Author Topic: Likelihood Distribution of Number of SpaceX Launches in 2015  (Read 12380 times)

Offline Proponent

  • Senior Member
  • *****
  • Posts: 5472
  • Liked: 987
  • Likes Given: 653
As previously threatened, I'm going to solicit members' views as to the likelihoods of various numbers of SpaceX launches in 2015.  Of course, this does not preclude anybody from starting a traditional single-choice poll as well.  It may be interesting to compare the results of the two.

The advantages I see for this type of poll are 1) it imposes no preconceptions as to the number of flights, and 2) it elicits more information than does a single-choice poll.  We'll see how it goes.

Rules:

1. For ease of tabulation, please use precisely the following format:
Quote
<forum name>: n pn pn+1 pn+2 pn+3 ... pN

<(optional) rationale>

where n is the minimum number of launches to which you assign a non-zero likelihood, pk is 100 times the likelihood assigned to k launches (e.g., a likelihood of 10% should be expressed as "10" [without the quotation marks]), N is the maximum number of launches having non-zero likelihood, and the other two fields are self-explanatory.  Please note that a colon terminates the forum name and that subsequent entries are separate by spaces, not commas or other punctuation.  Please leave a blank line between the first row and any explanatory text.

For example, suppose I think there's a 10% chance of 6 flights, 25% for each of 7-9, 15% for 10 and 5% for 11.  Then my entry would look like

Quote
Proponent: 6 10 25 25 25 15 5

They're probably gonna do better than last year, but I'm not expecting them to completely blow barn doors off.  If nothing else, for the time being there is still just one pad.  And I'm betting that Falcon Heavy isn't going to fly until 2016.

2.  I reserve the right to disregard entries not precisely in the above format as well as to interpret non-conforming entries in what seems a reasonable way.

3.  Likelihoods should sum to 100.  If they don't, they'll be normalized to 100 at my discretion (e.g., if they sum to 99, I may simply normalize; if they sum to 47, I may ask for clarification).

4. I commit tabulating up to 400 entries.  That's not to say I won't process more, but there has to be some limit.  Of course, everybody else is free to process the entries as may seem appropriate.

5. At the close of 2015, I will score each entry according the likelihood assigned to the number of flights which actually occurred.  I will post the results.

6.  A launch shall be deemed to have occurred whenever SpaceX Falcon intended for orbit or beyond leaves the pad in 2015 (GMT).  Whether is successful or not is irrelevant (that way we avoid disputes about the definition of "success").

7.  Entries will be accepted until 15:28:53 GMT 19 January 2015.  Once posted, entries may be modified until the poll closes.

8.  One post per member, please, until the poll closes on 6 January.  Please post questions or comments here (or PM me) rather than posting in the present thread.  This will make tallying the entries easier.

Go for broke with a single number, or play it safe with a nice, broad range -- it's up to you!
« Last Edit: 01/05/2015 08:47 PM by Proponent »

Online Lar

  • Fan boy at large
  • Global Moderator
  • Senior Member
  • *****
  • Posts: 10434
  • Saw Gemini live on TV
  • A large LEGO storage facility ... in Michigan
  • Liked: 7280
  • Likes Given: 5082
Lar: 0 0.05 0.1 0.2 0.45 0.8 1.4 2 4   5 7 9 11   12 14 12   10 6 3 1   0.5 0.5

Rationale: My lower bound is zero. There is a very small, but non zero probability that CRS-5 will fail in such a way that the entire year is shot (note that under Proponent's criteria the recent Antares/Cygnus failure counts as a launch... in order to hit 0 CRS-5 has to "fail" before it even is ignited, and fail so spectacularly that they call off all their launches). I think this a very low probability so I assigned it a .05% chance. The range is broad, it goes up to 20 but the peak is around 12-14 launches. That's what I know they are striving to hit.  The probability of more than 15 starts to fall off really rapidly so I stopped at 20

I fudged a few of the estimates to get it all to come out to 100 (I probably should have pre normalized)

Here's a chart the way Skybum did it

flights      prob        Cum         At least
 0       0.05       0         100.00%
 1       0.1        0.05       99.95%
 2       0.2        0.15       99.85%
 3       0.45       0.35       99.65%
 4       0.8        0.8        99.20%
 5       1.4        1.6        98.40%
 6       2          3          97.00%
 7       4          5          95.00%
 8       5          9          91.00%
 9       7         14          86.00%
10       9         21          79.00%
11      11         30          70.00%
12      12         41          59.00%
13      14         53          47.00%
14      12         67          33.00%
15      10         79          21.00%
16       6         89          11.00%
17       3         95           5.00%
18       1         98           2.00%
19       0.5       99           1.00%
20       0.5       99.5         0.50%
                                0.00%
         


