Wednesday, April 11, 2012
Trader’s corner: week two
Posted by Mark Himmelstein at 5:13amWelcome to Trader's Corner, your one-stop shop for bargains and busts. I've partnered with our good friend Oliver to look at the recent performances of a few players and why they could present a major profit opportunity for you. This won't just be your typical buy high/sell low column, though. As much opportunity as those situations may present, we'll also try to identify the hot streaks that figure to last and the cold spells that could spell doom.
Every two weeks, I'll look at a pair of players in each of four categories: Buy High, Buy Low, Sell High, and Sell Low. The first player will be my own selection and the second based strictly on the Oliver projections.
I'll keep a tally of all my recommendations, the date I made them, and the players' performances from that point forward. From time to time, I'll share the results in an attempt to evaluate how I'm faring and if there are trends to be found.
Each entry will include the player's rest-of-season Oliver projection in the standard rotisserie categories (in the format AVG-R-RBI-HR-SB for hitters and W/SV-ERA-WHIP-K for pitchers). Also provided will be the accompanying projected dollar values according to THT Forecasts' custom price guide for both the standard Yahoo! and ESPN formats.
Dollar values are based on a $260 draft budget with $2 allocated to each bench spot and a 70/30 hitter/pitcher split.
For the first few weeks of the year, I'll also include Average Auction Cost (AAC) figures for both ESPN and Yahoo, since they provide a reference for each player's market price heading into the season.
Week Two's edition of Trader's Corner will include a bit of discussion on players' early season performances, but established values will still be more heavily weighted towards players' preseason prices than 2012 results.
Buy high
Buying high is one of the most difficult and frequently overlooked strategies at a fantasy manager's disposal. We all love to discuss player trends that look promising in the offseason, but somehow, once the season begins, every sample size becomes too small and every unexpected performance a matter of mere luck. The consensus bias shifts from heavily weighting recent performance and "upside" to nigh unshakable temperance and prudence. For the savvy, risk-seeking owner, this can present a great deal of profit opportunity.
This week we'll examine at a breakout pitcher from 2011 that I think Oliver is overlooking and an outfielder who still doesn't seem to be getting the recognition he deserves.
My Pick: Matt Garza
Yahoo! AAC: $11
ESPN AAC: $8
Oliver RoS: 11-3.92-1.29-175
Oliver Yahoo! Value: N/A
Oliver ESPN Value: $4
Despite a poor record last year, Matt Garza was a fantasy asset. While I view the breakout as legitimate, Oliver disagrees. Let's take a look at why we differ.
Although Garza has had strong years in the past, he had never been nearly so good as he was last year. He saw a massive jump in strikeout rate last year—from a career 19.6 percent to a career-best 23.5 percent, a mark that was good for 12th in baseball. This was supported by a similar spike in swinging strike percentage—from a career rate of 8.7 percent all the way to 11.2 percent, tied for fourth-best among major league starters.
His groundball rate also increased from a career 41.1 percent to 46.3 percent while his walk rate was a stable and solid 7.5 percent. That all added up to a 3.19 xFIP and 3.31 SIERA, marks that were 11th and 15th in baseball, respectively.
Oliver sees Garza's 2011 as an outlier. It recognizes his less-impressive prior track record and assumes he will likely move back towards those rates rather than maintain his new-found success.
What I see is a drastic shift in approach that lead to Garza's improved results. Prior to 2011, Garza typically was throwing his fastball roughly 70 percent of the time and his slider a bit less than 15 percent of the time. In 2011, all of a sudden he reduced the frequency of fastballs to a bit over 50 percent of his pitches, while his slider usage jumped to nearly 25 percent.
This approach shift makes sense—over his career Garza's most effective weapon has been his slider. Based on FanGraphs' Pitch Type Linear Weights, his slider has been worth 0.71 runs above average per 100 pitches since he entered the league. His fastball has only been about half as effective, at a solid 0.39 runs above average per 100 pitches—still a quality pitch but not nearly to the extent of the slider.
