Fantasy Baseball projections. There are a million of them out there, all of which place importance in different areas. So whose do you trust?

Honestly, I don’t know. Baseball is a game played by human beings, which means unexplained variation is going to happen. However, there is no reason to avoid the past in an effort to forecast the future. This set of projections is numerically driven and free of opinion. That’s right; these are rankings/projections that have nothing to do with what I think.

Let me explain. Instead of attempting to project every player under the sun, I focused my research on players that were relevant last season. Makes sense, right? My player field consists of the Top 100 hitters on ESPN’s Player Rater in 2014, a group of players that will be drafted in most, if not all, leagues. Once I had a reasonable player pool, my goal was to assign a single number to every player for the upcoming season, thus allowing no room for confusion: the player with the higher number owns the better Fantasy projection for 2015. But how do we get that tell-all number?

First, I charted the five most common hitting stats (batting average, runs, home runs, RBI, and stolen bases) for each player and calculated what the average Top 100 player produced in each category. This chart is sorted according to Player Rater finish in 2014.

My goal was to normalize all counting statistics in an effort to weigh all equally, thus giving me the ability to label every player with one singular number for ranking purposes. So I used a “common multiplier” to normalize my data, according to batting average: that is, the number it would take to get the other four categories (runs, home runs, RBIs, and stolen bases) equal to batting average (as a whole number, not a decimal) for the average batter. This common multiplier was determined by taking the average Top 100 hitter stat line (.283 batting average, 77.5 runs, 18.4 home runs, 73.6 RBIs, and 13.1 stolen bases) and simply dividing the batting average by the stat of choice.

Example: The average batting average was actually 282.88 and the average run total was 77.45. I simply divided batting average by run total to give me my multiplier: 3.65. For the study, I will multiply a player’s projected run total by 3.65 so that it is weighted equally as batting average, thus allowing for all statistics to be tracked on the same scale.

Original Batting Average Multiplier: 1.00

Original Run Multiplier: 3.65

Original Home Run Multiplier: 15.38

Original RBI Multiplier: 3.84

Original Stolen Base Multiplier: 21.68

BUT … all stats are not created equal. Some of those totals were ballooned by outlier performances. In order to adjust, I modified the common multiplier to account for such abnormalities. I did this by charting how many players on the 2014 Player Rater scored above the mean for each statistic. Using 50 as a benchmark, I then adjusted the multiplier based on the percentage of players that finished above average.

Example: There were 45 players that finished the 2014 season with at least a 78 runs scored (remember, 77.5 was the average for the Top 100 players). That represents 10 percent fewer batters than we would have had in a perfect world (ideally, when taking the average of something, half the data falls above and half below), so I subtracted 10 percent of the multiplier. Thus, the runs multiplier, which was initially 3.65, is now 3.29.

Updated Batting Average Multiplier: 1.02

Updated Runs Multiplier: 3.29

Updated Home Runs Multiplier: 15.38

Updated RBI Multiplier: 4.07

Updated Stolen Base Multiplier: 17.35

This is what I’m talking about. Now you’re ready to calculate the “Soppe Score” (SS) that is listed as the final column in the table above. Simple multiply each statistic by the updated multiplier and add them all together. You’ll notice that the 2014 ranks are similar, but far from identical.

[table “1339” not found /]But enough dealing with statistics that have already been accumulated, those won’t help you win your Fantasy Baseball league this season. The projected numbers used for this study do not require a college degree to understand. I used the past three seasons of statistics and calculated a SS per at-bat. I then evaluated the player’s career and used his past to calculate a reasonable projection for an at-bat total. Multiply the per at-bat projection by his projected at-bat total and you’ve got yourself a logical Soppe Score.

FURTHERMORE … not all sets of stats are created equal. Numerically speaking, projections are most accurate when there is a large sample size, as it allows us to form a better opinion of who the player is and what he can be counted on to produce annually. Well, not all high-ranking players from last year (Jose Abreu, Dee Gordon, and Charlie Blackmon to name a few) have a track record of proven excellence, but at the same time we cannot completely dismiss their stellar 2014 stat lines. For this reason, I have two sets of data for each player that has been a regular for fewer than three seasons. One (italicized in orange) is the player’s production as a regular (for some players, this is simply their 2014 stat line, while others have two seasons under their belt) and where that production alone would land them in my projections for 2015. The reason I included this data is simple: if you believe what they did last season is the new norm, you know where to rank them. The other set of data (red) is how their “Steamer” projection (as found on FanGraphs) would grade out in my formula … with one exception. While the great minds behind the Steamer projections work as hard as anyone, even they can’t truly tell what a young player is destined to do, as there simply aren’t enough numbers to crunch. How are we supposed to know if that player will develop into an injury prone player or if 2014 is destined to be the norm/outlier? The truth is that there is no certain way, not this early in a career anyway. To account for this uncertainty, I’ve included a minor penalty for these players, a modification that admits that the sample size simply isn’t large enough to properly project performance. The thought behind such a deduction is not to try to forecast the exact regression that will occur, but rather to admit that there is inherit risk involved in taking a player with a limited professional resume.

I viewed the Top 10 scoring players in this formula and charted the percent change in their Soppe Score from year one to year two. The average player saw his SS change by 17.8 percent. I also looked at their average SS through two seasons and compared it to their year three score. The average player saw his SS change by 20.5 percent. These are the elite players, and even they were experiencing varying levels of success with limited amounts of consistency. This trend would lead me to believe that it is fair to assume that the possibility of volatile production for the current crop of young Fantasy stars is present. To account for this, I adjusted the Steamer generated SS by those aforementioned percentages (a 17.8 percent drop for players who are set to enter their second season as a regular and a 20.5 percent drop for those entering year three). Age bias? Maybe, but I place high value on track record and would rather draft a player that I know has had success at this level for an extended period of time than take a risk on a player who lacks a resume of Fantasy goodness (no matter how good the player is, there is risk involved in banking on a player who simply hasn’t proven the ability to sustain greatness). Here are the ranks, sorted by SS with the experience deduction (players who incurred a penalty are starred) included.

^^^ If you’re skimming through this article for the pure rankings, this is your table ^^^

But let’s say you’re happy with your own set of projections/rankings; that’s not an issue. No matter where you are in your preparation process, one major goal will be to identify strong value options and I can help you there too. Here is a look at what my formula thinks of experienced players (at least three seasons of being a regular) and their production compared to their 2014 stat line. I’ve also included the change in from their 2014 SS ranking to their projected 2015 SS ranking.