« Last Edit: 12/19/2014 09:04 PM by Lar »
"I think it would be great to be born on Earth and to die on Mars. Just hopefully not at the point of impact." -Elon Musk
"We're a little bit like the dog who caught the bus" - Musk after CRS-8 S1 successfully landed on ASDS OCISLY

Offline saliva_sweet

  • Full Member
  • ****
  • Posts: 555
  • Liked: 424
  • Likes Given: 1345
saliva_sweet: 4 100

Keeping it simple and pessimistic.
« Last Edit: 12/19/2014 02:37 PM by saliva_sweet »

Offline Garrett

  • Full Member
  • ****
  • Posts: 1108
  • France
  • Liked: 111
  • Likes Given: 77
Garrett: 6 10 20 24 21 14 8 3

rational: no crazy rational. I just hope they'll do at least 6 next year, so that's why I started with that. Last year I said 8, so I'm just hoping for that again, so it gets to be the maximum. I think 12 would be sweet, as it would be an average of 1 per month and that's as far as my expectations go. Also, I played with the numbers so that the distruibution would look kinda Maxwellian ;) (see attached image)

 
- "Nothing shocks me. I'm a scientist." - Indiana Jones

Offline Jimmy Murdok

  • Full Member
  • *
  • Posts: 154
  • Tokyo - Barcelona
  • Liked: 92
  • Likes Given: 85
Jimmy Murdok: 6 5 12 18 18 20 15 12

Edit/Lar: fix to match requested format.
« Last Edit: 12/19/2014 03:13 PM by Lar »

Offline averagespacejoe

  • Full Member
  • **
  • Posts: 201
  • Liked: 51
  • Likes Given: 0
averagespacejoe: 4 5 7 10 10 10 15 30 10

This is a bizarre format that shows me why I was never an engineer. However here it is.

Edit/Lar: Fix to match requested format
« Last Edit: 12/19/2014 03:19 PM by Lar »

Offline skybum

  • Member
  • Posts: 40
  • London, UK
  • Liked: 59
  • Likes Given: 43
skybum: 2 1 1 2 4 6 9 13 20 30 10 3 1

Or expressing this another way:

Number of flightsprobability of achieving
0100%
1100%
2100%
399%
498%
596%
692%
786%
877%
964%
1044%
1114%
124%
131%

Offline CraigLieb

  • Full Member
  • ****
  • Posts: 859
  • Dallas Fort Worth
  • Liked: 779
  • Likes Given: 974
CraigLieb: 3 1 1 2 6 17 22 25 16 7 2 1

Rationale:
Most likely 9 with more rounded on the low side with a more severe drop off in probability on the high side.
Like showing the chart. I am a fan of bell curves having been a Six Sigma Black Belt, but I don't think this is a situation where there are equal probabilities around the mean.  It will be more skewed to the low side due to "space is hard", "abundance of caution".. etc. Hopeful that they could swing for the fences and hit 12! can't bet that will be the result. 

Edit/Lar: Fix format to match requested format. and yes, I am getting tired of doing this.
« Last Edit: 12/19/2014 04:26 PM by Lar »
Colonize Mars!

Offline toruonu

toruonu: 4 3 4 8 11 14 15 13 10 8 6 4 2 1 1

It derives from Poisson distribution for 9 flights, but gets a manual modification trunkage at the higher end shifting some of the probability to lower end due to space being hard etc.
« Last Edit: 12/19/2014 05:21 PM by toruonu »

Offline mheney

  • The Next Man on the Moon
  • Global Moderator
  • Full Member
  • *****
  • Posts: 779
  • Silver Spring, MD
  • Liked: 378
  • Likes Given: 200
mheney:  7   2 4 6 10 20 25 20 10 2 1

Figuring 50% growth in flights per year is a good pace; I've got 85% of my money on 10-14 flights.
I'd be happy to be wrong on the low side.


