Who Helps?

  Guest Submission posted 8/28 by Randall Morrison, Exclusive to Footballguys.com

You are on the clock and only have 30 seconds left to make your pick. You know you need to draft your first QB. Among the available QBs, you see that John Kitna is projected to score 323 total fantasy points and Tony Romo is projected to score 321 total fantasy points. With only 2 points difference projected for the entire season, it appears the decision comes down to a coin toss or NFL team preferences. However, what if you could look at projected fantasy points for individual weeks rather than for an entire season?

How This Tool Evolved

I first thought of this idea last summer when reviewing two excellent resources on Footballguys.com: David Dodds' Offensive Projections and Clayton Gray's Strength of Schedule. I used my league's scoring system to convert the Offensive Projections to Projected Total Fantasy Points for the upcoming season. In reviewing the Strength of Schedule I realized that I could convert the Projected Total Fantasy Points to Projected Weekly Fantasy Points.

Here is a link to the Excel file.

Now, for example, the 2007 Offensive Projections for Kitna and Romo as of 7/19/2007 are:

The Weekly Strength of Schedule for Detroit and Dallas QBs as of 7/2/2007 is:

I then calculated the Projected Weekly Fantasy Points, which prorates the Projected Total Fantasy Points for each week of the season based on the Weekly Strength of Schedule data. The resulting weekly projections for Kitna and Romo are:

Notice Week 1, in which Dallas hosts the NY Giants, while Detroit plays at Oakland. Since the Strength of Schedule projects that the Giants defense is expected to allow QBs to score more fantasy points than the Oakland defense, it is projected that Romo will have a much better Week 1 than will Kitna.

Now that I had Projected Weekly Fantasy Points for each player, what could I do with this data? I needed to be able to quickly assess players during a live draft while under a time limit. I would not have time to look at each weekly score for all of the players I was considering. I decided to try to calculate the number of Weeks of Help each player provided, which would provide a single number for each player and thus allow quick comparisons of the available players. I started to look at the number of weeks in which a player scored above a set score for their position. I quickly noticed that the projected range of scores varied considerably from week to week, especially during bye weeks. I then decided to try to define Weeks of Help as the number of weeks in which the player scores in the top X players for their position. I decided to use three groupings: Good (blue bold), Fair (purple bold) and Poor (red italics). For QBs I defined Good as the top 8 scorers for a week, Fair as the top 16 scorers for a week and Poor as anyone who is 20th or worse in scoring for a week. I then assigned a weighting of 1.0 for Good, 0.5 for Fair and -0.1 for Poor. I also used Weekly Weightings in order to give more value to better projected performances in important weeks, such as the last week of the fantasy regular season (in case I need one last win to make the fantasy play-offs) and the fantasy play-offs. Based on these settings, the values of the Weekly Help for Kitna and Romo are:

(Note: My leagues only use 16 weeks, but the spreadsheet can accommodate 17 weeks if needed for your league.)

Finally, I had a single number comparing the value of all players on a weekly basis. I could now quickly determine that based on the projections that Kitna is a better pick since he helps in 10+ weeks, while Romo only helps for 7. In this case I would select Kitna.

Now that I have drafted Kitna, who would I draft later for a back-up QB? The first thing I had to do was ignore the QB Weekly Help values for Kitna's good weeks (5, 7, 10, 11, 12 and 14) since I already have my projected starting QB for those weeks. I modified the spreadsheet to automatically ignore the Weekly Help values for the weeks in which I already have sufficient Good starters at the position. For example, let's say I am ready to pick my second QB and the best options are Ben Roethlisberger and Eli Manning, both of whom are projected to score 304 total fantasy points. Although both players are projected to score the same amount of total points, looking at the Weeks of Help shows a difference. Without showing all of the calculations which were shown for Kitna and Romo (see the spreadsheet if you are interested in this data), the values of Weekly Help for these QBs are:

While both would provide help, Manning is more valuable since he has a Good rating in weeks 1, 3 and 15, as well as having a Fair rating for Kitna's bye in week 6. Notice that the help for week 15 has a weighting of 1.5 since this is the first week of the fantasy play-offs. Based on this data, I would select Eli Manning as my second QB rather than Roethlisberger.

Now that I had the calculations working to my satisfaction, I created a worksheet for each position. Now I could quickly see the best available player for each position, but wondered if I would have time to examine the worksheets for all positions during the draft to know who to draft next? To solve this problem I decided to calculate the top 10 available players for each position and display this data on the Top 10s Helps worksheet. Here is an example of the Who Helps worksheet that is based on the retained players for a keeper league in which I participate:

If this was the status of the spreadsheet when it was my turn to select a player, I could use this worksheet to quickly analyze the available players and make a selection. The highest values for Weeks of Help are at TE and Kicker, but I can easily see that there are plenty of other available players who can help me at those positions. Since I already have good players at QB and WR, I can afford to pass on some high quality players who are available, such as Palmer, Brady, Boldin and Harrison. Defense is not a real consideration for me at this point. It appears the best option would be to select McGahee. Even though his Weeks of Help value is only 7.7 and other players are available with a much higher value, I notice a significant drop-off in the Weeks of Help values for the remaining RBs and realize that the remaining help at RB is very limited at best. If I do not select McGahee at this point there will most likely be very little help remaining at RB for my next pick.

