The Trendy Fantasy blog
If you want industry leading statistical analysis to improve the state of your Fantasy teams, you've come to the right place. CBSSports.com data analyst Al Melchior will be breaking down numbers and trends all season with the help of some powerful interactive data visualizations in his data-driven blog.
Editor's note: If you want industry leading statistical analysis to improve the state of your Fantasy teams, you've come to the right place. Data analyst Al Melchior will update this blog multiple times weekly with trends and interactive visualizations to assist you in your quest for a Fantasy title!
Do you have any thoughts or suggestions for a future blog entry? Click here and drop us an e-mail.
Tuesday, Sept. 21 ... How to avoid homer-prone hurlers
When we look for an edge as we set our Fantasy rotations week in and week out, one place we frequently go to is to look at a pitcher's home stadium environment.
Those of us who put stock in using park factors as an analytical aid know that pitchers fare well in places like PETCO Park and Oakland-Alameda County Coliseum, while they pitch at their own peril at Coors Field and U.S. Cellular Field.
Certain cases reinforce our notions about ballparks making a difference in Fantasy. Look at what happened when Javier Vazquez moved from neutral Turner Field as his home park to homer-happy Yankee Stadium? The change in venues is an obvious factor, but less obvious is the fact that Vazquez's flyball rate soared from 36 percent in 2009 to 47 percent this year. Worse yet, he gave up homers far more often when he did allow flyballs. In his case, all of these factors have likely played a role in Vazquez's regrettable season, one in which he has allowed 29 home runs in 148 2/3 innings.
On balance, though, how much difference does a pitcher's home park really make? To look at the graph below, it appears not much, particularly at the extremes. Pitchers have been color-coded according to their home stadium's home run park factors. (Three-year trend data from the Bill James Handbook 2010 were used. Parks with factors of at least 110 have been classified as "home run friendly", those no higher than 90 are "home run unfriendly", and all others are "home run neutral.") Of the seven pitchers who have allowed no more than 0.4 home runs per nine innings this season, none pitches in a home park that has a recent track record of being a bad home run park, but two pitch in homer-friendly parks. Six of the seven, though, have posted flyball rates below 35 percent. Moving from the lower-left to the upper-right corner of the graph, of the nine pitchers with HR/9 rates above 1.5, four have pitched in homer-friendly environment, while two call pitcher-friendly parks home. Six of these pitchers allow flyballs on at least 40 percent of hit balls, and three of the four most prolific home run pitchers have flyball rates in excess of 45 percent.
This goes to show that you can take the flyball pitcher out of the home run park, but you can't keep him from giving up home runs. Witness Ted Lilly , who left the pitcher-unfriendly confines of Wrigley Field just to see his home run rate rise with a move to Dodger Stadium. Lilly will be a free agent this offseason, but owners shouldn't spend much time worrying about where he'll land, unless it could have a dramatic effect on his win total. Pitchers can see their home run rates fluctuate, but that has at least as much to do with chance as it does with changes in their venue. Many of the pitchers in the graph, from flyballer Lilly to ground ballers like Chad Billingsley , have seen their home run per flyball rates fluctuate greatly from year to year. (Click on some pitchers' marks to see how much their yearly rates vary.)
The lesson for Fantasy is to target pitchers based on their flyball rates rather than on the stadiums in which they play their home games.
Thursday, August 26 ... Who are the true men of steal?
Some e-mailers have recently asked about how much weight to place on a player's stolen base trend, given that there is so much inconsistency in players' steals totals from year to year. Steals do seem to spike and dip more than other stats, but the questions made me wonder how much there is to this perception.
At least judging by some players who were recent stolen base leaders, there is a lot of volatility in yearly stolen base totals, but there is also a surprising amount of play in the yearly home run totals and batting averages as well. I chose to look at these three stats, as they are the least team-dependent of the five standard Rotisserie categories. I have also chosen to look at the leaders in these categories from 2008, so that we could see how each player trended coming into the season as well as how each one did in a full season after being among the leaders.
