FB402: Beyond The Numbers

05/20/2007 10:42 PM - 

FantasyBaseball.com University Series
Contributed By: Scott Delp

Baseball has always been a game which has been analyzed by the stats that it keeps. It is no coincidence that fantasy sports, based so devoutly on stats, began with baseball. As the involvement of fans in fantasy baseball has grown, more stats have been invented where none were thought to be possible. Now we have VoRP, Runs Created, OPS and our own Ray Flowers has developed SWIP, which you can explore in some of his work on the fantasybaseball.com site. All of these are designed to help you better evaluate players.

Two important points might be made at this time. The first is that baseball is unique to all major sports in the world in that it occurs outdoors in venues of non-uniform dimensions. The second, and most important for fantasy purposes, is that although previous performance is a useful indicator, the key is finding and/or making good projections of what players will do in the future. It’s a lot like handicapping a thoroughbred race in that the past performance of each of the horses has been impacted by so many different factors.

All of this makes it imperative that successful fantasy baseball owners look beyond the stats and try to find the context in which those stats were accumulated. Following that, it is important to look for changes in context that might make the upcoming season significantly different from the previous ones. Most of the stat people who develop projections, will talk to you all day about ‘regression to the mean’, which simply means that they expect a player to basically perform this season in much the same way as he has in the past. The more of a career the player has had, the more confident the statistician will be in his projection.

You can start with any set of projections with which you are comfortable, but it is important to then take those stats in context and look for players who should, because of changes in context, perform significantly above or below ‘normal’.

What is the context you should be looking for? There are a number of factors that impact a player’s stats:

Venue - In most cases this is an easy one, but there are also subtle factors to consider when looking at those players who have changed teams. You can find park factor stats for the previous season or for a longer period of time. Sometimes the usual assumptions do not apply. While Coors Field was the park that increased runs the most of any park in the majors, Wrigley Field was actually the park that increased HR the most. Two other parks, in addition to Wrigley (US Cellular Field and Bank One Ballpark), increased HR by more than Coors while Citizens Bank Park finished tied for fourth with Coors in increasing HR. You can find park factor stats elsewhere in this guide. The park factor stats are very interesting and can add a lot to your consideration of context in adjusting your projections.

Role - Again, this can be easy. Any player either gaining or losing a full time job is going have his stats impacted, but the subtle factors are what make your projections more accurate. Does he have the make-up for the new role? Are his peripheral stats such that he is a good bet for success in an increased role? Another, sometimes overlooked, factor of role is position in the batting order. Knowing where a player is going to hit and who is likely to hit around him can be key in making accurate projections. Finally, does the player have legitimate options behind him on the roster? Teams have to be more patient with players who are not backed up with solid options and obviously will be less patient when they can turn to someone they feel will get the job done.

Injury - Did the player play through some injury last season? Is he nursing a new or old injury as we come into the new season? Is he coming off surgery? What do you know about his recovery? Any one of these answers that you don’t know may make the player a risk not worth taking, especially with an early pick or a huge auction salary. Injuries are a tricky part of fantasy baseball. Last season, by the time most leagues had their auctions/drafts, players like Curt Schilling, Barry Bonds, Oliver Perez, Kerry Wood and Mark Prior were having injury troubles which were (at least at first in most cases) downplayed. None made much of a contribution on the season and most of those players were either high picks or cost a good percentage of your auction budget. There are just too many players out there who are healthy and able to put up numbers to speculate on guys who are hurt. The only way to justify getting a guy like that is if everyone in your leagues reads this and the player lasts late enough in your draft to make him worth the chance you take.

Context stats - These should definitely be a part of your consideration of any player with no or limited major league experience, but they can also be very helpful in projecting players who fit into some of the other categories mentioned. K/BB ratios for both pitchers and hitters and ground ball to fly ball ratios for both pitchers and hitters are two very useful stats for making accurate projections.