Offline Tass

  • Full Member
  • ***
  • Posts: 361
  • Liked: 76
  • Likes Given: 153
Tass: 0 0.1 0.2 0.5 1 2 4 8 12 18 23 18 8 5 2 0.9 0.2 0.1

Offline nadreck

nadreck: 0 .5 .5 .5 2 3 5 5 5 5 5 5 10 10 10 20 10 2 1 .5


[edit]Ok Ok, rationale:  Yup I agree that they might stop for a whole year if there is an Anteres event and I consider that a .5% probability right now, and if they launch another 10 in a row without a bauble it is down to about .2 - .3%. Another thing is that while I think they will make 13 launches next year, each successive launch beyond that is really a strain on human (and other) resources for where the company is at right now. By the time they get out to 15 I think payloads will also be an issue. Which is why my odds drop so dramatically at the end. I do think they may launch an essentially "free" bird using a recovered core in the 2nd half of the year and that it will contain either a SpaceX test payload, an Elon or Elon favoured payload for free or cheap, or a "deal" payload for some company that had a standby bird to cover for a failed launch from some other carrier. [/edit]

[2nd Edit] fix my math [/edit]
« Last Edit: 12/19/2014 08:13 PM by nadreck »
It is all well and good to quote those things that made it past your confirmation bias that other people wrote, but this is a discussion board damnit! Let us know what you think! And why!

Offline topsphere

  • Full Member
  • *
  • Posts: 132
  • Liked: 66
  • Likes Given: 159
topsphere: 3 2% 2% 2% 3% 6% 9% 11% 15% 19% 13% 11% 3% 2% 2%

rationale - I figure +-2 from 11 will be very likely, but anything above 13 pretty unlikely. Couldn't be arsed to make it look like a symmetric distribution....
« Last Edit: 12/19/2014 08:18 PM by topsphere »

Online NaN

  • ... a free man
  • Full Member
  • *
  • Posts: 146
  • Liked: 80
  • Likes Given: 116
NaN: 0 0.04 0.08 0.16 0.32 0.64 1.28 2.56 5.12 10.24 18 25.96 17 10.24 5.12 2.56 0.64 0.04

Rationale: 9,10,11 carries the bulk of the likelihood. 7-13 as outside possibilities. Beyond that, a longer tail on the bottom than on the top since anything going wrong is more likely than everything going right.

Offline Billium

  • Member
  • Posts: 62
  • Winnipeg Canada
  • Liked: 36
  • Likes Given: 18
Billium: 7, 5.10.10.15.15.15.10.10.5.5

Rational - we have insufficient information to make an accurate prediction buts it's very reasonable to guess 2015 will have more launches than 2015.
« Last Edit: 12/20/2014 04:45 PM by Billium »

Offline cscott

  • Senior Member
  • *****
  • Posts: 2949
  • Liked: 2053
  • Likes Given: 664
cscott: 2 0.6 0 0 0 1.9 4.5 11.5 13.5 15.4 11.6 19.2 7.1 7.7 4.5 1.9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.6

Rationale: wisdom of crowds.  This matches the distribution of the votes in the "How many SpaceX flights do you anticipate in 2015?" as of this moment (2014-12-19 22:20 EST)... and happens to match my intuition as well (ie, broad peak between 9 and 12 launches).  For the sake of science (science!) I even included the "more than 35+" joker, including their vote as a vote for 36 launches.

Online aameise9

  • Member
  • Posts: 49
  • Potsdam, Germany
    • Cognitive Biology
  • Liked: 13
  • Likes Given: 55
aameise: 6 1.7 3.4 5.8 8.6 11.1 12.8 13.1 12.1 10.1 7.7 5.3 3.3 1.9 1.0

Rationale: Treating each of 50 weeks as independent events and taking the expected mean to be 12, I obtain a binomial distribution with parameters N = 50 and P = 12/50.

I think it would have been sufficient to guess mean and variance ...

Offline arachnitect

  • Member
  • Full Member
  • ****
  • Posts: 1440
  • Liked: 387
  • Likes Given: 462
arachnitect: 1 0.5 1 1.5 2 3.5 6 10 19 30 21 4 1 0.5

Offline sdsds

  • Senior Member
  • *****
  • Posts: 5479
  • "With peace and hope for all mankind."
  • Seattle
  • Liked: 579
  • Likes Given: 677
sdsds: 1 1 2 3 4 5 6 7 8 10 11 12 10 8 6 4 2 1
-- sdsds --

Offline MikeAtkinson

  • Full Member
  • ****
  • Posts: 1779
  • Bracknell, England
  • Liked: 542
  • Likes Given: 70
MikeAtkinson: 2 0.1 0.1 0.2 0.6 2 4  7 10 11 12 13 11 10 9 5 2 1  1   1

Low probability of something going really wrong and SpaceX only managing a few flights.

Not much to choose from 9-15 launches, 15 seems the maximum number they can realistically do now. Payload availability and other external influences will likely be the limiting factor.

Low probability of more than 15 launches due to possible multiple reflights of recovered stages, perhaps with customers paying a very low price.


Tags: SpaceX 2015 Predictions