While a single pitch's effectiveness can't be taken out of context from the rest of a pitcher's repertoire, Garza is still throwing more than enough fastballs to effectively throw his slider off of it. His slider was actually more effective than ever in 2011, an outstanding 1.54 runs above average per 100 pitches. And though any single-season pitch type linear weights mark should be treated as too small a sample size to draw significant inference, the reduced fastball frequency certainly doesn't seem to have diminished the effectiveness of his slider.
Garza's already made one start this year. Exactly 55 percent of his pitches were fastballs and 25.3 percent were sliders. He went six innings, yielded two runs, recorded five strikeouts against just one walk and, exactly half of the balls put in play against him were on the ground. Through one start, Garza appears to be sticking with the approach that made him so successful in 2011.
My money's on the success that Garza found last year following his revamped approach into 2012. His ERA may creep up a bit due to regression, but I'm betting it won't be nearly to the extent Oliver is predicting. If he can luck into a few more wins, he might be the cheapest Top 20 starter you'll find for 2012.
Oliver's Pick: Jay Bruce
Yahoo! AAC: $19
ESPN AAC: $18
Oliver RoS: .264-88-101-32-8
Oliver Yahoo! Value: $31
Oliver ESPN Value: $26
Jay Bruce is a good hitter. I know it, Oliver knows it, Dusty Baker knows it, and the pitchers of the NL Central know it. For some reason, the people responsible for stock fantasy baseball rankings don't seem to know it, and these rankings have anchored Bruce's perceived value quite a bit below where it belongs.
Bruce may not be the most balanced player in the game, but power is an increasingly rare commodity and is something Bruce provides plenty of without really costing you anywhere else. It's nice to get at least a .280 batting average and 20 steals, but .265 and eight are perfectly nominal for the current run environment. Plus, Bruce is still only 25 years old and still may have a bit of room for improvement.
There really isn't a whole lot more to say on Bruce. Oliver sees a slight uptick in batting average, but otherwise, those are basically his numbers from a year ago. As you can see from the dollar values, they're quite a bit more valuable than the amount you can be expected to pay for them. If you're feeling a bit light on power and have room for an outfielder on your roster, don't pass the opportunity to acquire Bruce at (literally) 70 cents on the dollar before his owner realizes what he or she has.
Buy low
Everyone loves a buy-low candidate. The problem is the owner who owns the buy-low candidate usually loves him, too, so you may not be able to buy as low as you wish you could. Still, it's always helpful to identify guys who could see their performance improve in the not-too-distant future.
Much like the Buy High section, this edition's Buy Lows include a mid-priced pitcher who could produce top-tier results and a hitter being punished too much for a substandard batting average.
My Pick: Mat Latos
Yahoo! AAC: $10
ESPN AAC: $9
Oliver RoS: 13-3.65-1.21-200
Oliver Yahoo! Value: $12
Oliver ESPN Value: $13
I'm not entirely sure why Latos is being overlooked in the 2012 market, but I'm even more bullish on his prospects than Oliver is.
The change from Petco Park to Great American Ball Park concerns me, but only slightly. If we were talking about a more extreme flyball pitcher, I would be a bit more concerned, but Latos has a career 43 percent groundball rate. That's just a touch below the major league average.
Latos has given up 201 and 220 fly balls in his first two seasons, eight percent of which have gone for home runs. If instead 10 percent of those fly balls go for home runs, you only have to tack on about four extra homers to Latos's expected total.
According to The Book, the run expectancy of an average home run is about 1.7 (though since Latos is better than the average pitcher at preventing baserunners, it's likely lower for him). That means we would expect four extra home runs to add about 6.8 runs to Latos's season total. That would have increased Latos's 2011 ERA from 3.47 to about 3.70.
