Marching With the Army

i3SXAH4AAxtWS

Turns out, Yu Darvish is really good. *legendary GIF from reddit*

Last Tuesday I participated in my first fantasy mock draft of 2014. It was “Opening Day” for Howard Bender’s Mock Draft Army. The idea of the Army is this: bring industry experts together with their readers to talk ball, strategy and life. If you’re into fantasy baseball, I recommend checking out Mr. Bender’s twitter page and personal website for updates. Seriously, it’s a truly kick-ass experience.

The inaugural draft was a 12-team snake, 5 x 5 roto, with 23-man rosters, no bench. The crew was elite, as was the banter. We talked young pitching, Ubaldo Jimenez,  and…  the weather. By draft’s end, the Human_Beans– that’s me– looked like a solid bunch to begin a season with. Shortly thereafter I began work on a spreadsheet to catalogue the results, anxious to find who Steamer–the projection system– would crown King of the Draft (results can be found here ).

Much to my chagrin, I placed near the bottom of the standings. Not surprisingly, Jeff Zimmerman of Fangraphs “won”, by a not-so-small margin (he used Steamer as a guide). I’ve spent last few days poring over the results, extrapolating meaning, and here’s what I’ve come up with:

Lesson 1: Yu Darvish is Good. Really Good.

Last year Yu had 37 more strikeouts than the next pitcher in baseball (¡¡¡277!!!). Every projection system so far has him repeating his dominance of the category. It is within the realm of possibility that he will be the first pitcher in a generation to reach 300 k’s in a season. Given the massive step forward he took in 2013 (he raised his already robust K% from 27.1 to 32.9, while lowering his BB% from 10.9 to 9.5) it stands to reason that 300 could happen as soon as this year. Fantasy-wise, drafting Yu in roto (Yu’s skill set is less valuable in head-to-head) puts you shoulders above the rest in a reliable, important pitching category (strikeouts typically augur well for run prevention as well). The k-cushion allows you to assemble the rest of your pitching staff on the cheap, perhaps stocking yourself with strikeout-neutral options like Alex Cobb or Doug Fister. There are plenty of guys whose rate stats will likely stack up to more expensive options, without the strikeouts. I typically subscribe to the adage “don’t draft pitchers early.” But for me, Yu Darvish provides so much flexibility that he is reach-worthy. I have no problem taking him as early as the mid-late 2nd round.

From a broader perspective, extreme “specialists”–single category outliers– can really put you ahead in roto formats. Jeff also took Billy Hamilton (in the 5th round!). Most of us felt it was early, but if Hamilton stays healthy and in the lineup, he is more than capable of stealing the 65 bags Steamer projects. Like Darvish and his k’s, B-Ham has the ability to secure an entire category on his own, giving you breathing room to be patient and let value picks free fall their way into your little mitts.

Lesson 2: Projection System(s)… Aren’t Perfect.

Steamer, like any system, does the best it can with the information it has. Some player performances it will nail, and others it won’t. In my experience, Steamer tends to be conservative. For example, I took Gerrit Cole with the 108th pick of the draft, just after Sonny Gray and Starlin Castro. When I made the selection, it felt like a coup, considering Cole’s skills, pedigree and promising 2013 performance. Steamer didn’t draw the same wry smile as I:

IP W L K ERA WHIP Player Comp
Steamer 173 11 10 149 3.88 1.31 None
PECOTA 156 9.3 10.5 134 3.33 1.19 Michael Pineda, Brian Matusz, Brett Cecil
ZIPs 163 134 3.48 1.26 Kevin Brown
Oliver 149 10 7 112 3.68 1.29 None
Fans 197 15 8 194 3.13 1.13 None

Steamer was the most pessimistic about my choice, and Oliver wasn’t far behind. However, projection systems tend to have difficulty with young players because of their limited track record. To make up for the lack of information, they usually tie in a MLE modifier (Major League Equivalent). MLE’s translate minor league stats into a “major league equivalent”, using past trends to comprise the equation.  If I’ve learned anything over the years from prospect gurus, it’s to not scout box scores. Looking at minor league stat lines to predict major league success is a dangerous business. In Cole’s limited time on the farm (200 IP over 2 seasons), he posted a healthy but hardly dominant 8.2 k/9. His “knock” as a prospect was that his “stuff” was always better than the results it yielded.

Skip forward to the last 9 starts of Cole’s regular season, and we see a k-rate spike from 6 to over 9/ 9 IP.  What happened? He started using his curveball to great effect. None of these projection systems account for the added usage of a dominant pitch, which could prove to be a statistical game-changer. Furthermore, between his 68 IP on the farm, 117 IP in the Majors, and 11 IP in Post season, Cole was altogether good for 196 IP. The Pirates eased him into the fold upon his arrival to the big club, so it is not unreasonable to count on a solid 200 IP from Cole in 2014 when the gloves come off. No projection system has him anywhere near that sort of output. With Cole, the pedigree is elite, and the stuff is some of the best in the majors. His Steamer projection seems more worst-case-scenario than reasonable expectation. I feel fine about drafting him where I did.

A good way to get ahead in your drafts is to find players like Cole who have added a new pitch, or have somehow taken a step forward in your mind, and plug-in numbers YOU feel comfortable with. Move them up your draft boards accordingly. It’s what I’ll be doing…

- @WordSmithSilva

References:

Fangraphs

Baseball Prospectus

This entry was posted in Fantasy. Bookmark the permalink. Follow any comments here with the RSS feed for this post. Post a comment or leave a trackback: Trackback URL.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>