the shortest variations: only 7 of the 15 events (47%) of the event series . Finally, that only 14 of the possible 24 variations (58%) are actually observed should be your third clue that the "pattern" is an artifact of the construction of the pattern and not an actual pattern in the tv series. Logically , if the variations are necessary, then there should be at least one of every case found in 76 episodes. Put another way, 42% of the "pattern" is unnecessary. GIGO imho.
Message 1: I spent a few years The paper does not test any predictions of, or provide and support to, intelligent design as the paper fails to demonstrate what it claims. The data documented is not usable because it is incomplete, skipping over many items, and it is preprocessed into element groups rather that the raw data; any data that does not fit is not reported. The grouping of optional items into elements, and of optional elements into events, and of optional events into pattern variations, is haphazard and arbitrary, with no rhyme or reason. No rationale is provided for these choices, nor is their opportunity to judge whether they are subjective or objective classifications. If the authors learn anything from this paper, it should be how not to do one. Enjoy.by our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
Sorry I don't buy it. You are hiding behind being offended to keep from dealing with the issues raised. Such a prima donna .
GIGO really means that you get out the quality of information that you put in.
For example you will find a record of elements that would invalidate the pattern in Message 190 for E1. ... Except that it doesn't necessarily invalidate the pattern, it just means you can skip to the next element. You have so many loopholes built into the grouping of multiple items in each element and into the grouping of elements into events and in the grouping events into subpatterns. When you look at the breakdown of what subpatterns were found you see very little actual correlation - the results are all over the map, favoring the shortest versions heavily. You are inconsistent in what you assign to people and what you assign to marks, but this isn't readily apparent until you spell it all out. Your groupings defy logic: how can a color represent a person? How can a color be affected? P.Wo can be hostile aliens and Klingons but Romulans (hostile aliens) are P.BW??? When your groupings are ambiguous and illogical and obscure, then it makes the results ambiguous and illogical and they obscure what actually occurred. So yes your "pattern" is garbage -- because you can't tell what is matched to an element from one event to the next. Nor is your "data" really the actual data, the actual items that were (supposedly) noted, rather they are the element packages they are massaged into, combined in a way that obscures the actual items data so that it is impossible to tell if the second or third or whatever reference to an "element" refers to the same element item. It is also incomplete because you only recorded what you want to use (or want others to use). That is shoddy, misleading, mistaken and incomplete record keeping, and it obscures the actual action: it is not an honest representation of the actual item by item data. So yes your "data" is garbage -- it is unusable to check your work (see below).
... This paper was revised multiple times during the last year through comments from people who really was concerned with the paper. If there wouldn't have been that much comments about it from April 2014 to March 2015, then it could not have been extended and revised that often. Therefore I'm really sure that the content of the paper is correct. I know all the real weaknesses in the old versions of the paper which were corrected during the last year. ... I'll bet that my comments apply to the original paper just as strongly. Curiously I started to try to replicate your pattern construction. The start is relatively easy: group all the elements (assuming the elements are rational grouping of items, which they aren't) that appear first in the episodes for event 1 and I got:
*P.Al(32),*P.BeC(1),*P.BW(2),P.BW-(1),*P.Da(1),*P.Pi(26),*P.Ri(3),*P.Tr(1),P.Tr+(1),*P.Wo(4),*P.Ya(1),M1(1),M10(2) It does not surprise me that the most frequent beginning element is *P.Al (42%) -- because it is so loosely defined that any one of 15 different items (>5 people, green, big, wide, a lot, lack of knowledge, do nothing, holiday, very old, starships, standby, science, stone, death, or #4) are grouped ad hoc into this "element" without rationale or reason, with no visible purpose ... other than to get (force?) a positive result out of a random mix of items, including (as it does) many things to be expected in an opening scene: >5 people, big, starships, etc) .... It also does not surprise me in the slightest that *P.Pi(26) is the next most frequent beginning element (34%) seeing as he is the commander. Likewise I would expect A crew-member in an opening scene (*P.BeC(1), *P.Da(1), *P.Ri(3), *P.Tr(1), *P.Wo(4 ... if actually Worf ...)) (10 more beginning elements, 13%) So that is 89% of the episodes with completely predictable results -- without expecting a pattern. Note that P.BW-(1), and P.Tr+(1) are errors because your rule is that you can't affect a person before they appear, you cannot have P.ap(+/-) before *P.ap ... and these are the very first elements in your data list. When I looked at the second elements recorded I found:
