Daily fantasy: splits decisions

One of the most debated issues in our game is the relevancy of the countless split stats that are readily available with just a few clicks on the web. With sabermetrics becoming increasingly used, there seem to be more and more ways to analyze, and attempt to predict, player performance. That said, I’ll outline the three splits most worthy of consideration while doing daily fantasy baseball research.

This information should prevent you from committing one of the ultimate no-nos when it comes to daily fantasy baseball preparation: “digging too deep.” I’ve seen it from so many players, both new and old. They think so long and so hard about so many different variables that their thinking becomes clouded. We’ve all heard the classic slogan “go with your gut,” and while I’m not advocating this completely, I do think there is something to be said for not over-thinking things. I’ve found that some of the best, most profitable teams I’ve put together, have come from very basic research in a reasonably short amount of time.

A quick note. Not included in these splits decisions are past week performance and specific pitcher vs. batter stats. If you’ve been reading me, you know I advocate these two statistical measures more than any other. If you haven’t, then shame on you, but also have a quick look at last week’s Movin’ On Up and previous article Step your prep up.

So here they are, the three most relevant splits, in no particular order.

Lefty/Righty

This is the most obvious, and most used split stat in daily fantasy baseball. If you spend any significant time in the forums or chats of the various sites, it will become clear to you that the first thing that pops into the heads of the experienced player is the dominant hand of the hitter and the dominant hand of the pitcher he is are facing. But while this is the most relevant split, it is also the most misinterpreted.

The biggest misinterpretation is that lefties can’t hit lefties. This is not the case. While some lefties look lost against lefty pitchers (see, Adam Dunn), others hit them quite well—so well in fact, that they could give you a leg up in your match-up over a competitor who blindly doesn’t play Adrian Gonzalez because he is facing crafty southpaw Ted Lilly. In reality Gonzalez has a career OPS of .933 against Lilly in more than 20 plate appearances.

The same type of assumptions can be easily made for pitchers, along with complete ignorance to pitching statistics. To stick with our Red Sox examples, lets use Kevin Youkilis. He thrashes typical left-handed pitching, to the tune of a .328 batting average and an OPS of over 1.000 over the past three years. Both those statistics are significantly higher than his stats versus righties over that same period. So some would automatically roster him against the likes of say, C.J. Wilson. But not so fast. Mr. Wilson has a very respectable ERA of 2.65 versus righties in 2011. And not only that, he has given up pnly extra base hit to Youkilis in his career. Bet you didn’t know that.

So when using lefty/righty splits make sure to analyze the guy in the batters box as well as the man on the mound. You might find some interesting things if you take the time to dig deep enough. Here’s a top five list you might find useful as you using lefty/righty splits in the future:

Top five batting averages for lefties against lefties for their career (since 2002)

(1) Ichiro Suzuki – .341
(2) Todd Helton – .311
(3) Juan Pierre – .303
(4) Joe Mauer – .301
(5) Robinson Cano – .298

For more on the relevancy of lefty/righty splits check out this cool article from Baseball Professor, Complete MLB Batter Split Stats vs. LHP/RHP

Home/Away

There are many, many things that you must consider when deciding whether to value the home/away stats of any individual player. Many variables that can skew these stats one way or the other have absolutely nothing to do with a player playing at home or away from home. Some of those will be presented in the next section that discusses stadium-specific hitting. Others include weather (temperature and precipitation), playing through injury and things as simple as pitcher match-ups But home/away splits are a variable on their own, independent from the stadium aspect, and these splits can be very telling for certain guys. These stats are most relevant when sustained over a long period of time. Some players just hit better with the support of fans and others may get ridiculed in their home stadium, making them far more successful on the road.

Here are three players who have shown statistically significant differences in batting average between home/away stats since 2002

(1) David Ortiz (.308 home/.265 away)
(2) Michael Young (.325 home/.285 away)
(3) Bobby Abreu (.303 home/.279 away)

You can find pretty much anything you need on Fangraphs. Check it out and see if you can find anything significant that I missed. That’s the fun thing about splits, there’s always more to be found. But for every significant thing you find, you will encounter and have to determine the relative unimportance of countless other splits. It’s a test of your inner patience. Good luck!

Stadium-specific hitting statistics

Last, stadium factors. We all have preconceived notions about which parks are hitter friendly and which ones are not. And for the most part you are probably spot on. But make sure when you pick players based on what stadium they play in, you also factor in the other variables on that given night (like weather and pitcher). Just to put your mind at ease, I wanted to present you with the three most hitter-friendly, three most pitcher-friendly, and four most neutral fields in the league, as rated byThe Sporting News.

