The Weekly Gut Check - Vol. 95
Projecting Player Performance
7/6/07

The “Gut Feeling” is often synonymous with a sense of desperation resulting from a lack of preparation. The Gut Check is a huge proponent of studying the numbers, but there’s a point where one can place too much emphasis on the wrong information. This can result in the undervaluing or overlooking a player’s potential. Therefore, The Weekly Gut Check is devoted to examining the frame of reference behind certain number-driven guidelines that fantasy football owners use to make decisions.

Although The Weekly Gut Check doesn’t claim to be psychic, he does believe that he can dispel certain numbers biases and help you make the best choices for your team. We’ll keep a running tally of The Weekly Gut Check’s insights. This way you can gauge his views as something to seriously consider, or at least seriously consider running the opposite way as fast as you can!

But how does one do projections? There are several ways. This article will serve as an introduction to some basic techniques. The next four articles will cover the statistical high points of the offensive positions since the advent of the 16-game season in 1978. This data will help the reader determine whether his projections (whether his creation or from another source) have a historical basis in reality.

Step One—Determine The Amount of Players To Project

One should figure out how many players he or she wants to project. Despite the fact this is a hobby, projections require work. The Gut Check recommends you work efficiently. In fact, he suggests you forget the idea of projecting performance, save yourself potentially hours of work, and buy the Cheatsheet Compiler and Draft Buddy. This application will provide you carefully weighed projections and all you have to do is enter your league rules. Plus you can view a variety of projections or import them from other sources. Even if you insist upon doing all your projections, the Compiler is a great format to maintain an organized list that you can transfer to the Draft Buddy when you’re ready to select your team.

If you’re absolutely against the idea of spending money on this product—despite the fact that even if you valued your time at minimum wage, 99% of you would spend a lot more money doing this work yourself than simply buying Mike MacGregor’s terrific product. Plus you support one of the last remaining free fantasy football information sites on the Internet. Other options include using Crank Scores or learning more about the Average Value Theory.

But if you are dead set on doing projections on your own without the help of these resources, yours truly is here to help (though he’d also recommend Projection Pal as a way to organize your results and make it easy to update as player values change throughout the preseason). Once again you should factor how much time you have to project performance. Your resources should be based on your available time. Yours truly would budget at least 2-4 hours per position (kickers and defenses possibly excluded) depending on whether you only intend to project starters or all players on the depth chart of a team. After these initial projections, one may need to devote an hour a week to update the results when players get hurt or their standing with the team changes. This depends on how far from your draft date you decide to project performance. Below, the Gut Check will recommend the resources that will best suit fantasy owners according to their time allotment and the amount of players I decide to project.

Step Two—Resources

Limited Preparation Time (4 hours or less): If you just don’t have enough time, get the Compiler and use the projections there as your base set. Make adjustments where necessary or import projections to the application from your on-line source of choice. Another option is to copy and paste online projections to a spreadsheet and you can make any desired adjustments. The Gut Check strongly recommends you take projections that include statistics for attempts, yardage, and touchdowns along with projected fantasy points. Otherwise, you cannot make logical adjustments. If you’re really hard up and lack the cash to buy the Compiler, use these projections or Mike Krueger’s rankings format, which is what the Gut Check would do if he were in such a situation. If your normal mode of operation is to buy a magazine and use those outdated rankings—which if you’re reading this article, yours truly finds very doubtful, you might as well spend the extra ten bucks and get the Compiler and Draft Buddy otherwise you simply just aren’t thinking straight.

Minimal Preparation Time (4-16 hours): Get the Compiler to make the most of your time and download several projections that you like. How do you know which projections are good ones? Check out the remainder of this series over the next four weeks—if you have that much time prior to your draft—and compare the information with the projections you review. This will help you spot the well-conceived figures from the rest. Even with the lists you like, begin to take note of the players with numbers that you disagree with and those will be the players of focus. With limited time, it doesn’t make sense to try to project any more than 10%-25% of the total players on your list. This is because a good draft list of players will have at least 64 QBs, 96 RBs, 128 WRs, and 64 TEs—when one includes the kickers and team defenses one will have 416 players due to the necessity of keeping depth chart players in mind. At least all the starting QBs, two-thirds of the RBs, three-quarter of the WRs, and two-thirds of the TEs should have projections. Considering it will take the average person 10 minutes to do a good job projecting a player’s performance—some players will take longer than others—one is looking at a minimum of 40 hours of work to project every player the Gut Check would reasonably recommend. If you want to do a decent job doing your own projections but only have 4-16 hours, then forecasting 10%-25% of the players you should have on your list will fall within your time requirements. As an alternative, the Gut Check would strongly recommend you consider using the Average Value Theory or the Gut Check’s Crank Scores as a basis for compiling a list without projecting performance. Of course, yours truly recommends you incorporate Crank Scores even if you do projections—rating player consistency is a good way to tier players with similar values.