The flipside is that the Reds provide a far superior supporting cast for Latos on both sides of the field. They'll not only provide quite a few more runs for the young right-hander work with, but they'll also likely save a few more with their gloves, as well. Over the last three seasons, the Reds as a team have posted a 6.7 UZR/150, second only to the Rays over that span. Their UZR/150 has been no worse than 5.4 in any single season. The Padres had a UZR/150 of 1.5 over that span.
If Latos throws 190 innings, that will account for roughly 13 percent of the Reds' season. If their average fielder has a UZR/150 of 6.5, that would be 5.0 runs better than the support Latos was getting from the Padres. That means each fielder would save an additional 0.65 runs per about 190 innings. Multiply that by eight fielders, and you get 5.2 runs saved while Latos is pitching. That mark would nearly eliminate the damage done by the extra home runs.
Then when you also consider that the Reds are a far superior offensive team to the Padres, it actually becomes likely the trade will have a positive impact on Latos's fantasy value in aggregate.
In two seasons thus far in the major leagues, Latos has posted strikeout rates of 25.3 percent and 23.2 percent, walk rates of 6.7 percent and 7.8 percent, and groundball rates of 44.7 percent and 42.8 percent, respectively. If that sounds an awful lot like what Matt Garza did last year, and what I'm expecting from him again this year, it's because it is. They're very similar-looking pitchers moving forward with very similar costs to acquire. Unlike Garza, Latos had a rocky first start in 2012. Don't let that fool you. If you need pitching, look to get him on the cheap while you still can.
Oliver's Pick: Kelly Johnson
Yahoo! AAC: $2
ESPN AAC: $2
Oliver RoS: .241-79-79-22-12
Oliver Yahoo! Value: $10
Oliver ESPN Value: $15
Kelly Johnson's batting average is a bit of an enigma. Over the last four years alone, it's ranged from fairly helpful to highly detrimental. Because of this volatility, the market seems to be shying away from him. Don't shy with it—your opponents' trepidations can sometimes be your gains.
It's easy to flip the context on Johnson. He's a second baseman hitting near the top of a potent Blue Jays lineup. He'll play his home games in a ballpark that boosts run production. And best of all, he's coming off consecutive seasons of at least 21 home runs and 13 stolen bases. It's stunning to think that such a player can be had for merely a song, even despite a weak batting average.
The downside with Johnson is that the batting average decline appears to be more than superficial. His .277 BABIP in 2011 didn't help. Considering his career mark is .311, we should expect some increase, but the prospects of a .285 batting average and true five-category benefits are likely a thing of the past.
Johnson had been a respectable contact hitter during his career up through 2009, never posting a strikeout rate over 20 percent over a full season. In 2010, his strikeout rate crept up to 22.1 percent, a fact that was largely masked my a .339 BABIP. In 2011, the strikeouts increased even more, to a near-dangerous 26.6 percent. Even with a BABIP boost, a positive batting average is a long shot as long as he continues whiffing like that.
Of course, this doesn't mean Johnson is valueless. As long as he keeps hitting home runs, stealing bases, and drawing walks at a respectable clip, he makes a fine mid-range second baseman. But because of all the batting average variance, it seems the market just has no idea what to do with him. Rather than attempt a somewhat difficult evaluation, owners are neglecting to evaluate him at all.
That's a poor reason to ignore a legitimate threat in both power and speed who should provide plenty of runs and RBIs, as well. If an owner in your league landed on Johnson cheaply but isn't seeing the value, make sure you punish him for it.
Sell high
There may be nothing more satisfying in fantasy baseball than selling a player at his peak value only to watch him crash and burn for another owner while you reap the benefits of said owner's former studs. It happens every year, whether it was Michael Pineda's second half swoon in 2011 or that time that closer saved 20 games in the first half only to blow four in a row and lose his job. You remember that guy, right?
This round of Sell Highs includes a pitcher whose price tag may not match his value and an example of why the keystone is the most misunderstood position in fantasy baseball this year.