Of the top 10 home run hitters from '08, four experienced a three-year spread of no more than five homers from their most prolific season to their least prolific. A fifth hitter, Adrian Gonzalez , was also fairly steady, as his home run count climbed from 30 to 36 to 40. In the batting average category, just four of the '08 leaders kept their three-year spread to 30 points or less, and only Nick Markakis ' spread was less than 25 points. Based on this small group of recent stat leaders, it seems that it may be realistic to expect top home run hitters to add or lose 10-plus homers in a given year and the best hitters for average to fluctuate by as much as 40 points.
Among the steals leaders, only Jimmy Rollins and Ichiro Suzuki have spreads lower than 20 stolen bases. Granted, some of the larger spreads (i.e., Jose Reyes , Jacoby Ellsbury , Michael Bourn ) are the product of injuries or variations in major league playing time, but the frequency of large double-digit swings is hard to ignore.
This is, of course, a very cursory look at variation for a handful of statistical categories, but this small study does produce two glaring results. For three categories that are relatively team-independent, it is reasonable to expect the top players to experience a notable degree of variation from year-to-year. Even for reliable performers like Ichiro and Albert Pujols , lockstep consistency is not the norm. That said, it appears to be especially difficult for top base stealers to maintain an elite level of performance year after year. Consider this: of the top 10 stealers from two seasons ago, six failed to exceed the 25- steal mark within one year (either before or after) of being among the stolen base leaders. This doesn't suggest that owners should punt on stolen bases, but it does prove that the most consistent producers in this category, like Rollins and Ichiro, are extra valuable.
Tuesday, August 17 ... The stat behind Hudson's success
Tim Hudson is on the verge of doing something unusual. When he faces the Nats on Wednesday, he could become the first pitcher in Braves' history to make six straight starts of six innings or more with no more than a single run allowed. This streak is impressive enough, but he has been as consistently effective as any pitcher all season. Hudson has allowed more than three earned runs in just two of his 24 starts.
If that weren't eye-opening enough, the way that Hudson is achieving success is extremely unusual. If the season had ended on Monday, he would rank as one of 15 starting pitchers from the last four years who completed a season with a WHIP below 1.10, but out of this group, only he and Chris Carpenter have managed the feat with a strikeout rate below 7.0. And with a K/9 rate of 4.9, he isn't just a little bit below the 7.0 threshold. To put it another way, he is unique among recent starting pitchers in his ability to limit baserunners without limiting contact.
Hudson is an extreme ground ball pitcher, but there are plenty of groundballers who do not sport impressive WHIPs or ERAs. (Hudson's 2.13 ERA is currently the second lowest in the majors.) His true secret to success is his consistent ability get batters to hit either grounders or flies, but not line drives. Over the last four seasons, batters have hit 1,813 balls off Hudson, but only 251 have been line drives. That 13.8 percent rate is more than a percentage point lower than the next lowest rate, and he, along with Roberto Hernandez and Hiroki Kuroda , are the only pitchers to compile a line drive rate below 16 percent between 2007 and 2010.
What this tells us about Hudson -- and to a lesser extent about Carmona and Kuroda as well -- is that we can expect him to pitch much better than his strikeout rate would suggest. He can achieve results similar to Jered Weaver or Josh Johnson but with about half the strikeouts. This also tells us that pitchers with chronically high line drive rates, like Jason Hammel and Aaron Harang , could wind up being perennial disappointments.
Wednesday, August 4 ... Are there simply more elite pitchers?
With the bevy of no-hitters and perfect games being tossed this year, we've all become familiar with the proclamation of 2010 as the "Year of the Pitcher." True enough, run scoring is down again this year, falling from 4.6 team runs per game last year to 4.4 this season. With the overall shift from hitting to pitching, it seems like the preponderance of elite pitching performances has been on the rise.