Moderate Preparation Time (20-40 hours): You should be able to forecast 50%-100% of a solid list of players with this amount of time. Obviously, the Compiler becomes a great tool for inputting your own projections and allowing the macros to do the work of tallying the fantasy points for your league scoring system. Yours truly knows this may seem like shameless promotion of an FF Today product, but it’s the best product of its kind on the market—and the Gut Check has tried several products that followed in its footsteps (Mike’s was the first of its kind). So whether it’s the Compiler or your own spreadsheet where you input your initial projections, the next step is to take a look at last year’s stats—use FFToday’s stats so you can look at both the first and second half of previous seasons—to see if a few theories apply to some of your players. If so, these concepts may impact your projections. Here are some concepts one should consider for step three of this article:

• Second Half Wonders
  
• Profiling Breakout Receiving Candidates
  
• RB Workloads—The concept of f-carries will help you spot backs that might experience a dip in production. Look for a 2007 version this summer, but this 2006 article will still help.

Once you’ve projected performance, acquire some average draft position (ADP) data to note whether your projections may be over valuing or undervaluing the player in comparison to your peers. When the Gut Check says this he doesn’t mean you are wrong, only that your views differ from the norm. Having awareness of your differences from the rest of the crowd will help you make wiser selections on draft day.

Emulating The Reclusive Lifestyle of Howard Hughes: This must mean you intend to forecast the performance of every player on an extensive list and will be committing the time to create a unique list you will feel strongly about your data. The Compiler, ADP data, the articles mentioned above, and Crank Scores to create separations among players one may have within a similar tier of fantasy points. In essence, you’ll do everything listed under Moderate Prep Time, but more extensively.

Step Three—Begin With A Simple, Clear Cut Method As A First Draft

Once you have your list of players to forecast it is time to make sure you’re using the right kind of stats to create a solid performance projection. Everything should be cause-and-effect in nature when projecting numbers. These are the key stats one should project for each position:

QBs—Attempts; completion percentage; passing yards; passing tds; rushing attempts; rushing yardage; and rushing tds.
RBs—Attempts; yards per carry; rushing yards; rushing tds; receiving targets; receptions; yards per catch; receiving yards; and receiving tds.
WRs and TEs—Receiving targets; receptions; yards per catch; receiving yards; and receiving tds.

Stats such as yards per carry, completion percentage, and yards per catch will help keep your expectations in check when coupling them with attempts or receptions. Another stat the Gut Check didn’t mention for quarterbacks, but is also very helpful, is viewing the yards per completion for each quarterback. This will help you see the differences in the passing game of the offense each quarterback uses. An owner will be able to see clear differences between west coast offenses and vertical passing attacks—important differences to create realistic projections.

An owner should begin with attempts (for RBs and QBs) and targets (WRs and TEs). He should take a look at 1-2 years of stats for the player in question. Let’s use Edgerrin James as an example player to forecast performance. We’ll start with two years of stats.

One of the Gut Check’s favorite runners of this era, James missed the 370-mark for f-carries in 2006 and will be playing behind a line that might have an upgrade with first round pick, Levi Brown. He’ll also have Matt Leinart with a complete year under his belt. The Gut Check believes James should improve upon his stats in what will likely be the swan song of his most productive years as a runner as he nears his athletic prime at 29 years of age.

One could be extremely methodical and calculate the average difference in yards per carry for a starting back between the year he had a rookie QB under center and that QB’s second season, but the Gut Check isn’t going to do this just for an example. Let’s just say James will improve his yards per carry from 3.44 in 2006 to 3.9 in 2007, a bit closer to his 2005 average and a good half-yard lower than his second half average with Arizona. To yours truly, this is a conservative estimate considering how James looked from weeks 10-17. Conversely, he’ll lower the yards per catch average to 5 yards because as James’ ground production improved, his yards per catch dipped closer to the average of his Indy days.

Now let’s focus on attempts, targets, and receptions. First, one should ask whether there is any reason—barring injury—why James won’t receive similar attempts this year. Yours truly doesn’t think so. If anything, coach Whisenhunt will likely place more emphasis on the run. This should mean James gets at least as many carries. The difference will be the increased efficiency of the running game. Forecasting James at 340 carries seems reasonable. At 3.9 yards per attempt, James would then gain 1326 yards—still starter-worthy ground stats. It would seem James has fewer receptions in this offense, but not significantly less. During the second half of 2006, James averaged nearly half the targets and receptions as the first. Let’s say James will perform closer to the second half of his 2006 stats as a receiver. This means he’ll average 2 receptions per game for a total of 32 and 5 yards per catch his receiving yardage will be projected at 160 yards—nearly the lowest of his career.

