Every year for the last several years, the annual task of creating player projections for the Xclusive Edge Premium Baseball Package and the upcoming Fantasy Baseball season has fallen to me. It is a daunting, tedious and difficult job that takes considerable time to complete. Yet, it is also a labor of love. I like to think of it as my opportunity to get the first in-depth look at the season ahead. Nobody can foretell the future, but putting together the projections is a pretty darn good attempt at it. For all the talk about players having streaks of good (or bad) luck, along with all the other things that can affect a player’s performance on the field, their overall statistics are actually fairly predictable.
The difficult part of putting together projections is actually all the things that happen after they’re done. Spring training injuries, depth chart changes and breakout performances during the spring can all affect playing time. Then there’s the free agent signings and trades that can totally change a player’s future production. So, keeping up with all of that and adjusting the numbers accordingly is much more difficult than coming up with the projections in the first place. One of the more frequent questions we’re asked at RotoExperts.com is about what goes into our projections. So I’m going to give you a little peek behind the curtain and describe the process.
Xclusive Edge subscribers can see our projections in a few different places depending on how you like to use them and how you prefer them displayed. Sortable projections are available here. You can manipulate them right on the site and sort them by different statistics. Our player database is also sortable in a number of ways and the projections are listed along with each player’s three-year statistics and other pertinent information. Finally, our recently revamped Cheat Sheet Generator offers you a variety of ways to sort and display our projections, including custom sorts for points leagues or sorted according to our 5X5 rankings. Plus, you can download them to Excel from here to manipulate them however you choose.
If you’d like to see and use our projections and you’re not an Xclusive Edge subscriber, it’s never too late to get an edge over your competition. Don’t forget, subscribing to the Xclusive Edge Fantasy Baseball package includes a full season of access to our premium content right up to the end of the regular baseball season. All season long you’ll get the very best advice, the very latest information and our award winning analysis to help you win your Fantasy leagues regardless of how you like to play Fantasy Baseball. Subscription information for the Xclusive Edge Fantasy Baseball package can be found here.
PROJECTIONS THE ROTOEXPERTS WAY
The projections process actually begins as soon as the last out of the regular season is recorded. So, while you’re sitting in your easy chair and watching the MLB playoffs and World Series, work on the projections for the next baseball season is already underway. We begin the process by putting together a statistical profile of every hitter and pitcher that participated in MLB play during the previous season. We take the past three years of player statistics to calculate each player’s average annual production; in other words, their mean. While that calculation serves as the basis of what is to come, we don’t ignore what’s happened prior to the last three years. One premise that many baseball analysts have come to embrace is that once a player displays a particular skill, whether it be power hitting, speedy base running, the ability to hit for a high average, superb control or limiting groundballs, that skill could reappear at any time during the player’s career. So, the entirety of a player’s statistical output is considered in addition to their three-year averages. Injuries or just plain bad luck may have caused a particular skill to temporarily diminish or disappear, but a return to good health or a change of luck (or scenery due to a trade) could easily allow the skill to return.
Once we have three-year averages and scanned their history to note their skill metrics, we can begin to look towards the future. At this point, factors like the ballpark they’ll play their home games in, the division where another sizable chunk of games are played, as well as the lineup around them all come into play. In fact, we project the run environment that they will be working in by projecting how many runs their team will score during the season. That number, divided by 162 will give us the number of runs scored per game by the team, which we then use to figure the number of runs and RBIs each individual player will contribute, first on a per game basis and then over the course of the whole season.
Next, we figure a player’s age into the equation. In general, a hitter will increase in production from ages 24 through 29 or 30, with their peak production years falling somewhere in the middle around age 27. After age 30, the majority of hitters begin to decline in production. Pitchers are similar in that they will typically become increasingly productive through their 20s, and then begin their decline phase once they get to the other side of 30 years old. Of course, every player is different and some players defy the typical aging curve of the average baseball player. However, we compare the age-adjusted production of the average pitcher and hitter to every player we project to get an idea of where they fall on the aging curve. We use this comparison to predict whether they are still on the upswing in production or perhaps entering their decline phase. At some point, every player starts to regress towards their average production and towards the age adjusted average production of the average ballplayer. Predicting regression is where player projections become a bit more like art and a lot less like science.
Once we’ve calculated what looks like a reasonable set of projected statistics for a player, the next and most difficult factor to project is playing time, the number of plate appearances a batter will accumulate or the number of innings a pitcher will pitch over the course of the entire regular season. This is when projecting ballplayers really does become more like art than anything else. We can make some assumptions based on what has happened over the course of general baseball history. For instance, a full-time hitter will accumulate somewhere around 600 plate appearances, more or less depending on where they bat in the lineup. A hitter that bats in the top third of the lineup will typically make 650-700 plate appearances, while batters in the lower third will make 550-600 PA. Keep in mind that these are all very rough numbers, subject to any number of variables that will change from player to player and team to team.