*P.Al(8),P.Al+(1),P.Al-(3),*P.BeC(4),P.BW-(1),*P.Da(1),*P.LF(6),*P.Pi(8),P.Pi+(1),*P.Ri(4),P.Tr-(1),*P.WeC(4),P.Wo-(2),*P.Ya(1),M1(3),M6(4),M7(2) And combining these to find the first two element I get:
*P.Al(40),P.Al+(1),P.Al-(3),*P.BeC(5),*P.BW(2),P.BW-(2),*P.Da(2),*P.LF(6),*P.Pi(34),P.Pi+(1),*P.Ri(7),*P.Tr(1),P.Tr+(1),P.Tr-(1),*P.WeC(4),*P.Wo(4),P.Wo-(2),*P.Ya(2),M1(4),M6(4),M7 (2),M10(2) *P.Al(40) and *P.Pi(34) now account for 74 of the 76 episode opening scenes. Shocking ... But there is a very distinct and disturbing problem at this point, caused by the way the (lack of) data is recorded -- the "data" only lists one ( →1← ) element for the episode events after the first season: the second "element" is not recorded for any "event" -- so this is not, cannot be, an accurate, correct, actual representation of what was on the screen for the opening scene, no matter how you massage the data. This makes it completely and irrevocably impossible to even begin to attempt to reconstruct anything remotely resembling a pattern from this poor excuse for data.
... This paper was revised multiple times during the last year through comments from people who really was concerned with the paper. If there wouldn't have been that much comments about it from April 2014 to March 2015, then it could not have been extended and revised that often. Therefore I'm really sure that the content of the paper is correct. I know all the real weaknesses in the old versions of the paper which were corrected during the last year. ... That nobody else noted the errors and omissions and obscuration of the data does not make the paper a good paper, it just means they did not find the errors and omissions and obscuration of the data, even though these problems make this paper virtually useless.
... Your other comments also reveal that you haven't read the paper completely. It would still take a long time to explain all this to you what you haven't understood by now. ... Curiously I don't think anything left in the paper that I haven't covered nor anything you can say based on it will be able to rescue this paper from the problems in it. Enjoy Edited by RAZD, : clrty Edited by RAZD, : ..by our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
@RAZD: The list of 26 occurrences in Message 190 invalidates the pattern at E1, it is not skipped to the next element. ... You mean these?
For E1 there are 25 occurrences that fit with the pattern (P.Al, P.BW, P.Da, P.LF, P.Pi, P.Tr, P.WeC, P.Wo, P.WSA, M1, M2, M5, M6, M7, M13, P.Al-, P.BW+, P.Tr+, P.WeC-, P.BeC, P.Ri, P.Ya, M4, P.BW-, P.Da-) and 26 occurrences that break the pattern (*P.En, M3, M10, M11, M12, M14, P.Al+, P.BeC+, P.BeC-, P.Da+, P.En+, P.En-, P.LF+, P.LF-. P.Pi+, P.Pi-, P.Ri+, P.Ri-, P.Tr-, P.WeC+, P.Wo+, P.Wo-, P.WSA+, P.WSA-, P.Ya+, P.Ya-). And here you make several errors. First P.BeC, P.Ri, P.Ya, M4 are not elements of E1. Second P.Al-, P.BW+, P.Tr+, P.WeC-, P.BW-, P.Da-, P.Al+, P.BeC+, P.BeC-, P.Da+, P.En+, P.En-, P.LF+, P.LF-. P.Pi+, P.Pi-, P.Ri+, P.Ri-, P.Tr-, P.WeC+, P.Wo+, P.Wo-, P.WSA+, P.WSA-, P.Ya+, P.Ya- cannot occur in E1 by your own rules. Let me correct you:
For E1 there are Third most of these purported deal breakers don't necessarily cause a break:
The only one that actually causes an unquestionable break is M12. One (rare?) element (It only appears 4 times in the whole (documented) data set). Possibly to make events 12 to 15 get counted when they otherwise would fail, ... in four episodes. Do you begin to see how your loading of multiple different items into element groupings contributes to fitting them into your purported pattern? When the number of actual items included in P.Al (15), P.BW (8), P.Wo (3) easily account for the majority of possible items? Do you begin to see how your use of obscure or rare items for some elements (eg M12) contributes to fitting them into your purported pattern? When the other items fail to provide a fit? Do you begin to see how your hop skipping pattern contributes to fitting the elements into one of the event subpatterns? Do you begin to see how these three different ways of muddling the information contribute to an appearance of a pattern when the reality is something quite different?