Intangibles - There are a lot of theories about the age at which players break out and the effect of the contract year on stats. Most of these have been proven to be statistically invalid, but enough players still fit the theory that they won’t go away. Make-up of a player was mentioned earlier, but fits here as well. Some guys can’t handle the big stage while some only perform at their best when they are on the big stage. For pitchers, the effect of the team’s defense can’t be overlooked. Derek Lowe last season was a prime example of this. It was easy to predict that his overall numbers would improve in the move from Fenway to Dodger Stadium, but a deeper look would have told you that a ground ball pitcher such as Lowe needed good infield defense behind him. The Dodgers were trying to patch a defense together from the start of the season and Lowe seemed to let every error get in his head. He pitched a lot better than his numbers indicated last season, but must find a better defense behind him and do a better job of shrugging off the errors for you to get full benefit of his pitching. Any move to a new team and the resulting change in chemistry also can have an effect on a player’s performance.

In order to be truly prepared for your draft, as many of these factors as you can reasonably quantify must be considered. It is the owner who goes this extra step that often prevails in fantasy baseball. Everyone is working from the same stats and lists to start with. Finding context and making reasonable adjustments is the difference that separates winners from also-rans.

Let’s look at some players who fit some of these categories this season and judge accordingly.

Kevin Millwood - Millwood has been somewhat inconsistent throughout his career anyway, but he has moved among a few different venues and we can gather some thoughts from the results. He began his career and had his best success (before last season) in Atlanta. Turner Field is a pretty fair park, though it does slightly favor hitters. He moved to Veteran’s Stadium and Citizens Bank Park with the Phillies and his stats took a predictable turn for the worse, especially with the move to Citizens Bank Park. In 2005, he moved to the Indians where Jacobs Field is among the better parks in the majors for pitchers. He actually gave up a few more HR/9 in Cleveland, but overall the park helped him to an excellent season and the ERA title in the AL. His signing with Texas this winter is a huge red flag. Ameriquest Field is among the best hitter’s park in the majors and Millwood’s stats should take a fairly healthy turn for the worse. Not only should you not expect another sub-3 ERA, it might not be realistic to expect anything under 4.

Scot Shields - Last season, in putting together an article similar to this one, one prediction was that Guillermo Mota was not a risk worth taking. That was based on the context that, historically, middle relief pitchers who had consecutive seasons of around 100 IP were serious injury risks. Mota’s stats from last season, for a pitcher who was an unquestionable closer going into the season, were appropriately dismal and he spent some time on the DL. Shields is the guy this season who fits that profile. In the second half last season, Shields’ K/9 decreased 30% from the first half, while his H/9 increased 17%. He pitched 105 innings in 2004 and 91 last season. In 2003, he was a spot starter and pitched 148 innings. He is an excellent candidate to be much less effective this season and/or to spend time on the DL. Let the rest of your league pay elite middle reliever money for him.

Adam LaRoche - At first glance, it may look as if LaRoche took a step back last season as he hit for a much lower average in 2005 than he had in 2004. The context here can go either way and it is important for you to do some research as the season approaches. LaRoche battled a shoulder injury for much of last season, but still hit for decent power and had 78 RBI in just 451 AB. If his shoulder is sound, he’s a good bet to have a significantly better season in 2006. If the shoulder is still nagging, he’s a guy to avoid.

These are just a few major leaguers who you can find some context on and possibly get some value or avoid a costly mistake. As you work with these ideas, you’ll find more and more of your own.

Rookies are a whole different breed and choosing the right ones is all about context. Picking solid rookies, whether for the future or to have them contribute to your team more quickly, depends in large measure on venue, peripheral stats (especially K/BB ratio) and age relative to the other players at a given level.