Injuries, platoons, days off from the grind given by the manager, and performance can all raise or lower the number of plate appearances a player will make. So, we look at a player’s injury history to see if there are patterns there. Some players will consistently make at least one trip to the disabled list every season. That can be figured into a playing time projection. We also look at the roster to see if the team’s management might be considering a platoon. The 600 plate appearances for your average player might be divided among two or more players if they are sharing a position. Some fourth outfielders or utility infielders might only make 200-250 plate appearances. Catchers can be especially difficult to project, as some play nearly full time while others may platoon, with the plate appearances divided up any number of different ways.
For pitchers, projecting innings pitched is a little bit easier as long as you have an idea of what their role will be and where they’ll be slotted in the rotation. Starting pitchers will make anywhere between 25 to 32 starts in a full season, more if they are the team’s “ace” and less if they are fourth or fifth in the rotation. Middle relievers are a bit tougher to project because their role can change many times during the season. So-called “swing men,” who may start occasionally and work in long relief, often pitch between two and five innings per outing and can pitch anywhere from 60-100 innings. Most closers work anywhere from 50-70 innings, while left-handed one out guys (LOOGYs) may only work 25-40 innings. Finally, setup men, who typically get most of the high-leverage innings that don’t go to the closer, will typically work 70-90 innings. Of course, this can vary greatly depending how the team’s manager doles out his assignments. However, studying a pitcher’s statistical history usually reveals the way he’s been used out of the bullpen and since we’re talking about relatively small samples anyway, a projection for a relief pitcher that is off by five or even ten innings isn’t going to make all that much of a difference in Fantasy.
To sum up the playing time conundrum in the projection process, it really is part science and math with a little art and creative thinking mixed in. To make a fair projection, one has to try very hard to leave any sort of bias out of the equation. Just because a “name” pitcher has a reputation for going deep into games and racking up strikeouts it doesn’t mean he will do it again. One cannot do a reasonable projection if you ignore the fact that the numbers say his performance is in a four-year long decline phase. If you are biased by the reputation, ignore the pattern in the numbers and make a projection that doesn’t match up to that pattern, you are not doing a very good job. However, if the player has been unhealthy for most of those four years and he’s suddenly experienced a recovery that allows him to play better than he has in recent years, you have to account for that somehow. That’s where the “art” of projecting performance comes into play.
Ultimately, though, the vast majority of projection work comes down to mathematics. It would take months on end to do all that work manually (even with a great calculator), so the work is done with spreadsheets. At this point, I’d like to give a shout out to Tanner Bell, who has a wonderful website he calls SmartFantasyBaseball.com. While we at RotoExperts are in the business of doling out our take on how to succeed at Fantasy Baseball, Tanner focuses on teaching his readers how to think for themselves by providing the tools and resources to create their own rankings, projections and more using Excel spreadsheets. In fact, Tanner’s projection spreadsheet, which he makes readily available for a nominal fee, formed the backbone of the projection system that I used this year. It worked so much better than some of the projection spreadsheets I’ve created in years past. I was able to easily modify it to meet my needs and Tanner was very accommodating with support and answers to my questions via email. In addition, Tanner offers other downloadable tools and has a resource page that lists many sources of information and information rich books, including at least one e-book that he authored. If you’ve ever had any interest in creating your own projections or rankings, or if you just want to learn a little bit about how the process works, I urge you to check out SmartFantasyBaseball.com.
One last item I’d like to address is the question of accuracy. Just how accurate are RotoExperts’ projections? Of course, we won’t really know until the season is over and we can compare the players’ actual statistics to our projections. However, I can tell you from years past that, for the most part, we’ve done a job that I am very proud of. No, we don’t nail it for every player. In fact, I’d say we come reasonably close two-thirds to three-quarters of the time – roughly 70 percent accurate, if you force me to give you a number. If you were to poll the various people and organizations that create projections every year, nearly all of them would be pleased with a 70 percent accuracy rate. Every baseball season is filled with instances and events that will ruin a projection – injuries, hot streaks, a trade, a new pitch or newly awarded playing time are just a tiny sample of the many things that can alter the numbers and render a projection moot.
All that aside, I believe that this year’s projections, which encompass many more factors with much more accurate information than any previous year that I’ve done them, will be the most accurate projections ever offered on RotoExperts. The refinements I was able to add, thanks in large part to Tanner Bell, made it possible to create projections that took into account many factors that I was unable to incorporate before due to my limited skills with spreadsheets. So, while I am looking forward to this Fantasy Baseball season because I love baseball and can’t wait to draft my teams and get things underway – I am also incredibly curious to see just how accurate the revamped RotoExperts’ projections turn out to be. You can bet that the day after the season ends, as I start downloading player statistics to begin the process anew for 2016, I will also be preparing to compare the 2015 actual statistics to our projections. My goal is 80 percent accuracy; a very lofty goal indeed, but I am confident we can approach that level of accuracy. You can bet I’ll be letting you know.
Got questions about our projections? The Xclusive Edge package? Anything else on RotoExperts.com?
The send me an email at firstname.lastname@example.org, or hit me up on Twitter @Tim_RotoExperts and I guarantee you’ll hear back from me.