THT's Fantasy Archives

---

Posted by Karl de Vries at 2:42am

Thursday, April 11, 2013

Blind resume: starting pitcher

In fantasy baseball, as in real baseball, there is a lot of surplus value to be found by acting objectively. While brand name players carry some assuring degree of competence, what we call "the known quantity" factor, they also carry little upside. Marquee players come at marquee prices.

That is not to say they are not immensely valuable. In fantasy, few players achieve, let alone achieve with regularity, the kind of production that Miguel Cabrera offers in 5x5 formats (and that is before you consider his third base eligibility). The extra few bucks you spend at an auction acquiring Cabrera above market rate have a better expected return rate than paying an extra few bucks to acquire a "lower tier" option such as Pablo Sandoval or even a next-tier guy like Evan Longoria. When you pay $45 and get $45 in production, you come out on top because, profit aside, so few players produce $50 worth of value.

The problem with marquee players arises not on the upside, but on the downside. If you spend market value on a top player who has a "down year," a real risk considering there are no guarantees in baseball (this is doubly true for pitchers), even if that down year is 75 percent his normal production, you are losing (or risking, in foresight terms) a lot of money that could have been better allocated elsewhere.

Players you draft are generally valuable because they are some degree better than "replacement level"—that is to say the best player available on the waiver wire in your league. The greater above the replacement level a player is, the more value he has. In some sense, this value can be looked at linearly. If say the replacement level RBI total is 60 in your league, then you can run a standard deviation analysis and calculate how much each additional RBI is worth.

However, this analysis can also be looked at exponentially. You have only a finite number of active roster spaces that can accrue stats for your team. The greater the production you can cram into a single player, the more value he has to your team. If you think of the value above replacement level a given set of players can offer, as you move further away from the replacement level, the pool of players to draw from shrinks.

That is one reason Miguel Cabrera is so valuable and often goes for above market. It is not just that he can put up $40-50 of value, but also that the "next best" third basemen, likely Adrian Beltre, is likely to produce $10 or so less in value. By spending a few extra bucks to cram additional production into your bottom line out of a single player, that gives you extra flexibility at the end of the draft and during the season. For this reason, players in the upper tiers have arguably "above market" value not reflected in their simple production-based dollar values.

The converse of this exponential view is also true. As a player deviates from his expectations and toward the replacement level in a down year, his value flattens. Those +20 RBI become +10, and all of a sudden you could have had that kind of production out of a much easier-to-replace/cheaper-to-purchase player plus additional dollars to allocate elsewhere.

Brand names, therefore, offer not only limited upside, but, if you buy into the exponential value or "uncaptured" value approach, they offer substantial downside—more so than good players in lower tiers who can be got at lower prices.

This is not to say, of course, that marquee players should be avoided. To the contrary, I practice the stars and scrubs approach to fantasy auctions. What it means is that marquee players need to be picked out with care. You should focus more on the downside than the upside when selecting a player and evaluating his value. Players like Cabrera, Robinson Cano and Prince Fielder end up atop my draft board in front of players like Mike Trout, Ryan Braun and Matt Kemp not because I think they are better players or even because I think they have comparable ceilings. Rather, I think the likelihood that my dollar investment in the former set of players is better maximized than my dollar value on the latter set of players. This is my personal risk assessment, as I lean toward calculated decision-making with small incremental victories as compared to the big win or big loss risk-taking approach.

In that vein, let's objectively look at a player who is traditionally ignored in fantasy formats but offers good value, a relatively high floor and no significant injury concern for his price.

This player, Player A, was ranked 320 overall by Yahoo in the preseason. In standard 12 team 5x5 leagues, he was thus deemed an undraftable player. Player A was the 156th most valuable player in fantasy last season. In the chart below, is Player A's 2012 fantasy season against the seasons of nine higher-ranked starting pitchers, none ranked outside the top 300. In fact, the second lowest rated player in the sample (Player C) is ranked nearly 50 spots ahead of Player A.



Striving to find players of comparable value, I avoided players with a history of "elite" overall production. I did not want players like Dan Haren, even if 2012 was an aberration year, in the mix. Second, I tried to avoid players who clearly offered more value than Player A in a single category. I did not want to select a player slightly inferior to Player A or comparable to Player A in all aspect save for, for example, a substantially better strikeout rate. Third, I selected only players ranked higher inside the top 300 overall by Yahoo. There is clearly some personal bias in my selection process, but I believe that I found some good value comparables for the purpose of demonstrating that Player A is undervalued, even without the use of Z-Scores to confirm my selection.