There is some evidence to support the idea that the upper ranks of starting pitchers is growing. In each of the last five seasons, there has been no year in which there have been more than eight starters to finish with an ERA below 3.00 and a K/9 ratio of at least 7.0. Currently there are 13 starters on that pace for both metrics. While we are deep into the season, we would be remiss not to notice that there are still two months left for players to regress to the mean, which means that some of these 13 exceptional pitchers may not look as exceptional by the time late September rolls around.
At least four of the current "elite" look to be at risk of a dropoff over the next several weeks. Even though he may only pitch the last five or six weeks of the season, Andy Pettitte is probably the most likely pitcher to see his stats come back to earth. He has been unusually good at stranding runners, leaving them on-base at an 80 percent clip, but that rate should head toward the lower 70s region more typical of his fellow Yankee starters and his historical numbers. Jaime Garcia strikeout to walk ratio, which is barely over 2.0, does not fit the typical profile of an elite pitcher, and his Component ERA (ERC) estimates that his ERA probably should be 3.17 -- good, but not great. (To see the ERC and K/9 trends for each of the 13 pitchers, click on their bars in the graph below and see the data displayed below the graph.) Mat Latos has benefitted from a strikingly-low BABIP all season, and odds are that his .238 rate will fall soon, even if he has maintained it for this long. Finally, David Price has cut a run and a half off his ERA due to a dramatic reduction in homers, but can we buy into his plunging home run per flyball rate (11.5 percent in '09 vs. 7.2 percent in '10)? According to data on the Hit Tracker website, his luck with giving up "cheap" home runs hasn't changed much, so the dramatic one-year improvement is hard to accept.
So are we moving into a more pitching-dominant era? It looks that way. But are there more dominant pitchers? I will be much more likely to believe that if the landscape looks the same come October. Until then, I will be viewing elite talents like Josh Johnson , Roy Halladay , Cliff Lee and Tim Lincecum (whose ERA has actually risen above 3.00) to be as just as scarce as I did back in March.
Tuesday, July 27 ... Supermodels of consistency
As you suffer through your Fantasy hitters' slumps, you may wonder if there's enough of a payoff at the end of the slide to make the wait worthwhile.
Not surprisingly, most of the hitters who have the greatest fluctuation in Fantasy scoring week-to-week are top producers, since their spikes are higher than those of merely average players. Three of the 20 highest scorers in standard Head-to-Head leagues also happen to be among the five most inconsistent hitters, as measured by average weekly variance in their Fantasy points (minimum 290 Fantasy Points through Week 16). Joey Votto , Aubrey Huff and Adrian Gonzalez have all experienced average weekly shifts in their Fantasy scoring of nine points or more this season. Alex Rodriguez and Carlos Gonzalez , no Fantasy lightweights themselves, round out the top five in weekly scoring variance.
In each of those five cases, clearly the ups outweigh the downs, even when the slumps are persistent. Huff and both Gonzalezes have experienced downturns this year in which they registered three consecutive weekly Fantasy Point totals that were below their weekly average after 16 weeks. Being inconsistent doesn't have to mean that one is prone to extended slumps, though. Neither Votto nor A-Rod have yet to experience a stretch of more than two weeks in a row with below-average production.
If you place a premium on consistency, the five hitters at the bottom of the graph may be more your speed. Though he has something of a reputation for inconsistency, no one has been more consistent from week to week than Prince Fielder this year. If it seems like Martin Prado , Evan Longoria and Shane Victorino crank out points consistently one scoring period after another, it's because they have, as they also qualify to be among the five most consistent hitters. Bobby Abreu has been equally steady, though owners wouldn't mind a hot streak from the underachieving Angel.
This is just more data to underscore the point that long-term trends matter in Fantasy. Even star players can get stuck in the doldrums for the better part of a month, but a recovery can come at any time. Huff may have provided more heartaches this year than Abreu, but after four months of play, Huff has given his owners a lot more to cheer about. When the next slump comes for Huff, CarGo or the others featured here, the best thing you can do is to let them play it out on your active roster.
