Our Algorithm

First of all, thanks for coming to the site. With the thousands of other fantasy football sites out there, we appreciate you coming to ours.

To make it worth your while, we are committed to constantly evolving the content to help provide you with the data you need to win your league. And all of it is free since 2007.

If you take a moment to browse the site, you can see that we specialize in “Strength of Schedule” data. Its a polarizing concept in the preseason, but no one can argue its value during the season. Here at fantasyomatic, we took it to a new level.

Listen to the explanation here:

The Fantasyomatic fantasy football articles on this site are primarily based on a proprietary engine we call our “Fantasyomatic Algorithm™”. We conceptualized the algorithm in 2001 and brought it to life on this site in 2007.

Although poor imitators have cropped up since we broke the seal, none of them use our exclusive approach to fantasy football predictions. We are the only methodology that uses exclusive Player, Defense and Matchups “Ratings” we refer to “FoM™” Ratings. These ratings have helped us decrease prediction errors by up to 30% against NFL data since 2009. We work on the algorithm non-stop to continue to add new variables to make it more and more accurate all the time.

The concept is based on a belief that a player will do better than average against a defense that is weak against their position and worse than average against a defense that is strong against their position.

This is commonly called “Strength of Schedule” (SOS). You will see this talked about for overall team strength of schedule but that does not apply here. For an NFL Team, SOS can determine if they will have an easy road to wins. But lets face it, in Fantasy you don’t care if the TEAM wins. You want to see if your PLAYER is going to well in a game and over a season.

You may already subscribe to this theory and not even know it. Think about all the sites you visit throughout the season. The START/SIT information is almost completely about what defense a player faces that week. Specifically how well that defense does against their position. Of course, the player’s capability to score points is the other factor.

If you look at a given game you can easily see that a player could do well if they play a weak defense against their position. But our engine also aggregates that information across an entire season (for drafting), remaining season (for waivers and trades) and weekly (for starts and sits).

On every fantasy football website, tv show or radio show during the season, start/sit advice ALWAYS boils down to who your player is playing that week. This tool puts that data in your hands before anyone else.

How To Use This Data

Using our preseason data, you can determine:

Using our in season data, you can determine:

  • What players have the best weekly matchups
  • What players have the most value as waiver pickups
  • What players are the best BYE week fillers
  • What players you should trade for and what played you should trade away

What you should do with this data:

  1. We do not advocate only drafting players because of their SOS. Instead, use this information to help as a tiebreaker when deciding between players in the same tier or relative draft position. Never reach out of a player’s tier just because of SOS. Our Draft Board shows players we like for various reasons and we will make recommendations on players to take before their ADP and we take many factors into account, not just SOS.
  2. Players who have an easy SOS in the first few weeks, but have a bad SOS in the last 3 weeks or playoffs, should be considered great players to draft, then trade after their first few weeks before they hit their tough stretch. Ideally, for a higher rated player who starts off with tough match ups (thus temporarily driving his value down), but finishes the season with easy ones. “Buy Low, Sell High.”
  3. Players with a easy fantasy playoff matchups can help you win your championship. After all, the most important wins are the last three of the season. You KNOW you will sweat the match ups each of those weeks, why not plan ahead.
  4. On a weekly basis, check our site for the players with the best weekly matchups. Use this information to help you make lineup decisions between two similarly valued players. Best matchup = highest probability for success. Many times you can identify weekly sleepers based on their opportunity and matchup that week.
  5. When evaluating trades, check a player’s “Remaining Strength of Schedule” to see if you can expect higher or lower than average performances. A good trade would be when you trade a player with a bad remaining schedule for a player with a good remaining schedule.
  6. When BYE weeks come around, use the data to find players with good match ups that week as high value spot starts while your studs sit.
  7. When you are looking at waivers, look for players who have easy opponents for the remainder of the season, this can help you find out if a player was a one week wonder or a waiver gem to help you win your league.

FOLLOW US ON TWITTER FOR RELEASE UPDATES!

How do we generate the data?

During the season:

The data algorithm is based on predicted fantasy score values updated each tuesday am after weekly performances are complete.

Each week most sites rankings are based on Cumulative Points Against for that season. Here at fantasyomatic.com™, we spent the entire offseason developing a brand new way of looking at matchups for the purpose of predicting player performance.

Our new fantasyomatic algorithm™ takes matchups to a new level that you won’t find anywhere else.

What makes it so different:

  • Instead of just looking at the defense the player is facing each week, we look at what other players in his same position that defense has already faced  this season. This tells us whether they are ranked low/high because they faced tough/easy players or if they actually should be ranked low/high
  • We then also look at the defenses that player has already faced this season to determine whether his average fantasy points per week are because of easy/tough matchups and then project what he would score if he faced every defense that all other players at his position faced.
From this analysis each week, we would define the “perfect” matchup at each position as:
A player who performs well against “tough” matchups facing a defense that performs poorly against “easy” players 

The statistical concept we are using here is a complicated one, its called “regression analysis“. Its a perfect concept for predicting fantasy player performance against defense opponents based on their season history.

In statistics, regression analysis includes any techniques for modeling and analyzing several variables, when the focus is on the relationship between a dependent variable and one or more independent variables. More specifically, regression analysis helps one understand how the typical value of the dependent variable changes when any one of the independent variables is varied, while the other independent variables are held fixed.

Most commonly, regression analysis estimates the conditional expectation of the dependent variable given the independent variables — that is, the average value of the dependent variable when the independent variables are held fixed.  In all cases, the estimation target is a function of the independent variables called the regression function. Regression analysis is widely used for prediction and forecasting.