Looking at the mix above, you will not find Player A clearly standing out in any category in comparison to the other nine players. He ranks third in wins, but his win total is only a half win greater than the sample average. On the flip side, you do not find this player lacking in any category in comparison to the other nine players. His WHIP, strikeout rate, walk rate, groundball rate, SIERRA and xFIP are all comfortably within the middle/middle-plus of the pack. Even in terms of the value that Player A provided in 2012, Player A is the sample median.

Player A is not an exciting or game-changing player. Few pitchers who strike out fewer than 7.0 batters per nine innings are. Player A is, however, the kind of key cog you hope can round out your pitching staff with quality innings so you don't have to micromanage and gamble on day-to-day match-ups and the waiver wire.

Of the 21 qualified pitchers who posted a ground ball rate of 50 percent or higher last season, nine had a higher strikeout rate than Player A. Only six had a strikeout rate greater than 7.0:Those six players are A.J. Burnett, David Price, James Shields, Adam Wainwright, Edinson Volquez and C.J. Wilson. Only Price, Shields and Wainwright had equal or better walk rates as Player A did last season.

So far this season, Player A has started two games and pitched 12.2 innings. He has won both outings, and and has a 13:4 strikeout to walk ratio. Player A started 210 games between 2006 and 2012 (an average of 30 per season). His average innings pitched over that span is 184 per season, and he has pitched under 175 innings pnly once (161.1 innings that season) since becoming a full-time major leaguer in 2006. While no 200+ inning work horse, Player A threw 1,291.1 innings between 2006 and 2012—16th highest overall among pitchers over that span. Player A is also not an apparent injury risk. The only time that Player A has been on the disabled list was for a shoulder sprain in the year he threw 161.1 innings.

Care to guess who Player A is? Let's reveal the 10 players listed in the above chart in reverse order

Player J is Derek Holland. Holland owns a career 4.69 ERA. Despite peripherals that indicate he is likely better a better pitcher than the results have shown to date, none of his FIP, xFIP, SIERRA or tERA check in below the 4.0 threshold (OK, his career xFIP is 3.99) in an era in which the average ERA is barely north of 4.0. Yahoo's preseason rankings valued Holland comparably to where he performed in 2012.

Player I is Clay Buchholz, who owns one elite fantasy campaign, one good half-season, one bad one, one average-at-best half-season, and a set of peripherals that do not match his minor league hype. Buchholz has a real nice curveball, but he still needs to work on his other pitches (and stay healthy) if he is ever going to be anything more than a spot starter with flashes of brilliance. Buchholz barley ranked in the top 500 among all fantasy players last season, but was ranked more than 75 spots ahead of Player A in Yahoo's preseason rankings. Buchholz, to this point in his career, is all name and without a reliable track record. At best, he is a medium reward, high risk player.

Player H is Tim Hudson, who is a much better real life pitcher than fantasy pitcher due to his chronically low strikeout rates and reliance on ground balls and a good defense. Hudson is actually a good comparable for Player A, albeit with a few more ground balls. Hudson's 2012 season was pretty similar to Player A's in terms of value provided. He was the 145th most valuable player overall in 2012, 11 spots ahead of Player A. Yet Hudson ranks 100 spots higher in Yahoo's preseason rankings. I would probably take Hudson over Player A—but it is a darn close call.

Player G is Alex Cobb, an interesting player to be ranked 100 spots ahead of Player A. Cobb's minor league numbers indicate that he has more strikeout upside (with good control), and his major groundball rate has been better than Player A's, albeit in under 200 innings pitched. Nonetheless, Cobb has no record of major league success, and ZiPS pegs him pretty close to Player A's expected output (albeit with a few more strikeouts). In terms of upside, I would rather have Cobb over Player A, but if filling out a roster for quality innings rather than upside risk plays, I would rather have Player A. Player A also pitches in the National League for a team with a good defense; Cobb pitches in the AL East.

Player F is Doug Fister, the only player in the list with substantially better control than Player A. Because he had a career minor league strikeout per nine rate of 6.7, I view Fister's post-Seattle strikeout rate skeptically. The lower a player's strikeout rate, the more he has to be better/reliable in other categories. I am not sure I want to bank on any wins of a Tigers pitcher who relies more on the defense behind him than his own skill in converting outs. I view Fister as a slightly better, post-renaissance Joel Pineiro. I am probably in the minority, but I would rather have Player A than Fister.

Player E is Trevor Cahill, who strikes me as another good value comp for Player A. While neither has huge strikeout potential, I think Cahill has better strikeout upside. My biggest knock is that his control is average at best. Cahill's career 55 percent groundball rate goes a long way toward erasing free passes via the infield double play, but Arizona's bandbox is not the most forgiving place to make mistakes. Furthermore, the Diamondbacks defense is likely no better than league average. Player A's team's offense and defense rank among the best in the league, and his home park is pretty neutral. While Cahill is arguably the better "skills" pitcher with youth on his side (he is 26, compared to Player A's age of 31), I think Player A will be more valuable than Cahill in 2013—just as he was in 2012.

