Home | Blog | Software | Reviews and Features | Forum | Help | Donate | About us
topbanner_forum
  *

avatar image

Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
  • December 08, 2016, 12:20:36 PM
  • Proudly celebrating 10 years online.
  • Donate now to become a lifetime supporting member of the site and get a non-expiring license key for all of our programs.
  • donate

Author Topic: IDEA: Discrepancy Trainer  (Read 4575 times)

mouser

  • First Author
  • Administrator
  • Joined in 2005
  • *****
  • Posts: 36,412
    • View Profile
    • Mouser's Software Zone on DonationCoder.com
    • Read more about this member.
    • Donate to Member
IDEA: Discrepancy Trainer
« on: November 15, 2009, 10:43:26 PM »
Ok i'm trying to post smallish ideas here, in hopes people will find something they like to do for their NANY 2010 projects.

Here's an idea i had a while ago for a utility designed to help people become more accurate and realistic about their predictions and promises in terms of deadlines.  Most coders I know underestimate the time it will take to get something done, or how much we can get accomplished in a given period, etc.  At the start of the day it feels like you can achieve anything.. at the end of the day you wonder where the time went..

So, here's the idea:

Program shows a list view with columns for:
Item description, Measurement description, Predicted Value, Actual Value, Accuracy

I'll give you an example:
I start out by adding a new item with the following column values:
  • Item description = Finish reading new book on Planning Algorithms
  • Measurement description = Number of days to finish (starting on Nov 15, 2009)
  • Predicted Value = 21

That says i hope to finish this book in 21 days.

Then when i actually finish the book i go back to the program and add how long it really took:
  • Actual Value = 63

Then the program can calculate the last column for me:
  • Accuracy = 30% (shown in red since i took longer than i said, meaning bad)

You dont have to use days, thats why i have a separate column for measurement description.  You could instead have an entry for how many pages of a novel you can write in one week for example, or how many pounds you can lose on a diet in one month, etc.



Other bonus features:

1) show a running average of accuracy at top.
2) have view modes for: show all, show incomplete, show complete
3) add a date-added column
4) allow soring on columns

JavaJones

  • Review 2.0 Designer
  • Charter Member
  • Joined in 2005
  • ***
  • Posts: 2,717
    • View Profile
    • Donate to Member
Re: IDEA: Discrepancy Trainer
« Reply #1 on: November 15, 2009, 10:51:08 PM »
Cool idea, and useful for more than just programmers, to be sure!

- Oshyan

nudone

  • Cody's Creator
  • Columnist
  • Joined in 2005
  • ***
  • Posts: 4,117
    • View Profile
    • Donate to Member
Re: IDEA: Discrepancy Trainer
« Reply #2 on: November 16, 2009, 04:54:13 AM »
yes, very interesting. i think i'd find it useful too.

SleepingWolf

  • Supporting Member
  • Joined in 2007
  • **
  • Posts: 107
    • View Profile
    • Read more about this member.
    • Donate to Member
Re: IDEA: Discrepancy Trainer
« Reply #3 on: November 16, 2009, 03:25:51 PM »
Great idea, perfect for students to test their heavy mettle! One question, how are you getting the above accuracy? I'd have thought it should be (63/21)*100 = 300%, not 30%. But I'm no expert, just guessing here.
-------------------------------------------

Choose life -or at least respect it.

mouser

  • First Author
  • Administrator
  • Joined in 2005
  • *****
  • Posts: 36,412
    • View Profile
    • Mouser's Software Zone on DonationCoder.com
    • Read more about this member.
    • Donate to Member
Re: IDEA: Discrepancy Trainer
« Reply #4 on: November 16, 2009, 07:30:45 PM »
i just did the reverse (21/63) but there are lots of functions you could use, as long as they were heuristics for accuracy.

skwire

  • Global Moderator
  • Joined in 2005
  • *****
  • Posts: 4,668
    • View Profile
    • Donate to Member
Re: IDEA: Discrepancy Trainer
« Reply #5 on: December 31, 2009, 05:34:34 AM »
I have this done but am taking requests for the final name of this before posting it.  So...suggestions, please.

cranioscopical

  • Friend of the Site
  • Supporting Member
  • Joined in 2006
  • **
  • Posts: 4,367
    • View Profile
    • Donate to Member
Re: IDEA: Discrepancy Trainer
« Reply #6 on: December 31, 2009, 06:30:57 AM »
Hone-est?
Reform's Cool?
Adumbrate?
Goal Keeper?
Bestimate?
Great Expectations?

skwire

  • Global Moderator
  • Joined in 2005
  • *****
  • Posts: 4,668
    • View Profile
    • Donate to Member
Re: IDEA: Discrepancy Trainer
« Reply #7 on: December 31, 2009, 07:26:52 AM »
I like Bestimate. 

