Check Out Our Shop
Page 1 of 2 1 2 LastLast
Results 1 to 25 of 41

Thread: Cute Analytics Exercise

  1. #1
    Join Date
    Mar 2006
    Posts
    3,452

    Cute Analytics Exercise

    From my econometrics prof. Sooooo nerdy, but I found it very amusing, especially the last paragraph

    http://www.clicktracks.com/insidetra...urce=nws072007

    Analytics According to Captain Kirk
    By Matt Bailey (Founder, SiteLogic)

    In my seminars, I enjoy teaching analytics because the fun is in finding effective and memorable methods to help people understand the concepts. One of my favorites is an analysis of the Red-Shirt Phenomenon in Star Trek.

    What? You don't know about the Red Shirt Phenomenon? Well, as any die-hard Trekkie knows, if you are wearing a red shirt and beam to the planet with Captain Kirk, you're gonna die. That's the common thinking, but I decided to put this to the test. After all, I hadn't seen any definitive proof; it's just what people said. (Remind you of your current web analytics strategy?) So, let's set our phasers on 'stun' and see what we find...

    The basic stats:
    The Enterprise has a crew of 430 (startrek.com) in its five-year mission. (Now, I know that the show was only on the air for 3 years, but bear with me. 80 episodes were produced, which gives us the data to build from.) 59 crewmembers were killed during the mission, which comes out to 13.7% of the crew. So, that will be our overall conversion rate, 13.7%.

    Data Segmentation:
    However, we need to segment the overall mortality (conversion) rate in order to gain the specific information that we need:


    Yellow-shirt crewperson deaths: 6 (10%)
    Blue-Shirt crewperson deaths: 5 (8 %)
    Engineering smock crewperson deaths: 4
    Red-Shirt crewperson deaths: 43 (73%)
    So, the basic segmentation of factors allows us to confirm that red-shirted crewmembers died more than any other crewmembers on the original Star Trek series.

    However, that's only just simple stats reporting - ready for some analysis?

    In-depth Analysis:
    Analysis involves asking questions about the data. Analysis attempts to bring reason and cause to the reported data in order to find why something is happening. With that data, one can improve the situation based on the intelligence gained from the analysis.

    Q: What causes a red-shirted crewman to die?


    On-board incident - 42.5%
    Beaming down to the planet - 57.5%
    There were also many fights during the mission; on the Enterprise, on planets, and various space stations. The fights were also divided between alien races or crazed crewmen (usually wearing red shirts).


    There were 130 fights over 80 episodes.
    18 of the 130 fights resulted in a fatality.
    13 of the 18 fatal fights resulted in a red-shirt fatality.
    Q: what was the rate of red-shirt casualties?

    18 red-shirt fatality episodes:
    8 multiple fatality occurrences; involving 34 red-shirted crewmen.
    9 single red-shirt fatality situations.
    It was found that red-shirted crewmembers tended to die in groups. In 17 red-shirt fatality episodes, 8 were multiple incidents, 9 were single incidents. In a little less than 50% of the fatal red-shirt situations, multiple crewmen were vaporized.

    Q: What factors could increase/decrease the survival rate of red-shirted crewmen?
    Besides not getting involved in fights, which usually proved fatal, the crewmen could avoid beaming down to the planet's surface, which is inherent to their end. However, that could result in a court-martial for failure to obey orders.

    Besides not beaming down, another factor that showed to increase the survival rate of the red-shirts was the nature of the relationship between the alien life and captain Kirk. When Captain Kirk meets an alien woman and "makes contact" the survival rate of the red-shirted crewmen increases by 84%. In fact, out of Captain Kirks' 24 "relationships" there were only three instances of red-shirt vaporization.

    The caveat to this is when Captain Kirk not only meets the local alien women, but also starts a fight among alien locals. The combination of these events has led to the elimination of 4 crewmembers (3 red-shirts).

    Here are the statistics:
    Red Shirt Death episodes = 18
    Episodes with fights = 55
    Probability of a fight breaking out = 70%
    Kirk "conquest" episodes = 24
    Kirk "conquest" + fights = 16
    Kirk "conquest" + red shirt casualty= 4
    Red shirt death + fight + Kirk "conquest" = 3


    And the data trends;
    Probability of a red-shirt casualty= 53%
    14% of fights ended in a fatality (with a 72% chance the fatality wore a red shirt)
    Probability of a red-shirt "incident" when Kirk has a "conquest" = 12%


    The red-shirt survival rate is slightly higher when Kirk meets women than when a fight breaks out. This trend necessitates the question: How often did Captain Kirk "meet" women? In 30% of the missions.