A Hardball Times Update
Goodbye for now.

Hitter-friendly

(1) Yankee Stadium
(2) Coors Field
(3) Chase Field

Pitcher-friendly

(1) Petco Park
(2) Safeco Field
(3) Oakland Coliseum

4 Neutral

(1) Tropicana Field
(2) Kaufmann Stadium
(3) Sun Life Stadium
(4) Nationals Ballpark>

Make this an important consideration when finalizing the tough decisions for your daily lineup. Give one of my favorite sites a try:

FanDuelDraftStreetDraftZoneFantasy Sports LiveSportsGeek

Free to contact me to complain, chat or talk about your triumphs. See you next time!


4 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Will Hatheway
12 years ago

Why did you choose The Sporting News? I am interested in their methodology versus, say, Stat Corner, ESPN, or THT’s own annual. Specifically, I am curious about Yankee Stadium as the most hitting friendly park, because I wonder about the extent to which the offensive prowess of those in the AL East impact it (I hope to god that the Yankee’s own lineup’s prowess was factored out).

Anything to enlighten me? Or thoughts on which methodology is best and why? Thanks in advance…

Will Hatheway
12 years ago

Another question (really, just curious):

You imply that A) streaking is real and that B) small sample sizes of specific batter/pitcher matchups pale in the face of their predictive relevance… do you have data to back these up? I really appreciate your sharing that they are the most important things and also sharing sites that let you capitalize on that knowledge, but am wary of taking someone’s word without having evidence to back it up, so any would be much appreciated. Thanks!

Kevin Cearnal
12 years ago

Okay Will here is my best attempt at giving you what you asked for. Please keep in mind that I am bluntly honest and that I do not consider myself a stat junky like many of my colleagues:

Question 1:

Sporting News was the most readily information that was sitting in front of me at the time. Plus it was the first projection for 2011 park rankings I could find. It actually came from their 2011 Fantasy Prep Magazine. The system they use for evaluation is called Ballpark Power Index (BPI) and it is unique to the other ranking systems for ballparks I have found. From the magazine, “BPI is calculated by averaging its season rankings among all major league fields in three categories: home runs per game, runs scored per game and slugging percentage. By using slugging percentage instead of batting average, extra-base hits—particularly home runs—carry more weight.”

ESPN on the other hand keeps a running count on ballpark factors individually (i.e. homeruns, runs, hits, etc…) in comparison to one another. If you look here you can see the stats thus far in 2011:

http://espn.go.com/mlb/stats/parkfactor

As you can see these results are far different than the preseason projections offered by Sporting News. But individual stats are rated far differently I believe.

When I consider park factors I simply examine what offenses are playing in the top 6 or so hitters’ parks as well as the top 6 or so pitchers in pitcher friendly parks. I’ve never given extensive thought to what ranking system would most accurately project to daily fantasy. Too many other factors to consider. Plus I have a regular job I must spend a little time on.

If you want any more specific information on this topic please let me know.

Kevin Cearnal
12 years ago

Question 2:

A) Streaking is real. Period. At least thats what I whole-heartedly believe. I’ve heard every argument and read every article against it, but as a daily fantasy baseball player, 2 concepts have formed my unflappable opinion on the matter. The first is that I have played baseball (fairly competitively) and I’ve seen and experienced streaks. They happen. They just do. The second is that I would pick a player I believe is ‘streaking’ over a player with good pitcher priors EVERY SINGLE time in daily fantasy. Say for instance your choosing between Pujols and Votto (same price) and Votto has far better priors against his pitcher for the night, but Pujols has been red hot. With all other factors being equal (park, weather, etc…) I would take Pujols without question in that situation. Experience (personal profits) have told me that is the right move.

B) Small sample sizes have more predictive value than no sample size. Although I wouldn’t give a great deal of onus to 4 or 5 ABs. Its all about weight the different splits. Lets say for example that Carlos Quentin is in Safeco facing King Felix. And he has fairly good splits against him (i dont know if this is true or not). I certainly wouldn’t take him over say Nelson Cruz in Arlington against some pitcher he’s never faced like Andrew Oliver. I’m just saying that you should always consider these priors before making selections. But I’m not at all advocating following them blindly. I’m more of a gut guy than almost any other player I know in fact, but I still stand by my statement and opinions/rankings of the splits I provided.

My backup is my profit margin. Thats what matters most to me.
Thanks for working my brain Will, and please do let me know if you want to discuss anything further.

Kevin