My Pick: Ian Kennedy
Yahoo! AAC: $17
ESPN AAC: $13
Oliver RoS: 14-3.58-1.18-198
Oliver Yahoo! Value: $15
Oliver ESPN Value: $16
Ian Kennedy sure had a fine year last year. He notched 21 wins, a 2.88 ERA, 1.09 WHIP, and 198 strikeouts. He induced swinging strikes at a high rate, avoided bases on balls, and threw plenty of first pitch strikes. So what's there to be bearish about with this emerging ace?
Honestly, I'm not that bearish on Kennedy. The selection flows largely from the two buy candidates I chose for today's edition. I simply don't see a huge difference between Latos, Garza, and Kennedy, and actually prefer the former two.
Garza's 2011 xFIP was by far the best of the three, at 3.19, while Kennedy's 3.50 mark was nearly identical to Latos's 3.52. Meanwhile, Latos was actually better in 2010 than he was in 2011, while Kennedy has never been as good as he was in 2011 by any measure.
Let's dig a bit deeper, though. There are two major factors with Kennedy that point to negative regression.
The first is his combination of his flyball proclivities and low home run-per-fly ball rate from 2011. He typically generates ground balls on fewer than 40 percent of the balls he allows to be put in play. Unlike Garza and Latos, this mark is well below than the league average.
Only 7.7 percent of Kennedy's fly balls left the park last year. Although not quite the bandbox Great American Ball Park is, Chase Field also produces quite a few long balls (and has actually had the higher combined park factor over the last few seasons). As such, whatever home run regression we're tabbing Latos for, we also have to tab for Kennedy.
However, unlike Latos, Kennedy doesn't have the stable track record of generating strikeouts and walks, with career rates of 20.4 percent and 8.1 percent, respectively. Kennedy also won't get the bonus of an improved defense to help offset the extra home runs we can expect him to give up.
In fairness, Kennedy did see sizable spikes in both his rate of first-pitch strikes and swinging strikes 2011, which supports the increase in strikeouts and decrease in walks. But, unlike the case of Garza's breakout, it's difficult to tie this to a major change in Kennedy's approach. His fastball velocity and pitch mix were essentially the same as they had always been.
Looking at Kennedy's Brooks Baseball card, we see that he did have a career-best whiff rate on his fastball at 7.31 percent. That's a solid number, but not enough to explain Kennedy's sudden dominance. His fastball had more movement than it did in the past—both on the horizontal and vertical planes—and while this likely helped, it seems an unlikely reason for the drastically improved results.
And this brings me to the second major regression factor that's begging for attention but also easy to miss. Kennedy's fastball, which averaged just 90.3 mph on the radar gun, graded as 28 runs above average in aggregate by Pitch Type Linear Weights in 2011. Since 2002, the earliest FanGraphs provides data on average pitch velocities and pitch type linear weights, only five right-handed pitchers have aggregated at least 25 runs above average with a fastball that averaged fewer than 91 mph in a single season.
Three of those pitcher seasons—Derek Lowe in 2002, Brandon Webb in 2006, and Tim Hudson in 2010—can essentially be ignored. They all had groundball rates of over 60 percent and represent a unique class of pitchers who generate outs in a specific way that has more to do with movement than velocity.
The other two examples were Chris Young in 2007 and Mike Mussina in 2003.
Young is a complete oddity and not much like Kennedy, either. Although much of his recent career has been lost to injury, most of the success he has enjoyed can be attributed to his career .248 BABIP. He's also 6-foot-10 and has surprisingly mediocre command. I don't see much to glean here.
Mussina presents a much better comparison. He likely had many years before 2003 where he got quite a bit of mileage out of a fastball without huge velocity. Still, 2003 was one of the best years he ever had and not a great example of a season he was able to replicate with regularity. His 3.09 FIP was the second-best mark of his career (xFIP also isn't available before 2002), and included a minuscule 4.7 percent walk rate.
It's very difficult to believe Kennedy can continue to generate such spectacular results—specifically via swings and misses and fly balls—with on a fastball having such modest velocity. Since pitch type linear weights are not defense independent, this issue probably overlaps the expected increase in his home-run rate. If anything, though, it probably better indicates the full extent of the expected regression. While it's possible the added movement on Kennedy's fastball was the major factor in his improved results, without other improvements, it's probably not a means to sustained dominance.