But this doesn’t make sense for one of the better receiving backs in football. Additionally, Matt Leinart will likely be smarter about checking down to his back with another year under his belt. Look for James to have the number of targets and receptions closer to his 2006 amount when Kurt Warner was under center. Look for 3 catches per game for a total of 48 and forecast a slight increase in yards per catch at 6.25. This will put James at 300 yards receiving.

Touchdowns are trickier to forecast. One method is to calculate touchdowns per game. For James it would mean .375 scores per game on the ground and .0625 scores per game through the air. But the Gut Check believes the rate of touchdowns should be predicated off attempts and not games. Therefore, James ground scoring rate last year was 1 score every 56 attempts and through the air, 1 score every 38 catches. In 2005, James had a much higher rate on the ground with one score every 27 carries. The Gut Check once again looks to James’ rate of scores in the second half of 2006, when the Cardinals running game seemed to improve. The rate of scores per carries was 1 for every 50—not much better. Still, touchdowns come from defenses that are off balance due to efficient offensive play. Leinart’s improvement should increase the overall offensive efficiency, but how does one link Leinart’s ascent to James’ scoring production?

One method is to look at Whisenhunt’s last team’s--the Steelers--ratio of rushing and passing scores between the starting QB and starting RB in 2006. Roethlisberger had 18 scores to Parker’s 13 in 2006 and Big Ben’s 17 to Bettis’ 9 in 2005. The 2006 ratio is 1.38 QB scores for every RB score. For the sake of simplicity, let’s use this ratio and apply it to Mike Krueger’s current projections for Matt Leinart’s 18 passing tds—which will give James 13 scores. The same could be done for James’ receiving scores, which would give him 3 receiving tds.

For those of you keeping score, the Gut Check’s preliminary projections for James would be:

Yours truly has a bit more optimistic forecast for James as a scorer. Will the Gut Check lower his expectations for James in this area? Let’s go to the final step.

Step Four—Refining Projections

This step is where you need to consider factors such as injuries, players on the depth chart competing for time (highly regarded rookies), and recent historical performance of the player in question and his current offensive system. In the case of Edgerrin James, his surrounding personnel remain nearly the same with the exception of Levi Brown and a new coach. There are several cases of rookie linemen jump-starting the ground game with their performance, but it is no guarantee that Brown will follow suit. But the addition of a coach who has success with emphasizing the run and played in offenses steeped in a ground attack—Joe Gibbs Redskins of the 80’s and Dan Henning’s run-oriented offense with the Falcons—the Cardinals offense should be more efficient. James should have a stranglehold on the starting job. J.J. Arrington enters his third season and has done nothing to indicate he will earn more playing time. Marcel Shipp is a consummate backup.

So in this case, the Gut Check would only look to adjust James touchdowns. Although the Steelers ratio could be applied to the Cardinals, Pittsburgh’s defense keeps the offense in games and provides them time to run the football. The Cardinals defense is a group with talent but they aren’t as effective—they allowed nearly 30 yards more per game on the ground and nearly 20 more yards in the air than the Steelers. As a result Arizona allowed opposing offenses more time of possession than Pittsburgh’s and this was one cause for the Cardinals offense to throw more with limited time. In addition, Pittsburgh lacks to great redzone threats on the outside like the dynamic duo of Fitzgerald and Boldin. This in itself should detract from the Gut Check’s forecast of James’ touchdown potential. Knocking 4 scores off Edge’s totals (3 rushing and 1 receiving) seems pretty reasonable.

As a result, James would likely have somewhere between 234-258 fantasy points in a traditional scoring league (non-ppr), depending whether you refine Edge’s projections as mentioned above. This would place James somewhere between 4th and 8th overall at his position in a set of projections for 2007. The Gut Check doesn’t find this too surprising, because James was the 3rd-most consistent back in fantasy football in 2005 and Willie Parker, Whisenhunt’s starter in Pittsburgh, was the 4th-most consistent back in 2006. James returning to a higher level of consistency would be a good bet. Still James’ ADP will likely make him a bit of a draft day value since it is likely his projected stats are higher than the public perception of many that don’t do forecasts. The implications could mean you can snag him as a possibly a #2 RB for your fantasy squad. If you trust the methodology of your projections, this will be a bargain. And that’s why good projections can make or break your fantasy season.