... If you had read appendix B or Message 261 just once, then you would know that nothing is skipped. Riiiiiiiiiiiiiiight. You mean the part where you add more into the elements?
RAZD writes: *P.Al(40) and *P.Pi(34) now account for 74 of 76 episodes. Shocking ... Yes, the results of your analysis are that irrelevant that I'm shocked you even mention them. au contraire it is very relevant to the fact that your purported pattern is a bogus fabrication. And yet that is the only analysis your pathetic data record allows anyone to do. And that makes it is pointless to pursue any further attempt to replicate your results in any way. For all I know you made it up. Just as P.Al is made up from 15 disparate and quite unrelated items, any one of which "triggers" an appearance in order - apparently - to force an "appearance" into some kind of structure you have invented. Do we know which item was used to "document" the first appearance of P.Al? Nope. Do we know which item was used to "document" the second appearance? Nope. Is there any reason to expect one to relate to the other? None at all. This holds for the majority of the elements and it holds for ALL of the events, if not the majority of the sequences ... We see *P.Al documented ... what variation of the 15 possible items is involved? We see *P.Al documented ... which of the pattern variations applies ... only 4 "events" don't contain P.Al in one form or another. Similar for *P.BW with 8 possible items involved and only 4 "events" don't contain P.BW in one form or another. That kind of multiple overlap produces a lot of elasticity in making data (the real data of items) stretch to fit into a manufactured pattern. Then there are the items that make no sense. Colors are not people, but you use them to "document" people ... and then you use colors to "document" your "marks" ... which can't be affected +/- like people ... so some colors can be affected and others can't ... for no discernible reason. It is all ad hoc and arbitrary and inconsistent. Here's a challenge: instead of having multiple personality people and marks have one item definition for each one -- you go back through your real data (the item by item documentation for the full length of each opening episode that is not in the paper) and then tell me what pattern you have. Start with episode 1x01. Then compare it to episode 1x02 ... let me know what you find.
The <0.711 probability ... Not one person here believes that your calculation is valid, and what you have presented does not give anyone reason to believe it. For instance it appears that you calculated it based on the occurrence of your 15 events and 24 variations, when it should be based on the individual items ... which I know you did not do because you don't provide the data on them, you ignore the real data once you've fabricated your "elements" and "events" ... it's necessarily a bogus calculation. The more I try to follow your perambulations the more I see how bogus it really is. Enjoy Edited by RAZD, : clrty Edited by RAZD, : .by our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
Do you have ever seen this image from page 6?
Yes, it fails to show that you can (and most often you actually DO) skip to E3 (or E4 or E5) ... so it is not a complete or honest representation of your overall pattern.
There is no transition from E1 to E3. There is no transition from E1 to E5. There is no transition from E1 to E9. Yes, but you just move your data from E1 to E3 (not a transition AND not a deal breaker because you can skip E1 and E2) and then you have a valid transition from E3 to E9 ... Or you move it to start at E4 (which is also allowed by your "pattern") and then you have a transition from E4 to E5 ... Now I shouldn't have to explain this to you -- your "pattern" allows this, I suggest you LOOK at your paper. Critically. The way you should have before you presented to the world. Same for the other dealbreakers you listed. That you apparently haven't looked into all the ramifications of your system does not give you any special standing to say I haven't. It just shows a level of sloppy work and carelessness in reviewing your own work and asking yourself why do you do the things you do: why do you lump colors with people? Why do you lump 15 different and unrelated things in one "person" when few of them are actually aspects of people. Just for starters. Every one of those supposed deal breakers except M12 can be rerouted to another pattern that starts in a different place -- and that is what I showed. I suggest you READ what I said not just dismiss it as wrong because you want it to be. To properly refute what I said you need to SHOW that it is wrong. Sadly 90+% of your posting is bitching about others rather than dealing with the criticism -- that is what defines prima dona behavior. If you can't be ruthless with reviewing your own work then don't be surprised when others are -- that is how peer review works. Can you answer these questions? (you should have the answers in the paper):
Obviously I could continue for the full 15 "events" but you should get the idea. If you spend some time thinking about this you might see why your paper is full of holes. Why are fire and water "marks" but stone and earth are "people"?Why are green, black, white, silver, yellow "people" but black(?), and red are "marks"? Arbitrary and inconsistent classifications are not the way science is done.