Our formula is :

(intercept) + PLAYER difference (from BASELINE) + Defensive difference (from BASELINE)

We will use this data to determine weekly “Top 100” lists comparing predicted points for every player in all positions (so you can make starting decisions between players at different positions). We will also use this data for weekly positional rankings. Then we will use the data to forecast remaining schedules to help you find excellent waiver pickups  that are off other owner’s radars, and to help you evaluate future player value for trades. This will also serve as your guide to late season and fantasy playoff schedules.

FOLLOW US ON TWITTER FOR RELEASE UPDATES!

We will begin using this data AFTER week three, for the rest of the fantasy season. We will have the data there for you for when you start to look at waivers and BYE week fillers and for your rosters for the rest of the season.

Why wait until Week Three?

One important note about the first three weeks (AJ Mass- ESPN): Since 2002, there have been 54 teams that allowed 2 RBs to reach the 100-yd milestone over the first three weeks of the season. Of these teams, just one finished in the top ten in rushing yards allowed by end of the season-and that team was the ’02 Vikings, who finished 10th, largely due to teams focused on beating them through the air (they had 11 games which the opposing QB threw 350yds+). Meanwhile the rest were holding opposing RBs under 100 yds per game. With each successive game throughout the season, more pretenders fall by the wayside and a definite top tier emerges. Therefore, for the past 7 seasons, teams doing well against a position during the first three weeks, will be the real deal the rest of the way.

Bottom line: According to the last 7 seasons, most of the teams that end up finishing in the top 10 against a position are already in that group after week three.

The same is true for bad defenses.

More statistically speaking, three weeks is what we deem to be  a “mathematically reliable” data sample. Every week after that makes our analysis more accurate until we get the most accurate in time for late season decisions…when it counts th emost.

Thus, the fantasyomatic engine™ calibrates after WEEK THREE. So trust your draft decisions or use our preseason rankings to get you through the first couple weeks. Then let us help you win your championship every week after that.

In 2012 we add even more to our formula! We went back to the lab on the algorithm and came up with some adjustments. We were watching some trends in 2011 and realized room for improvement. Not that the algorithm was flawed in any way, just that we discovered a couple areas where we could make things even more precise. that factor in:

  • Home vs Away performance

  • “Playing Surface” performance ( Grass vs Artificial vs Dome vs FieldTurf)

  • “Spike” game adjustments (players who come out of nowhere to score big points in one game)

The adjustments were purely mathematical and after applying to 2011 data, we realized a 10% increase in accuracy!

Injuries are also factored in should any significant defensive player go down for a period of time.

Before the Season Starts:

Before the draft is A LOT more complicated. Our approach is less mathematical and more based on scoring EVERY defensive player in the league. We concentrate on the defenses and establish “Strength of Schedule”  based on upcoming year data while not factoring in previous year data in any way.

If our preseason ranks have any similarity to the previous year’s rankings, that is due to similar personnel on the team’s roster. Conceivably a team “could” have the same rank as the previous year, if they have the exact same defensive personnel. However, that rarely happens.

Its a very time consuming process, and here is how it works:

1. We wait until the post-free agency rosters are set and then enter every player in the NFL into our database.

2. We enter the defensive “rating” for each defensive player. Depending on their position these rating could be for Pass Coverage, Run Defense, Pass Rush, etc

3. We update and continue to update NFL defensive rosters

4. We enter the NFL schedules for the entire league

5. Then aggregate the totals for all defenses “against” each offensive team/position for every player in the NFL

6. We rank all offensive players based on the strength of their opponents against their position

7. These rosters, ratings and ranks get updated all the up to kickoff of week 1

We use these ranks until we get a “mathematically reliable” dataset of THREE WEEKS of current year defensive performances. Then we update every week for the rest of the season.

Player Ratings from ProfootballFocus.com
Depth Charts from OurLads.com

 See our entire breakdown of all NFL defenses and strength of personnel here

LEGEND:

PR Score:
Total of all Starting D Line players 2011 QB Sacks, Hits and Pressures +Total Starting D Line Pass Rush Rating +Total Starting D Line Pass Coverage Rating +Total Starting Free Safety Pass Rush Score
PR Rank:
1 = Worst Fantasy QB Matchup, 32= Best Fantasy QB Matchup
RD Score:
Total of all Starting D Line Players Run Defense Score +Starting Strong Safety Run Defense Score
RD Rank:
1 = Worst Fantasy RB Matchup, 32= Best Fantasy RB Matchup
PC Score:
Total Starting DB Pass Cover Score + Starting Free Safety Pass Cover Score + Starting Strong Safety Pass Cover Score
PC Rank:
1 = Worst Fantasy WR Matchup, 32= Best Fantasy WR Matchup
TE PC Score:
Total Starting LB Pass Cover Score + Starting Strong Safety Pass Cover Score
TE PC Rank:
1 = Worst Fantasy TE Matchup, 32= Best Fantasy TE Matchup

…any much, much more.

IT DOES NOT PREDICT:
– Weather
– Injuries
– Hold Outs
– Retirement (or unretirement)
– Arrests

There is a TON here, but Im going to break it all down for you from now to the end of the Fantasy Season. So bookmark this page come come back OFTEN. This is the ONLY place you will find this algorithm.

Listen to the explanation here:

Keep one thing in mind, its not “THE” combination that wins you a championship…its simply “A” combination. These are the tools to help you get a championship combination.

– The Fantasyomatic.com Staff (2007)

If you want to join the FantasyoMatic email posting update list, register here :