Ath

  • Supporting Member
  • Joined in 2006
  • **
  • Posts: 2,783
    • View Profile
    • Donate to Member
Re: IDEA: Discrepancy Trainer
« Reply #8 on: December 31, 2009, 09:07:04 AM »
+1 for Bestimate

Paul Keith

  • Member
  • Joined in 2008
  • **
  • Posts: 1,982
    • View Profile
    • Donate to Member
Re: IDEA: Discrepancy Trainer
« Reply #9 on: December 31, 2009, 09:21:29 AM »
Yep, I also had a recent idea like this but I dropped it because I don't think it would work.

...or at least not any different than any other generic Project Manager.

Sorry if I put a damper on the topic. As others also said, there are kinda clones of the concept out there.

Anyway, this isn't really a contribution but maybe someone can get something out of the concept.

My idea is the same as mouser's except before you do this, you create a "value for your life" rating which involves putting in your birth date and your age expectancy (what age do you expect to die) and your current age.

The program then chucks out a generic value based on some pre-configured statistics; maybe the total value being the overall amount of seconds in your life.

I didn't really had a number in mind but I remember playing a pirated version of Real Life back then (sorry I couldn't find the official website) and it was a game where you could for example play as a poverty class Ethopian and the game will give you a list of things happening in your life from childbirth to death based on a series of scenarios.

The thing was repetitive in that you can often find yourself getting the same disease but I thought it showed how life statistic is possible to an extent.

I thought without all these additional data like your nationality and your class, the childbirth and life expectancy data could be minimized to the point of being a compact program no different than a timer. (as an additional feature, you can tweak this value by adding in your expected "necessity time" which reduces the total value of your life)

Now where this concept differs with mouser's idea is that instead of a predictor rating you as bad or good, you put in tasks and once you finish them, it deduces your "life value" based on the value of the action.

The idea is that it's supposed to represent how much that task is taking away from your life.

It's similar though in that the input method is nearly the exact way you would input a task in mouser's idea.

The difference is that there's no good or bad rating.

Instead what you have is an optional enjoyment value or a "how much did this finished task add value to my life".

This would be set by a predefined value in relation to your life value via asking a series of Barnum statements regarding your perception life based on hrs./min./sec. (depending on what you want to input) and it would divide these numbers from your total life value.

The end result being a generic value for how much your "max enjoyment is".

This value in turn impacts how much a finished task will add to your life.

Ex.

Let's say your enjoyment value is 256 min.

Now you finished a task that is worth... 120 min.

It will reduce the life value by 120 min. It doesn't impact the enjoyment value. (...or if you really want, you can optionally star the entry so that instead of the total min. reducing your life value, you get 256 - <task value> = only the value of the sum is reduced from the life value.)

However let's say the value of your task exceeds 256. Let's say it's 259. Then 259 - 256 = it adds 3 min. to your life value.

Again like mouser, min. here is just an arbitrary value. The concept isn't so much to track your life but to record the truly positive and the truly negative anchors of your life so that you can use these as an indicator to self-simulate this:

Quote
An unusual use of anchoring was studied by Ellen Langer in her study of two groups of 75-80 year old men at Harvard University. For 5 days, both groups were isolated at a retreat, with one group engaged in a series of tasks encouraging them to think about the past in general (to write an autobiography, to discuss the past etc), and the other group engaged in a series of tasks which anchored them back into a specific past time - they wrote an autobiography up to 1959, describing that time as "now", watched 1959 movies, had 1959 music playing on the "radios", and lived with only 1959 artefacts. Before and after the 5 days, both groups were studied on a number of criteria associated with aging. While the first group stayed constant or actually deteriorated on these criteria, the second group dramatically improved on physical health measures such as joint flexibility, vision, and muscle breadth, as well as on IQ tests. They were anchored back physically to being 50 years old, by the sights and sounds of 1959. (Langer, "Mindfulness", Addison Wesley 1989)

http://en.wikipedia..../wiki/Anchoring_(NLP)




skwire

  • Global Moderator
  • Joined in 2005
  • *****
  • Posts: 4,668
    • View Profile
    • Donate to Member
Re: IDEA: Discrepancy Trainer
« Reply #10 on: December 31, 2009, 10:57:25 PM »
Please direct any replies or suggestions to the NANY post for Bestimate here: http://www.donationc...ex.php?topic=21219.0