Draft Loadout Features: Trade Calculator

I have gotten lots of requests for this tool so I finally sat down and figured out a way to implement it!  Here is my fully featured Trade Calculator.  Supports trades up to 6x6 between 2 teams as well as trading draft picks.  This is the only trade calculator I have run across that is dynamic to your personal scoring system and rankings.

Here it is in action:

1. Open up your app and make sure you have already setup a scoring system and rankings system you would like to use.  Make sure you have selected them prior to going to the tool or it wont know what to use.

 

2. Add your players or picks to both teams.

 

3. It will determine the winner of the trade at the bottom.

 

Notes:

  • This app is based on what a player is likely to do in 2017 based on where you have them ranked and your scoring.  If you are in a Keeper/Dynasty League it is still a great tool, but you may need to spend a little more time.  If you are grabbing a rookie like Chad Williams you could enter him as John Brown instead to do the trade based on what you project him to be when you plan to utilize him.  Similarly you could enter John Brown as Larry Fitzgerald for the same reason.

  • If you are trading rookie picks or future picks I recommend finding yourself a source for Dynasty ADP and figure out where rookies are being drafted in a full startup draft.  Use the ADP from where rookies in that stage of the draft are being taken as your "pick" in the tool.

 Some Examples:

** All of these examples are done using 12 man PPR settings and generic ADP based Rankings ** 

 

 

 

How it works: (About to get in the weeds.  Math nerds only beyond this point...)

 

After you go through the whole process of establishing a league scoring system and customize the player rankings to suit them you are ready to go!  Please see the tutorial article for the software on how to do this.

 

I needed a way to somehow translate what players relative values across positions based on who is starting them.  This is what I came up with:

 

  1. I first used the software to spit out the Value-Over-Replacement Draft board into excel and charted out the Value measures to see how they were distributed.

  2. Since the values are based off of players you are starting for optimization purposes, once you get into bench players beyond them the values are below zero.  This may seem odd, but all we are doing with these is comparing each player to the player at that position that you can grab as your last starter.

  3. Since we are developing trade values negative ranges don't really make sense so I applied a vertical shift by subtracting the lowest value in the series from every item.

  4. Now because every scoring system is unique and this numbers could very wildly I decided to scale the dataset so the maximum value peaked at 10,000.  This is done by multiplying every value by 10,000 / shifted maximum value.

  5. We are starting to look pretty good.  We have shifted and scaled a logarithmic value distribution without altering its shape.  No matter what settings you use your values will have a range of 0 to 10,000

  6. I did some testing at this point and realized that I needed something else.  Bench players values were coming out way to high relative to the elite tier player.  Since in Fantasy Football all we really care about it who you are starting I decided to dig on MFL's site for there % started statistics.  I found the percent started measure for the top 200 fantasy players and charted them out to see what the trend was.  It turned out to be a 2nd Order Polynomial.

  7. I created a % started measure for every ranked played in my database extrapolated from MFL's data and multiplied each trade value time the % of owners that started the player at that rank in the previous year.

The results came out really well I think.  Please let me know if you have and comments/feedback!

  

Enjoy!

 

 

 

Please reload

Featured Posts

Week 11 DFS Targets

November 16, 2019

1/6
Please reload

Recent Posts

November 16, 2019

November 10, 2019

November 3, 2019

November 2, 2019