If you call me prima donna, am I then allowed to call you stupid head? You can call me what you like, it won't change the FACT that almost all of what you said about deal breakers for E1 was false, nor will it change the FACT that I showed where and why it was false. You can call me what you want, but that won't change the FACT that your purported "data" is completely useless for replicating your work, and does not present the actual data but an unexplained filtered grouping of the data, as I have shown and documented. You can call me anything you please, but that won't change the FACT that your classifications are arbitrary, filled with apparently random choices. and inconsistent from one element to the next, from "people" to "marks" and from one event to the next. Science is based on consistency and logic, not on random whim. And you can call me whatever will make you feel better, but it won't change the FACT that the real data is not reported, that the real data is obscured by several layers of hand waving that only serves one purpose: to create the appearance of a pattern when there is none. Now you have an opportunity to prove me wrong: you can provide us with the real data, the actual items you tagged, the individual people, the individual colors, all of it as single data items and not lumps that you have conjured\developed (those "lumps are part of your pattern, NOT data: you have not reported any real data). Curiously I won't hold my breath, I expect you'll be too busy being outraged and throwing another temper tantrum. Enjoyby our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
RAZD will have to confirm, but I think you may again be misinterpreting what he is saying. He isn't saying that your paper contains transitions from E1 to E3, or from E1 to E5, or from E1 to E9. He's saying that your paper fails to consider the evidence for such transitions. Actually what I said was:
Third most of these purported deal breakers don't necessarily cause a break:
The only one that actually causes an unquestionable break is M12. One (rare?) element (It only appears 4 times in the whole (documented) data set). Possibly to make events 12 to 15 get counted when they otherwise would fail, ... in four episodes. When you compare the elements in E1 to those in E3, E4 and E5 you see that most of them overlap one or the other So instead of "breaking" the "pattern" at E1 it moves the start from E1 to E3, E4 or E5 and then these purported deal breakers become normal transitions to another event in the multitude of variations of the pattern. If you look through Appendix A you will see that only 11 of the 76 episodes are documented starting at E1 ... <15% of the 76 episodes documented. AND none of those 11 episodes used the E4-E8 option. Then 34 episodes start at E3 ... (45%, total <60% of the episodes), followed by 15 episodes starting at E4 and 16 starting at E5. Starting at E1 is the exception, not the rule. What you really have are 5 different starting sequences:
Leaving the rest of the variation to whether or not E10 and/or E11 are included. Most of the episodes involve massive skipping of (4 or more of) the purported 15 "events" with the most common "pattern" recorded being case 15: ,,E3,,,,,,E9,,E11,E12,E13,E14,E15 which is only 7 events long, and it occurs 23 times (30%). Next most common are case 19: ,,,E4,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15, which is 11 events long and occurs 12 times and case 23: ,,,,E5,E6,E7,E8,E9,,E11,E12,E13,E14,E15, which is 10 events long and occurs 12 times as well. These three cases account for 58 of 76 episodes (76%). More damaging imho is the fact that there are 24 possible variations in the pattern as described, but only 14 are actually observed, the rest are produced by the rule that have been made to cram the 76 episodes into the pattern, which means that 42% of the "pattern" is not required to explain the data. That's sloppy. The "pattern" has been made so open and adaptable that it is difficult to find a real deal breaker instead of shifting to another variation. When you look at the distribution of pattern variation length in number of elements you find
Only 3 episodes out of 76 are longer than 11 events of the purported 15 event string (4%) and they are all 12 events long. The overall average length is only 8.8 events, showing a heavy bias to the shorter variations being counted. This is a very bad match between the model and the data - even massaged as heavily as this data is. In other words 96% of the episodes documented skipped 4 or more events. Skipping 4 or more events is the rule, not the exception. That's poor form, and it points to another reason the probability calculation is erroneous -- it assumes a greater degree of complexity than actually exists in the data as presented. Furthermore he stopped counting when he got to E15, no matter how much of the episode opening scene was left. This sad fact is also obscured by the lackadaisical way the purported data is recorded. The appendix only records elements rather than the items in the element definitions (most with multiple variations which thus are not recorded), and only the first season data records more than one element for what are listed as the assigned events, so you have no idea who or what else is in those events. That's unexcusable. As a result I have no doubt that the way his purported probability is calculated from the data recorded while ignoring the actual data and using packages of data as single element and events, is false. Enjoy Edited by RAZD, : sum Edited by RAZD, : . Edited by RAZD, : AND 5 startsby our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
RAZD Member (Idle past 1435 days) Posts: 20714 From: the other end of the sidewalk Joined: |
I already stated I won't take part in a "discussion" anymore. You, RAZD, still make the same mistakes as in the comments before. Some of your comments are just wrong, some of your comments are true, but you claim they disprove the paper. I don't even know why you are still commenting. I already stated I won't take part in a "discussion" anymore. I could explain you your mistakes, but I will not as explained in Message 289. ... Predicted behavior.
