OK, the eighth release is ready.  It contains 37 keys.

Version 5 fixed the issue where users with older versions of Excel were getting #REF errors in the Most Likely field (Thank you to unborn who found the root cause of the problem).  Version 4 fixed the way the Best and Worst possible scores are calculated.  I had an issue as the range was too wide which was caused because I forgot to change something when I adapted this from the 2013 Sergeants test that had 30 research questions.  The Most Likely score is unchanged.

For those who are new to this, the spreadsheet works as follows:

Download the Excel file linked below.  Once you open it, enter your own answers in the space provided on the left.  The spreadsheet breaks down the test as follows:

1) It assigns a "Confidence Level" to each question.  High means that a vast majority of people selected that answer.  Moderate means that it is close but the top answer is still decently above the others.  Low means that it is still a potential double answer, or too close to call.

2) Your correct and wrong answers are broken down by these confidence levels, so you can easily see how you did on the questions that are most likely accurate and the others that are less so.

3) Potential double answers are also counted and how many where you have at least one.

4) Wrong answers are colored red on the left.  Doubles where you have the other one is colored orange.

5) The biggest thing here is the potential score.  I broke it down into three categories: Best Possible, Worst Possible, and Most Likely.  The Best and Worst are the two extremes (such as having all 20 of the research questions as wrong, getting all doubles; or having 20 research as right and losing doubles respectively).  The most likely scenario uses some general probability to determine how many of the research questions and doubles will affect you negatively.

Historically, this key ends up being about 90% accurate with the final published answer key from DCAS.  I try to account for a margin of error so the Most Likely score you see should be very close to what you actually get when the proposed key comes out.

http://www.ottrack.com/downloads/AnswerKey-2014-Capt.xlsx

Thanks for doing the 2013 Sgts answer key bro, it was dead on!

PM ed it to you. Thanks again.

Come on people, where's the excitement? The last captain's test, I had about 15 keys on the first night. Does everyone really want to wait 5 weeks? You should be impatient like me.

Let's go...less drinking, more PMing.

Lets go

I just PMed you

Here's another

I have pm you but don't know how to take it off from here

thanks, btw, how do you account for the 20 that do not count?

Answer sheet attached

EveryoneLovesChris wrote:

---

Question...are you getting answer keys from hardcore studiers only? I didn't open the book do you still want mine? PS what kind of nerd comes up with a spreadsheet like this?!? Lol just kidding....but seriously....

---

I'll take keys from anyone.  Really bad answers will get drowned out by the majority and those that you got right will reinforce the selection.  As for the type of nerd, it takes a special kind.

stillwaiting wrote:

---

What's the percentage of the 20 questions in your opinion so that I could plug it in and compare to the difference with the other keys

---

I'm not sure what you're asking.  The spreadsheet already factors in how many research questions will likely affect you.

I am still trying to figure out my score based on Gamblor keys can somebody please enlighten me. My breakdown is I got 34 different answer breakdown like this:

High=22

Low=4

moderate=8

Do you have another way to download the file because for some reason I can't access the link?

I just pm'd you gambler, nice job btw

I'm at most likely 69 with 4 possible doubles. ..I am the definition of on the fence...but man would that such if he's dead on. Let us pray

Gamblor wrote:

---

Basically, I use a percentage of the 20 questions against the total you have wrong to make a statistical guess on how many of your wrong answers would not actually count. Depending on how many you have wrong, the number applicable adjusts. It's mostly probability, so there is a margin of error, but it's fairly accurate. From the 2013 sergeants exam, where they had 30 research questions, my estimates of individuals' most likely score was either spot on, or one or two off.

---

hey. what's the formula for the most possible score? my worksheet took it out for some reason. i can get my most possible score. 

-- Edited by TrYAGaiN on Monday 15th of September 2014 12:13:05 PM

TrYAGaiN wrote:

---

hey. what's the formula for the most possible score? my worksheet took it out for some reason. i can get my most possible score. 

-- Edited by TrYAGaiN on Monday 15th of September 2014 12:13:05 PM

-- Edited by TrYAGaiN on Monday 15th of September 2014 12:36:02 PM

---

 If you have an older version of Excel, it may sometimes have an issue.  You can send me your answers and I'll run them for you.

My worst possible changed drastically in a good way. Any explanation for the adjustment?

wereinbacklog wrote:

---

My worst possible changed drastically in a good way. Any explanation for the adjustment?

-- Edited by wereinbacklog on Monday 15th of September 2014 11:56:47 PM

---

 I still had 30 research questions factored in on the best and worst possible fields (I adapted the sheet I made for the 2013 Sergeant's exam).  I changed it from 30 to 20 everywhere else, but forgot about those two.  The change brought them both in by 10, closer towards the most likely score.

Thanks Gamblor.  

Hey Gamblor, here's two more answer sheets.

Here is mine!

Twenty N Out wrote:

---

Thank you Gamblor for all your hard work on this spreadsheet. My most likely puts me at barely passing right now. I hope it holds true. I have a question. Does DCAS have to identify which questions were for research before they administer the test?

---

 I have no idea what DCAS' procedure is for identifying research questions.  I know that we don't find out until the protest review session.