Player D is Jarrod Parker, a young pitcher with a lot of promise and upside. Although Parker's rookie season strikeout per nine rate, just under 7.0, was nothing special, his lower-level minor league numbers indicate that he has a little whiff potential to grow into. Parker's upper minor seasons do not give much confidence in an elite strikeout rate, though 2009 and 2011 were bookend seasons to his Tommy John surgery. Parker exhibited good control in the minors, and has regained his control form now three seasons removed from surgery. Oakland is a great place to call your home park, and the A's have a pretty good defense. Parker's skills upside is greater than Player A's, and his "real life" situation (playing in Oakland ahead of a good defense) make him a solid young pitcher to own. Without a doubt, I would take Parker over Player A at the same cost, but I would not pay multifold over for Parker.

Player C is Kyle Lohse, who barely hooked on a with a team this offseason coming off back-to-back seasons that rank among the three best in his decade-plus long career. Lohse is probably the best comp to Player A based purely on his strikeouts and walks. Lohse is decidedly below average in strikeouts for his career (5.7 strikeouts per nine), but so is Player A, who owns a marginally better career rate. Both are stingy with walks. Lohse, however, is four year Player A's senior, and he plays in front of one of the worse defensive teams in the game. Furthermore, the Brewers bullpen is terrible. Lohse is a pretty neutral groundball/flyball pitcher, so Player A takes the edge there as well. I would not bet that either will be better than the other by the end of the season, but at the same price, I would rather have Player A. At a cheaper cost, Player A as the better choice is a no-brainer.

Player B, Jeremy Hellickson, has had a pretty impressive start to his major league career on the surface (3.15 ERA and 1.20 WHIP with slightly above average control), but his career peripherals (4.52 SIERRA, 4.53 FIP, 4.53 xFIP, 4.59 tERA, 6.0 K/9, 38.7 percent groundball rate) paint the story of a pitcher who's been a bit more lucky than he has been dominant.

(Of course, some pitchers are able to outpitch their peripherals and the luck-neutral measuring systems are not great at capturing/evaluating "extreme" pitchers at the margins accurately. Matt Cain (and many of his Giants teammates over the past several years), for example, has thrived off of a notoriously low home run per flyball rate. This offseason, Jeff Zimmerman observed that pitchers who work the edges of the strike zone tend to have an apparent peripheral-results gap. Hellickson was one of the pitchers identified as an "extreme edge percentage" type pitcher, which may go a long way in explaining the nearly run and a half differential between his career ERA and career FIP/xFIP/SIERRA/tERA.)

Hellickson is also on a very good defensive team, and Tropicana is underrated in terms of its offense-suppressing environment. However, in a vacuum, there are more red flags than positive signs when it comes to Hellickson. Edge percentage and youth (age 26) aside, Hellickson's strikeout rate is pretty low (even by American League standards) and his career strikeout to walk rate (1.94) is below average (the league average has grown from 2.2 to 2.5 during Hellickson's major league career). Hellickson is also a slightly flyball pitcher (0.94 career groundball to flyball ratio) with solid, but unimpressive velocity (91.1 miles per hour career). That is not to say Hellickson is a bad pitcher; just that he is not the "ace" his results to date have indicated. He has not pitched any better, from a skills standpoint, than Player A has for his career. To me, Hellickson is to Player A what to Blind Pig is to Pliny The Elder.

Player A...drum roll please...is Paul Maholm. Maholm's not had the most impressive career, and he's certainly had more lackluster years than good ones, but he's been a solidly above average pitcher for two straight seasons. Maholm is a pitch-to-contact player who is no longer playing in front of a horrible defensive alignment, which lends some reason to believe in his recent success. From 2006 to 2011, he pitched for the Pirates, who ranked in the bottom third among all teams in defense in all but one year. In the aggregate, the Pirates were one of the five worst defensive teams during Maholm's tenure.

Given his skills base, his current team's defense and his offense-neutral Atlanta home park, there is a lot to objectively and subjectively like about Maholm. The fact that he is on one of the best defensive teams in the National League goes a long way toward bolstering his win totals. And it counts for something when you realize that he will never have to pitch against that offense. The Braves also play a disproportionate number of their games against the Marlins (who are not very good), the Mets (who are not very good and are rebuilding) and the Phillies (who are decidedly average).

There are really no red flags cautioning against Maholm repeating his 2011-2012 performance level, a performance level ranking in the top 150-200 range overall among players, in 2013. That's a pretty solid return for $1. How he was deemed undraftable in the preseason, especially when players like Clay Buchholz are ranked inside the top 250 overall, is beyond my comprehension.