    As the data shows, Captain Kirk "making contact" with alien women has an impact on the crew's survival. The red-shirt death rate is higher when a fight breaks out than when Kirk meets a woman and a fight breaks out. Yet the analysis shows that meeting Kirk meeting women only happens in 30% of the missions.

    Conclusion:
    We can reliably improve the survivability of the red-shirted crewmen by only exploring peaceful, female-only planets (android and alien females included).

    Reporting the Data:
    Now, researching the data can be fun and informative. However, that is only half of the battle. The interesting part comes when you have to communicate not only the data, but your conclusions in an effective, persuasive manner. The best analysis won't go far if you can't communicate the conclusions in a manner that people understand.

    There are a few options at our disposal. First, the PowerPoint Method.


    There are a number of things wrong with the typical method of presenting data. For starters, this presentation could bore even the most hardened Starfleet manager (CEO). The typical corporate PowerPoint slide design is obnoxious and does not leave room for information, the charts are redundant, even unnecessary, and it does not do a good job of communicating the information or the analysis.

    In most cases, PowerPoint is NOT the recommended tool for communicating analytics data. It is not the right tool for the job. Communicating analytics data involves providing conclusions based on facts, tests, comparisons, and research. In order to display the necessary data, a better method must be used, and not one that forces redundant bullet point and "snazzy" charts.

    Visualizing the Data:
    There are some necessary elements required in developing a chart for this type data:

    A list of the specific episodes
    Events that happened in each episode

    The number of events that happened in each episode

    An easy way to identify data, then compare and contrast actions in all episodes

    By seeing all of the available data in one chart, associations, patterns and conclusions can be drawn simply by comparing the relationships as they are presented. This is something that I learned from Edward Tufte - 1. More information is needed to simplify data presentation. 2. Unless all of the data is presented, there is no data integrity.

    Information is Primary to Design
    This is critical in developing a chart of information - the information is primary. List the necessary data elements first. Then, develop the design around the information, and not the other way around. Otherwise, a beautiful chart will lack the critical information necessary to support your conclusions. The graphing software that I found extremely effective for communicating the episode data for this Star Trek analysis is Microsoft's Office 2007, and in Apple's OS X graphics software.



    click for full-size version
    I like this chart - eliminating the need for a legend is critical to allowing the information to flow. The data is the same color or object as the information we are trying to convey. Because there is no suitable color for Captain Kirk's affairs, we substituted a very flattering picture. Fights are represented by tiny phasers, which are not the best representation because of the size, but can easily be determined by the process of elimination. This chart allows conclusions and observations that simple charts, numbers, and explanations may never bring to the surface. It allows for easy comparison, both to other shirt colors, and in relation to other episodes. It also looks as though Kirk was a very busy man.

    In the first year of the series, red-shirt casualties were lower than other color-shirted crewmembers. The second and especially the third seasons were especially brutal. In the third season, only red-shirted crewmembers died; maybe because the other colors enacted better safety protocols, or maybe because they avoided the bridge when a new planet came into view, for fear of beaming down with Cpt. Kirk.

    Summary:
    Of the elements that helped to provide this analysis, segmentation was key. Segmentation of groups allows for comparisons. Comparisons allow you to spot trends that may be different from the rest. Asking questions of the data allows you to dig into specific trends and spot additional factors that affect the original analysis. Unless we dug into Kirk's personal life, we may never have spotted the contrast of Kirk's attraction to alien females as it related to saving red-shirt crewmen's lives.

    Remember, when you have to account for lost crewmembers, your report needs to account for the how, the why, and the ability to draw specific conclusions as to how to affect the trends in the future. Depending upon your approach, you could either doom the project, and future red-shirted crewmen, or you could be visiting planets full of peaceful alien women.

  2. #2
    Join Date
    Dec 2003
    Location
    Nhampshire
    Posts
    7,873
    How much of a nerd does it make me that I read that whole thing, understood it, and actually taught a similar session on Insurance Data. I should probably wedgie myself.