... Some of your comments are just wrong, ... A comment curiously devoid of factual substantiation, and therefore worthless whining.
... It is more fun to offend with other persons together than to do it alone. I will argue from now on against the paper and therefore against RAZD. You appear to confuse my discussion with support of the paper ... once again you seem to have trouble with basic logic. You can -- of course -- argue against the paper, in fact I strongly recommend it, as this is what you should have done before attempting to publish it. You can even develop your own arguments, which may or may not parallel mine.
RAZD, you still use the short forms of the author. For example: P.Da, P.Ri or P.BW. You can't do this. If you want to write *P.Da, then write *(Data). Or if you want to write P.BW+, then write (colour black/white, silver, ice, cold, invisible, 6)+, but not P.BW+. ... What a good idea: why don't you go back and do that for the whole paper ... but you should fully implement it: if you want to write P.BW+, then write ("P.BW" = {Romulans or colour black or color white or color silver or ice or cold or invisible or the number 6} only 1 needed at a time)+ ... Of course you still have the problem with how + can apply to some of those items and the problem of how you would know that whichever item is +'d is the same one that was *'d ... so there would still be that ambiguity in the data.
I'm looking forward to continue this offensive discussion and offend you for being not offensive enough about this paper and the author. You said yourself I'm allowed to call you however I want in Message 307. I'm looking forward to a long offensive discussion. And don't forget the excrements which makes this place that special. Go for it. That still will not address the criticisms nor rescue the paper from the systemic problems that are pervasive throughout, from the "abstract" to the last appendix. Looking forward to your next criticism of the paper, and the insightful suggestions you can make to improve it further... especially as you have so much experience with it. Why don't you start with this:
If "A" represents the data from 76 episodes ... ... and"B" represents the pattern with all (24 of) it's variations ... the model developed to explain the data ... ... by what stretch of logic and imagination can you call this anything more than a rough approximation? Wouldn't you agree that the probability of any new episode fitting within B is much greater than the probability of it fitting within A? Or, by what rational or logic should B⊄A be included with B⊂A in the pattern\model used to explain A? Isn't that like painting a target around A on the side of a barn and then saying that anything that hits the barn is a bull's eye? And, by what rational or logic should 15 items be lumped into one element ("P.Al" = {>5 people, or green, or big, or wide, or "a lot," or "lack of knowledge," or "do nothing," or holiday, or very old, or starships, or standby, or science, or stone, or death, or #4} only 1 needed at a time) if it isn't done to hide the fact that 15 different episodes each had a different one of these distinct items, and the lumping is used to disguise the fact that they are different items being recorded? Why not list them separately unless the intent is to deceive? Enjoy ps - fill in whatever snark you like _____________________ such fun Edited by RAZD, : . Edited by RAZD, : .. Edited by RAZD, : ... Edited by RAZD, : .... Edited by RAZD, : /by our ability to understand Rebel☮American☆Zen☯Deist ... to learn ... to think ... to live ... to laugh ... to share. Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click)
|
|
|
Do Nothing Button
Copyright 2001-2023 by EvC Forum, All Rights Reserved
Version 4.2
Innovative software from Qwixotic © 2024