Kennedy is being drafted alongside potential aces like Madison Bumgarner and Matt Moore. If you bought him at that price, you probably should shop him around and see what you can get. You may be able to net a pitcher like Garza or Latos, who should produce very similar results, and then some.
Oliver's Pick: Robinson Cano
Yahoo! AAC: $46
ESPN AAC: $36
Oliver RoS: .305-81-91-22-4
Oliver Yahoo! Value: $29
Oliver ESPN Value: $25
The reason Oliver is selecting Cano this week isn't so much because it disagrees with the evaluation of Cano in particular, but in the market's evaluation of the keystone as a whole. Oliver concurs with the market than Cano is in fact the best second baseman going into the year. What Oliver disagrees with is that this makes Cano is a first-round value.
In the ESPN format, Oliver ranks a whopping 21 players at second base as superior to the replacement level. That's more than either third base or shortstop. Oliver also sees a fairly moderate performance variance between these 21 players. The market is simply neglecting to account for these factors.
The result of the market's inability to recognize this lack of projected variance is that most high-end second baseman are tabbed as overvalued and most low-end second baseman as undervalued. There's roughly a $20 projected profit swing from a player like Cano, who may cost you upwards of $10 at his market price, against players like Kelly Johnson, Danny Espinosa, and Ryan Roberts, who can all be had for pocket change and net around $8-10 in profit.
The general conclusion is that if you spent top dollar on a second baseman, it might be wise to shop that player around and see what's available. You can probably get a solid replacement second baseman as a secondary throw-in as part of a larger deal, or even find a valuable replacement on the wire. You'll be much better off with Carlos Gonzalez and Jose Altuve than with Robinson Cano and Lorenzo Cain.
Sell low
If selling high is one of the most enjoyable acts of a fantasy baseball season, selling low is one of the most painful. Admitting sunk cost is difficult, but there is opportunity in these situations when the admission is managed. Many times, other owners will pay above a player's projected value out of a misguided instinct to buy low or on name value alone. Even if the return price is below the price you paid, it may still be well more than the price you'd earn in keeping a broken player on your roster.
We'll wrap up this week with a couple outfielders who will give you solid production in a category or two but who probably won't provide enough total value to justify their prices.
My Pick: Matt Holliday
Yahoo! AAC: $23
ESPN AAC: $21
Oliver RoS: .304-73-79-20-5
Oliver Yahoo! Value: $17
Oliver ESPN Value: $18
It was just a year ago than Holliday was a virtual lock to be drafted in the second round of a 12-team draft. After missing a month 2011 with a bad quad, the price to acquire him has dropped a bit but probably not enough to justify the investment.
Holliday spent most of his mid-to-late 20s as a true five-category performer—someone who could hit 25-35 homers, steal around 15 bases, and bat over .300 while putting up gaudy totals in both runs and RBIs.
At least one element of Holliday's game has almost entirely evaporated since those days—speed. After attempting no fewer than 14 steals in any full season prior to 2011, he attempted only three last year. While it's tempting to write this off to his balky quad, he'd previously set a career low in both stolen base attempts and success rate in 2010 by getting caught five times in 14 tries. Expecting more than a handful of steals is no longer reasonable.
There are a few other troubling signs for Holliday, as well.
First, he's no longer a true 30-home run threat. He hasn't reached that mark since 2007. It seems unlikely he does again any time soon.
Second, his strikeout rate of 18 percent in 2011 was the highest it's been since his rookie campaign. With no speed left and power that's more good than great, Holliday absolutely can't afford to start striking out more frequently. Batting average is a huge part of his value, and if that starts to decline, it will be almost impossible for him to justify his price. It's not quite all-hands-on deck in the category yet, but it's a risk that bears watching.