  3. #3
    Join Date
    Dec 2006
    Location
    crown of the continent
    Posts
    13,947
    Three econometricians went out hunting, and came across a large deer. The first econometrician fired, but missed, by a meter to the left. The second econometrician fired, but also missed, by a meter to the right. The third econometrician didn't fire, but shouted in triumph, "We got it! We got it!"
    Something about the wrinkle in your forehead tells me there's a fit about to get thrown
    And I never hear a single word you say when you tell me not to have my fun
    It's the same old shit that I ain't gonna take off anyone.
    and I never had a shortage of people tryin' to warn me about the dangers I pose to myself.

    Patterson Hood of the DBT's

  4. #4
    Join Date
    May 2006
    Location
    Flavor Country
    Posts
    3,032
    I didn't even read that whole thing but the fact that you used the word "cute" to describe an analytics exercise about Star Trek, suddenly makes me very attracted to you. I guess I must be a nerd at heart or something.
    "They don't think it be like it is, but it do."

  5. #5
    Join Date
    Oct 2005
    Posts
    11,811
    Quote Originally Posted by Tye 1on View Post
    Three econometricians went out hunting, and came across a large deer. The first econometrician fired, but missed, by a meter to the left. The second econometrician fired, but also missed, by a meter to the right. The third econometrician didn't fire, but shouted in triumph, "We got it! We got it!"
    Heh...funny

  6. #6
    Join Date
    Mar 2006
    Posts
    3,452
    you know, taking a second look at it, a cursory view of the data makes me wonder whether there weren't more red-shirt people in the first place.

    in other words, 73% death-rate among Red-Shirt crewperson is telling, compared to the 8% of the Blue-Shirt crewperson only if the gap does not simply reflect a difference in the make-up of the crew.

    any evidence of that?

  7. #7
    Join Date
    Apr 2005
    Location
    A Material World
    Posts
    1,668
    meh, not even a t statistic, let alone an F test.
    "Unfortunately, Meadows mgmt/marketing found out about the PR stash and published it on their trail map."

  8. #8
    Join Date
    Dec 2006
    Location
    crown of the continent
    Posts
    13,947
    Something about the wrinkle in your forehead tells me there's a fit about to get thrown
    And I never hear a single word you say when you tell me not to have my fun
    It's the same old shit that I ain't gonna take off anyone.
    and I never had a shortage of people tryin' to warn me about the dangers I pose to myself.

    Patterson Hood of the DBT's

  9. #9
    Join Date
    May 2004
    Location
    Colorado Cartel HQ
    Posts
    15,931
    Is this the bacon thread?

  10. #10
    Join Date
    Aug 2005
    Location
    WHEREAS,
    Posts
    12,936
    c,

    you are a bigger nerd than I am and that is sayign something.
    Quote Originally Posted by Roo View Post
    I don't think I've ever seen mental illness so faithfully rendered in html.

  11. #11
    Join Date
    Mar 2006
    Posts
    3,452
    Quote Originally Posted by BlurredElevens View Post
    Is this the bacon thread?
    boo... get out, don't cunt up my star trek statistics thread!!!

    I wonder what kind of results the study would have gotten if Picard would have continued Kirk's ahem...tradition? Probably would have to replace all the red shirts on the Enterprise every couple of years (Although the fact that Picard himself wore a red shirt sort of complicates matters).

  12. #12
    Join Date
    Oct 2005
    Posts
    11,811
    Quote Originally Posted by BeanDip4All View Post
    ...my star trek statistics thread...
    I think we reached the end of TGR. Are there any threads left to be had? I think with the recent Rangoon weather gambling and Star Trek statistical analysis, we might just have tied things up around here.

  13. #13
    Join Date
    May 2004
    Location
    Colorado Cartel HQ
    Posts
    15,931
    Quote Originally Posted by BeanDip4All View Post
    boo... get out, don't cunt up my star trek statistics thread!!!
    You need to go to The Hilton in Las Vegas.

    I'm guessing that would be your equivalent to boner city.

  14. #14
    Join Date
    Jun 2006
    Location
    Ventura Highway in the Sunshine
    Posts
    22,445
    Quote Originally Posted by BeanDip4All View Post
    you know, taking a second look at it, a cursory view of the data makes me wonder whether there weren't more red-shirt people in the first place.

    in other words, 73% death-rate among Red-Shirt crewperson is telling, compared to the 8% of the Blue-Shirt crewperson only if the gap does not simply reflect a difference in the make-up of the crew.

    any evidence of that?
    Give the girl a prize. That was the first thing I thought about, which is why I love statistics, with a little creativity you can make them reach any conclusion.