How to Use WhoHelps

I believe the best way to see the value of WhoHelps and to get a feel for what it does is to open the spreadsheet and practice going through several drafts. The good news is that although my discussion of how WhoHelps evolved is lengthy, actually using the spreadsheet is fairly quick and simple. The Instructions worksheet contains information to help you get started.

During a draft these are the basic steps I follow:

1. I use the Top 10s worksheet to see which players are my best options for draft picks.

2. I go to the position worksheet(s) (e.g.: QB, RB, etc.) for the players in which I am interested and quickly examine the data.

3. Once I have made my selection I enter the player's name in column A of the My section at the top of the position worksheets. For example, if I pick LaDainian Tomlinson I would enter his name in the first open line of the My RBs section of the RB worksheet, as shown:

While you can type the player's name, the best method is to type =, left-click on the player's name in the list of players in the weeksheet and then press the Enter key. Alternatively you can enter the cell reference for the player, such as =A45 for Tomlinson. Either method will cause the spreadsheet to automatically copy the player's name. Notice that once a player's name is entered, his NFL team and bye week, the number of Weeks of Help he provides and the Projected Weekly Fantasy Points for each week are automatically displayed. Also, the Need Help data above the Projected Weekly Fantasy Points is updated for each week. Notice that in this case I still need 2 Good RBs for week 7, which is San Diego's bye week, as well as 1 additional Good RB for most other weeks.

4. I also need to enter a Status Code of M in column C for the players I select, which indicates that this is My Player. In this case I would enter M in cell C45.

5. As players are drafted by other owners I enter a Status Code of D in column C to indicate that the player has been Drafted. You will notice each worksheet has a Legend for the Status Codes I use with WhoHelps. In addition to M and D, I also use A for Auction and K for Keeper when appropriate. With the exception of having to use M for players on your team, you are free to enter any code you want for players on another owner's team.

With practice you will be able perform these steps very quickly. In a redraft league in which I am the commissioner, last year I was able to use Draft Dominator, enter the draft picks on the league's web site for someone who was out of town and run the draft, while also using WhoHelps to help with my draft picks while on a time limit. And yes, I won both the regular season and play-offs in that league. The key was to practice before the draft and be very familiar with the steps I needed to use for WhoHelps.

As you practice with WhoHelps watch what happens as you assign players to your team and to other teams. If you are interested in how the spreadsheet works, all of the formulas are available for you to explore. The Instructions worksheet also contains important notes about the spreadsheet, such as Protection being on to prevent changes to key data during your draft, as well as a description of the color scheme used for the worksheets. Of course you can unprotect any worksheet, but I strongly recommend having the worksheets protected during a draft. This will help to prevent accidental data changes during your draft which could be difficult to correct under the time constraints of a fantasy draft.

If you are in a keeper or dynasty league another predraft step you should take is to enter all of the retained players before your draft. This will allow you to see the relative values of the available players and can help you to try several draft scenarios before your draft. Last year I did this for keeper and dynasty leagues, saved a copy with the keepers (so I would have a copy with the correct rosters) and then experimented adding different players to my roster to see the effect on my team.

In addition to helping me quickly make a draft pick based on who is already on my roster and the available players, this offseason I found another use for this spreadsheet. In a keeper league I was undecided about whether to keep Chester Taylor to use as a flex RB/WR/TE or release him to get extra auction dollars. I compared Taylor's Projected Weekly Fantasy Points with the three WRs I was keeping (Driver, Houshmandzadeh and Wayne) and found that his Projected Weekly Fantasy Points were always lower than all three of my WRs. This helped me decide to release Taylor and have extra auction dollars to try to get an upgrade at RB.

Customizing WhoHelps

In designing WhoHelps I tried to make everything fully customizable. All of the projections and data used for calculations can be customized as needed. For example, if your league starts 2 QBs rather than 1, you can change that value of the SetUp worksheet and the spreadsheet will automatically change the calculations. Most of the options on the Setup worksheet are discussed on the Instructions Worksheet. If you disagree with any of the Offensive Projections you are able to go to any projections tab (e.g.: QB Projs) and change the projections as you see fit. Of course you can also update the projections when the FootballGuys.com web site is updated.

In customizing WhoHelps for your personal use, the only feature that is not easy to change is the font characteristics for Good, Fair and Poor weekly scores and positive and negative values for Weeks of Help. The fonts are formatted using the Conditional Formatting option in the Format menu. Of course you can change the conditional formatting, but that would involve quite a bit of work.

Conclusion

The WhoHelps spreadsheet is a fully customizable drafting tool that dynamically adjusts to your roster and the available players as your fantasy draft progresses. Can it guarantee you a championship? No, it cannot. Drafting for fantasy football is part art and part science. WhoHelps can help with the science part. It is up to you to learn and apply draft strategies, roster management and other topics which compose the art portion of winning a fantasy football championship.

Will WhoHelps make every draft pick for you? No, it will not. It can show you the relative value of available players based on your current roster, season projections and schedules, but ultimately it is up to use to use this information to make the best pick for your team. For example, WhoHelps will not help you select the back-up for that high risk, high reward player you selected who has a history of injuries. That goes back to the art and science of a fantasy draft.

Finally, remember that the Offensive Projections and Strength of Schedule which form the basis of this spreadsheet are only educated guesses. Nothing in the WhoHelps spreadsheet is guaranteed to represent the actual player stats for this season. This spreadsheet is intended to help you make the best use of the Footballguys.com projections during your fantasy draft.