Friday, July 23 ... Putting Strasburg in perspective
Now that the Stephen Strasburg hype machine has downshifted out of overdrive, now is as good a time as any to see if his performance has justified all of the hoopla.
Yes, we know that he has been very good, but how has he stacked up so far against other young phenoms? The graph below plots the ERAs and strikeout per nine inning ratios (K/9) for starting pitchers during their first season in the majors. All starters with at least 50 innings in their first year going back to 2000 are included. Strasburg's K/9 rate hasn't been as gaudy over his last six starts as it was for his first three, but his overall rate still dwarfs that of every-other first year pitcher, other than Mark Prior. The ex-Cubs' 11.4 mark is still more than a full strikeout below Strasburg's.
Strasburg's 2.32 ERA is the second-lowest among all of the pitchers in this data pool, trailing only Zach Duke 's 1.81 ERA from 2005. Like Strasburg, Duke pitched fewer than 100 innings that season, and with more starts, he surely would have seen his 2.6 percent home run per flyball ratio -- and his ERA -- rise. No other first-year starters have mustered an ERA below 2.50, but Jered Weaver , Felix Hernandez , Barry Zito and Brandon Webb all came within a half-run of Strasburg's current mark. Even if the Nats' righty regresses towards a 3.00 ERA, he has some pretty impressive rookie comparables in Weaver et. al.
Assuming he can maintain something close to his current level of performance the rest of the way, Strasburg will have given us the best first-year performance by a starting pitcher in this young century, and the contest isn't particularly close. In Fantasy, he has been traded straight up in CBSSports.com leagues for the likes of Carlos Quentin , Curtis Granderson , Prince Fielder and Jimmy Rollins in recent days. While he is likely to be shut down by early September, owners making a final push towards their Head-to-Head playoffs should note that they could acquire the best young pitching talent in years for their stretch run in exchange for a slightly less-than-elite position player. And now we can slap these kinds of superlatives on Strasburg with something more than mere conjecture to back them up.
Friday, July 16 ... Post-break breakouts and busts
Sometimes our memory serves us well in Fantasy, and sometimes it distorts the truth.
Owners may remember last season that Matt Holliday and David Ortiz caught fire after the All-Star break, while Justin Morneau and Aaron Rowand fell off the map after solid first halves. We might be tempted to label Holliday and Ortiz as "second-half players," while viewing Morneau and Rowand as less potent performers over the home stretch. Are these all cases of a single season mirroring a longer-term trend, or are these recent memories distorting our perception of reality?
In nearly all of these cases, we can trust our memories. There aren't very many hitters who consistently perform better in one half of the season than the other, but each of the 10 depicted in the bar graphs below have been exceptions in recent years, and Holliday, Ortiz, Morneau and Rowand are included among them. All have posted an OPS differential -- that is, the difference between their post-ASB OPS and pre-ASB OPS -- of at least 100 points two times or more over the previous five seasons. That's not exactly metronome-like consistency, but few players have more than one truly lopsided season, even over a five-year span.
Not only have each of these players had multiple lopsided seasons, but nearly all have maintained a strong pattern of either fast starts or fast finishes over their entire careers. Aside from Rowand, each of these hitters could be classified as either a first-half or second-half player. Ortiz's career second-half OPS splits are not much higher than his first half's, but he has hit much better in the second half (OPS differential of at least 0.075) for each of the past four seasons. Holliday, Mark Teixeira , Robinson Cano and Adam LaRoche are not just one-time second-half wonders; they have been turning it on late for much of their careers. Alternatively, Morneau, Brandon Inge , Brad Hawpe and Chase Utley have a long-standing pattern of starting hot, then finishing cold.
These trends present more bad news for owners of Morneau, Utley and Hawpe, as each of them is already dealing with health issues. Then again, perhaps for Utley, the layoff could mean a payoff down the road, as his thumb surgery requires him to rest for the better part of two months. Perhaps this year, a well-rested Utley can come on strong in the season's final weeks.