Let’s look at four pitchers whose names are being tossed around this winter. Matt Cain is supposedly a potential stud. Yusmeiro Petit and Anthony Reyes are expected to end up in the Marlins’ and Cardinals’ rotations respectively and Chuck James seems to be major league ready for the Braves, though he currently does not have a spot. Cain and Petit are both just 21 while Reyes and James are 24. Hype aside, Cain and Petit seem to be the lesser of the four pitchers at this point. Cain has struggled with walks throughout his career and walked more than 4.5/9 last season in the PCL. It would be easier to write that off as an aberration of the hitter-friendly PCL except that he had similar BB/9 numbers in the more pitcher-friendly Eastern League in 2004. He may turn out to be a fine pitcher as his career continues, but this is not the season to spend a lot more than a minimum flier on him in your drafts. Petit is similar in that he’s been hyped for the past couple of seasons. He is a pitcher with big K numbers though he does not throw especially hard and he has gotten by on a deceptive motion. The rebuilding Marlins are likely to use him in the rotation this season, but his numbers from last season might make you want to stay away from him. His HR/9 took a jump and his K/9 took a slide as he moved to AA last season, again in the Eastern League. Even more alarming were the numbers from his brief stint in AAA at the end of the season. It could be that he was simply not ready for the jump (meaning he’s also not likely ready for the majors this season) or it could be that more experienced hitters will be less fooled by his delivery. Either way, he’s not a guy to target any time before the end of the draft. Reyes and James have had consistent, sustained success at every level of their progressions, including outstanding K/BB ratios. Reyes even maintained his success in the PCL last season and seems clearly a better bet than Cain this season. James pitched most of last season at AA and likely needs some more time in AAA, but he would still be a better pick late than Petit or even Cain as Cain is much more likely to lose his rookie eligibility this season.

Some general thoughts on rookies and minor leaguers:

1. Never take a pitcher who has not pitched above Single A. At this stage of development, a pitcher is just too big a risk. Pitcher’s tend to be less likely to skip a level and more likely to be injured than hitters at this level.

2. Don’t take a hitter who had not played above Single A unless he produces solid numbers in at least 4 categories AND has a solid K/BB ratio. You are looking for impact players with your minor league picks. If you just want roster filler, there is plenty of that at the end of your draft. Casey Kotchman is a great example of this theory. He never showed any power or speed at the lower levels. He was a top 5 prospect for a couple of years running despite that. Now he looks like the best he could hope to be is Keith Hernandez, but he shows little sign that he can get to that point. He seems to be heading closer to John Mabry territory.

3. Take players who can have an impact in at least two categories AND who have solid K/BB ratios. Power and/or speed guys who also hit for a decent average and don’t strike out a lot more than they walk that project as top/middle of the order guys who will score a lot of runs. Guys who don’t fit according to this rule would be Dallas McPherson, Ryan Howard and Mike Jacobs. All have awesome power, but terrible K/BB ratios. Guys like that are at best Rob Deer and at medium Russ Branyan—at worst, you never heard of them. Jeromy Burnitz, who is one of the better fantasy low average-high power guys in MLB, at least showed the ability to walk more than half as many times as he struck out in most of his minor league stops. Yes, Howard showed a lot of success in the majors last season, but the smart money says he does not sustain it. He struck out 100 times in 312 major league AB. Before you start comparing him to Adam Dunn, remember that Dunn is actually 10 days YOUNGER than Howard and he walks a lot more than half as many times as he strikes out. Though Dunn’s average is becoming a concern, Howard is much less likely to hit as well as he did last season. This category applies to pitchers as well. Look especially for pitchers who strike out at least 3 times as many batters as they walk. Without command at that level, it’s difficult to make an impact in the majors. Following this guideline would have caused you to pass on Randy Johnson, but you would not have wanted the stats from his first 5 full years in the majors on your fantasy team anyway.

Start with a list of top 100 minor leaguers and then go through the stats and do your own research. Throw out the guys who don’t fit what’s outlined above and you’ll have a list of potential difference makers from which to choose.

Click Here to check out the entire FantasyBaseball.com University Series!


Comments are closed.