Finally, his groundball rate inched up a bit in 2011, too. The 45.9 percent mark he posted last year was the highest he's had since his sophmore season of 2005, when he hit "only" 19 home runs in 526 plate appearances. Again, like the strikeout rate, this may be just noise, but it's something that bears watching.
At the very least, there's more risk than ever before that both his power and average decline in the near future, something fantasy owners absolutely cannot afford.
At age 32, the increasing possibility of injury and decline are starting to loom large for Holliday. If he does tally 600 plate appearances, it's still likely he hits around .300 with 25 home runs. But with no speed to back that up and no Albert Pujols to drive him in so regularly, there's not nearly as much supporting those totals as there once was.
The best-case scenario seems to be that Holliday justifies his cost but provides little to no profit. Oliver believes you can actually get similar, if not better, production out of players like Jay Bruce, Michael Morse, Pablo Sandoval, and Jason Heyward for a percentage of the cost. If someone is willing to buy him as the more-than-$20-player he used to be, don't hesitate to sell.
Oliver's Pick: Brett Gardner
Yahoo! AAC: $10
ESPN AAC: $10
Oliver RoS: .259-73-38-6-36
Oliver Yahoo! Value: N/A
Oliver ESPN Value: $8
Oliver is a bit unsure of how to handle Gardner. It views him as less than a Top 40 outfielder, meaning he won't carry significant value in the three-outfielder Yahoo! format. In the five-outfielder ESPN format, it sees an acceptable player, but not someone worth paying a significant price for.
In fact, Oliver is quite fond of Gardner on a rate basis. The problem is almost exclusively one of volume, as Gardner is only projected to receive 542 more plate appearances this season.
Unfortunately, in this case, it seems Oliver may be correct on the playing time issue. Joe Girardi is fairly set on getting Andruw Jones into the lineup against left-handed pitching, and this will largely come at Gardner's expense. Then even when Gardner is in the lineup, he's been penciled into either the eighth or ninth spot in every game thus far. This is not a good combination for his chances of accruing volume in the counting categories.
There's an interesting twist here, though, and a good example of something raw projections fail to account for. Gardner will likely in fact be more useful in the Yahoo! format, where you have deeper benches and fewer total games to fill, than the ESPN format, where the opposite is true.
Although the totals he's expected to produce leave him at roughly the replacement level in Yahoo!, the format's deeper benches and lower cap on games played give you the flexibility to roster other useful players who can fill in on days Gardner sits. On the other hand, in ESPN, the deeper starting lineups and diminished flexibility that comes with fewer bench spots means volume per player is absolutely crucial.
In the latter situation, there's not much difference between Gardner and a player who would give you identical totals in 650 plate appearances. In Yahoo!, that's a massive difference, since you can micromanage when you deploy Gardner and when you deploy other supporting players.
As a result, contrary to Oliver's recommendation, I would be more inclined to take a chance on Gardner in the Yahoo! format than ESPN. But when you combine the raw projection with these decision theory aspects, Oliver would recommend staying away unless the price is severely reduced in either case. If an owner in your league is short on speed and willing to pay a premium for Gardner's steals, make sure you don't stand in his way.
THT Forecasts
If you're curious about the projections and dollar values provided, make sure to check out the THT Forecasts section. For $14.95, you get full access to the Oliver projections for thousands of major and minor leaguers, including six-year major league equivalency (MLE) forecasts on every player card.
And best of all for us fantasy junkies, you get full access to THT's custom fantasy price guides, which allows you to create your own price guide based on your league settings and play-style preferences using the Oliver projections, with projections and dollar values updated throughout the season.
Mark Himmelstein is a philosophy student studying at SUNY Empire State who chooses to employ his analytical mind towards the wonderful world of Fantasy Baseball, rather than wasting it on something superfluous. He welcomes questions and comments via E-mail.
.
.
Unless you’re trading good players for under preforming good players, IMO you should hold onto your roster and let the first few games play out.
Of course I could just be saying this cause I own 4 of the 8 guys named in the article.