Sunday, July 11 ... Reality check, please
A month ago, I began a series of Lucky/Unlucky "reality checks" on this blog, revisiting stats for the players I targeted in the weekly Lucky/Unlucky column. After having reviewed my first set of picks for both hitters and pitchers, it's time for a second helping. This time, we'll take a look at the hitters I thought were due for surges and nosedives back in Week 7; after the All-Star break, I'll come back with the pitchers from Week 8's column.
At first glance, it looks like everything has gone as expected for these eight hitters, as every single one has seen his batting average on balls in play (BABIP) move in the expected direction over the last eight weeks. However, both Shin-Soo Choo and Andrew McCutchen have left their early streaking ways far behind, posting BABIP rates well below what they have produced in the past. Just as their earlier rates should not have been taken as signs that both players were taking huge steps forward, nor should we treat their recent slumps as evidence that there is something wrong. This is what the Lucky/Unlucky analyses are all about: spotting trends that may look significant but are nothing more than small-sample blips. Owners can go back to worrying about when Choo will return from his sprained thumb and how McCutchen will be able to score runs hitting in the Pirates' lineup. Both players will eventually settle into levels in between their two extremes.
J.D. Drew has also evened things out, although in a slightly different way. Early in the year, Drew was striking out in 30 percent of his at-bats, but seemed to be having a normal year due to an abnormally-high BABIP. Back then I wrote, "Something has to give here; either Drew starts whiffing less or his Fantasy stats will pay the price." He went with Option A, so his batting average has remained steady, even though his BABIP has regressed. Drew's Fantasy stats haven't improved or worsened much in recent weeks, but now we can have more confidence in him maintaining his current pace.
Of all of the players in this group, Lance Berkman and newly-minted Mariner Justin Smoak have the most perplexing stats. Both first basemen looked primed to take off, but despite their improved BABIPs, neither one is getting as many hits as one might expect. Berkman has been on a power tear recently, hitting five home runs over his last four games, but he still doesn't have his usual value because of a depressed batting average. The modest rise of his BABIP has been neutralized by a career-high 23 percent strikeout rate. While it's good to see Berkman's power back, we may just have to accept that won't ever be the player he was even two seasons ago. Smoak, too, has been victimized by a creeping strikeout rate, and the 21 percent rate he maintained through the end of May was probably too much to expect from the rookie. After all, he whiffed 23 percent of the time in his first season in Triple-A last year. The Mariners and Fantasy owners alike may have to accept on-and-off struggles from Smoak.
Runs Created per 27 Outs (RC/27) -- An estimate of how many
runs a lineup would produce per 27 outs if a particular player
occupied each spot in the order; ex. the RC/27 for Miguel Cabrera
would predict the productivity of a lineup where Cabrera (or his
statistical equal) batted in all nine spots; created by Bill James
Component ERA (ERC) -- An estimate of a what a pitcher's ERA would be if it were based solely on actual pitching performance; created by Bill James
GO/AO -- Ground out-fly out ratio
GB/FB -- Ground ball-fly ball ratio
Batting Average per Balls in Play (BABIP) -- The percentage of balls in play (at bats minus strikeouts and home runs) that are base hits; research by Voros McCracken and others has established that this rate is largely random and has a norm of approximately 30%
Isolated Power -- The difference between slugging percentage and batting average; created by Branch Rickey and Allan Roth
Walk Rate -- Walks / (at bats + walks)
Whiff Rate -- Strikeouts / at bats
Al Melchior will be providing data-centric advice columns Fantasy owners all season. Click here to send him a question. Please put "Melchior" in the subject field.
Who has the best and worst schedules in baseball for Week 18? Chris Towers takes a closer look...
The impending trade deadline may shuffle some rotations, but there are still some two-start...
Heath Cummings is falling in love with cheap rookies on Friday's FanDuel slate.
The Marlins' new acquisition has been on the upswing lately, but will he get a further boost...
The All-Star speedster has been losing Fantasy value, and things are unlikely get better now...
We're running out of reasons for benching flamethrower Lance McCullers. Al Melchior also sizes...