    "There are lairs, damn liars, and statistics."

    No, I am not going to watch 80 episodes of Startrek to find the answer.

  15. #15
    Join Date
    Mar 2006
    Posts
    3,452
    Quote Originally Posted by BlurredElevens View Post
    You need to go to The Hilton in Las Vegas.

    I'm guessing that would be your equivalent to boner city.
    actually, i was thinking how cool it would be to run a statistical analysis of the Stormtrooper effect, cause then we can have even more fun having super ULTRA nerdy Star Wars v. Star Trek STATISTICS debates.

  16. #16
    Join Date
    Apr 2005
    Location
    Sector 7G
    Posts
    5,660
    This is hysterical because I am a scientist (by training and title) and work with a bunch of engineers (and other scientists). The power point version of this is now printed and hanging outside my cube with the title, "Why it's good to be a scientist or and engineer."

    Everyone here is geeky enough to get it to....
    This is the worst pain EVER!

  17. #17
    Join Date
    Mar 2006
    Posts
    3,452
    Quote Originally Posted by Lonnie View Post
    The power point version of this is now printed and hanging outside my cube.
    ya know, i would have preferred a more detailed break down of the statistics for instance, where are the Yellow shirt stats?

    and the power point thing... shesh! create a mock MySQL db... or a LEAST use excel, man! :P

  18. #18
    Join Date
    Nov 2003
    Location
    Stuck in perpetual Meh
    Posts
    35,244
    Did somebody kill OGRE?

  19. #19
    Join Date
    Apr 2005
    Location
    A Material World
    Posts
    1,668
    Quote Originally Posted by Tippster View Post
    Did somebody kill OGRE?
    He was wearing a red shirt.
    "Unfortunately, Meadows mgmt/marketing found out about the PR stash and published it on their trail map."

  20. #20
    Join Date
    Mar 2006
    Posts
    3,452
    Quote Originally Posted by Tippster View Post
    Did somebody kill OGRE?
    Quiet you! I bet you are a Trekker yourself and just too afraid to admit it on a public forum... As a matter of fact I can see those model kits sitting in your office from here... the complete range of Starships Enterprise, from NX-01 to 1701-E!

  21. #21
    Join Date
    Apr 2005
    Location
    A Material World
    Posts
    1,668
    Isn't refridgerator spelled refrigerator?
    "Unfortunately, Meadows mgmt/marketing found out about the PR stash and published it on their trail map."

  22. #22
    Join Date
    Nov 2003
    Location
    Stuck in perpetual Meh
    Posts
    35,244
    So what WERE you doing in Vegas? Is Brett on to something?

    Do you own rubber Vulcan Ears?

  23. #23
    Join Date
    Apr 2005
    Location
    Sector 7G
    Posts
    5,660
    I sure it's been seen here 1,000x before, but worth posting again....

    Scottie, bringin' the POWER!


    Charlie Murphy RULES!
    This is the worst pain EVER!

  24. #24
    Join Date
    Mar 2006
    Posts
    3,452
    Quote Originally Posted by Tippster View Post
    Do you own rubber Vulcan Ears?
    Dude. I am like seriously offended... I am *so* much way more Klingon than Vulcan! Sure, they talk a big, fight-y game, but they mostly just go around yelling “kreplach!” at one another, suggesting that they are actually just Eastern European Jews, interested in getting together for a nosh

  25. #25
    bklyn is offline who guards the guardians?
    Join Date
    May 2005
    Posts
    5,762
    I'm just a simple girl trying to make my way in the universe...
    I come up hard, baby but now I'm cool I didn't make it, sugar playin' by the rules
    If you know your history, then you would know where you coming from, then you wouldn't have to ask me, who the heck do I think I am.

Similar Threads

  1. Cute Knut!
    By runethechamp in forum The Padded Room
    Replies: 8
    Last Post: 03-22-2007, 12:40 PM
  2. Can we get rid of cute chat?
    By AsheanMT in forum Ask TGR
    Replies: 2
    Last Post: 02-26-2007, 01:02 PM
  3. average heart rate vs calories
    By mntlion in forum General Ski / Snowboard Discussion
    Replies: 47
    Last Post: 04-26-2006, 01:12 PM
  4. I have a cute blond therapist
    By StokeSmack in forum Gimp Central
    Replies: 4
    Last Post: 04-02-2006, 11:14 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •