- Matthew Berry, Fantasy
- 0 Shares
It's all a crock.
And anyone who tells you different is trying to sell you something.
Whether it's a fantasy baseball sleeper or a political agenda or just why Olivia Wilde is hotter than Joanna Krupa, it's all bunk and hooey, especially since Olivia went brunette. Look, you wanna go back to Olivia in blond hair making out with Mischa Barton on "The O.C."? Then we can talk. But right now, it's The Krupa and it's not close.
The point, and I'm pretty sure there is one, is that every single day, on blogs, cable networks, Web sites, podcasts, Twitter, talk radio and more -- in coffee shops, bars, dorms, office break rooms and everywhere else -- people spout numbers and facts, research and statistics, poll numbers and pie charts to convince you that their point, their analysis, their opinion is the correct one.
More specific to our little fantasy world, you have already read and heard many stats, numbers and research. Tons of analysis and opinion, sleepers and busts, rankings defended and mock drafts explained, all delivered with certainty because of the facts supporting it.
So not only is there a point, there's only one point and it's this: It's all a crock.
Capital Ka, capital Rock.
The simple truth is that there is no way to get a fully accurate statistical projection of a player. Glancing at the stats section here on ESPN.com, along with Baseball-Reference.com, FanGraphs.com and the internal databases we have access to from Stats LLC and the Elias Sports Bureau, I count well over 100 statistical categories for every player.
From the basic stuff like ERA and home runs to more advanced things like GB/FB, K/9, plate appearances/strikeout and BABIP to really hard-core things like XBT% (extra bases taken percentage), O-swing % (percentage of pitches outside the strike zone a guy swings at) and wRAA (runs above average based on weighted on-base average).
More than a hundred of those stats, all with insight into how a player has performed and, by extrapolation and interpretation, might perform in the future.
And that's before you get into things like park factors, where the player is hitting in the lineup, his teammates, the manager's approach, how someone is scoring the game (we've all seen the clear error scored as a "hit" for the hometown hero and vice versa), weather and, of course, opportunity and playing time.
So it's a lot of information, some of it (such as how a play is scored or weather) almost impossible to predict or truly quantify.
On top of which, to project based on past performance assumes (always dangerous) that the player is 100 percent healthy (many hide or are unaware of injuries) and has the exact same approach as before (no adjustments to batting stance or trying a new pitch, gripping the ball slightly differently, trying to hit more line drives than home runs, etc.).
And even if you somehow know all of that -- I mean you've done nothing but study since November and you're a fantasy baseball Rain Man -- there is no stat for what I call "The Glug Factor."
Whether it's a drinking problem or a contract negotiation or the wife found out about the girlfriend or, my favorite, he is trying a new religion with Madonna, there's all sorts of "glug glug" stuff that never shows up in a box score.
Remember 2008, when Lance Berkman stole 18 bases? He had never (and still hasn't) stolen double digits in any other year of his career. In fact, he had stolen only 14 bases in the previous three years combined!
That year, I went to the All-Star Game and got some time with Lance. He couldn't have been nicer, so I asked him, point-blank, where the speed had come from. I'm paraphrasing here, but the gist of it was this:
"I've always had the speed to run; I've just never needed to before. But this year, I realized pretty quickly we had some challenges offensively and we would need to generate more offense, so I decided to start running."
The emphasis in that last sentence is mine.
But let's pretend, for a second, that you have such statistical knowledge that Bill James calls on you for advice. That you know all the gossip, all the variable factors, what's going through every player's mind, and how each of them will handle the rigors of the season physically and emotionally. What do you do with it?
Because fantasy is much less about projecting stats and a lot more about the value you (and your league mates) place on those stats. How many positive stats can you get for as little as possible, and how well does any one player mesh with the rest of your team?
I mean, 30 home runs is better than 20 home runs, right? Not if the 30 home runs costs a first-round pick and the 20 costs an 18th-round pick. Because you can't look at the 30 and 20 home runs in a vacuum. You have to compare the 30 home runs with what else you would have gotten in the first, the 20 home runs with what else you might have gotten in the 18th and every other player you roster. And what did those players hit for average? If the 30 home runs gets you three points in the standings but costs you two in batting average, was that tally really more valuable than the 20 that got you two points in the standings to go along with one in batting average? No stat, no projection, no analysis exists by itself. And the universe it exists in changes with every draft pick, every player purchased, and every player dropped or picked up during the season.
Hold those thoughts while I introduce another one.
The good folks over at Wikipedia describe selection bias this way: "Selection bias is a distortion of evidence or data that arises from the way that the data are collected. It is sometimes referred to as the selection effect. The term selection bias most often refers to the distortion of a statistical analysis, due to the method of collecting samples. If the selection bias is not taken into account, then any conclusions drawn may be wrong."
Emphasis at the end there, once again, is all mine.
Let's look at a well-known player I have yet to see being taken in the second or even the third round this year:
From 2008 to '09, Player A's on-base percentage dropped by almost 20 points, his batting average dropped 30 points and, despite getting 59 more plate appearances, he had one fewer hit and had his highest strikeout total in the past four years. Yes, you can point to fact that he tied a career high in stolen bases, but it was the same amount as he had the previous three years combined and, as we always say, you never pay for a career year. As this veteran player heads into his 10th MLB season, we might be seeing the beginning of the end.
So obviously, whatever you do, avoid Albert Pujols.
It's an extreme example of selection bias, but hopefully it proves the point. Anyone can argue for or against any player just by choosing which stat(s) to highlight.
Hence, the crock. It's all a crock, kids.
Because it has to be. When I am on "SportsCenter" or "Baseball Tonight" or even doing the Fantasy Focus podcast, I have a finite amount of time. I can't speak/type/gesticulate wildly all day any more than you can listen/read/tolerate me. It's the same for everyone. We all have time limitations, so we have to make choices.
I study all the stats, do the research and talk to as many folks as I can, then I choose which stats I want to show/discuss/butcher. If I like the guy, it's positive stats; if I don't, I highlight the negative. I've done it below in what I've chosen to show you. And all analysts and opinion givers -- in fantasy sports or anywhere else -- do the same thing, whether they admit it or not.
Your job is to see whom you trust, whom you don't, whom you agree with and who you think is a moron. Take it all in and make your own call because you're the one who has to live with it.
Everything that follows is absolutely true. They all are facts. Some are about baseball players and teams; some are just for fun; and not one of them tells the whole story. There are 50 in all. What you do with them is up to you.
1. Only five pitchers in baseball last year had more quality starts than Randy Wolf.
2. His home ERA (pitching in Dodger Stadium) was 3.63. His road ERA? 2.78.
3. According to our Park Factors, Dodger Stadium was 28th in terms of runs allowed. 27th? Miller Park in Milwaukee.
4. Of course, Miller Park was also top 10 in home runs allowed, so let's not go nuts.
5. I can't believe I just spent four facts on Randy Wolf. The article gets better. Probably.
6. Post All-Star break, only seven pitchers had more strikeouts than Jonathan Sanchez.
7. Sanchez had a second-half ERA of 3.83.
9. Last season, David Wright had 20 more strikeouts despite playing 16 fewer games than the year before.
10. It was the third year in a row his strikeouts have increased. Sounds bad, doesn't it? Not nearly as much as I just made it sound because his contact rate (basically how often he doesn't strike out) was an identical 81 percent for three consecutive seasons until he fell off a cliff last year. I was just using selection bias to tell you I don't trust Wright. I just happen to have a conscience about it. Don't worry, won't happen again.
11. Playing winter ball in the Dominican Republic this offseason, Francisco Liriano had 47 strikeouts and allowed five walks in 37 innings pitched.
12. When it comes to Francisco Liriano, I might as well be Charlie Brown and Liriano is Lucy with the ball. I know better, but I can't help myself. When he has killed me yet again, I will call an emotional news conference in September in which I will be awkward and stiff and hug Nate Ravitz afterward.
13. As R.J. Anderson pointed out in his excellent Insider piece, of the 72 home runs that Joe Mauer has hit in his big league career, 46 percent have been hit to left field, including 58 percent last year. Each of the "left-field home runs" went farther than 350 feet, according to HitTrackerOnline.com
14. In the Twins' new home stadium, the left-field wall is 328 feet, or 15 feet closer than it was at the Metrodome.
15. Last season, 17 of Johnny Damon's 24 home runs were at Yankee Stadium.
16. Yankee Stadium was first in home runs allowed last season. Comerica Park was 18th.
17. Last season, Ben Zobrist hit 27 home runs in 501 at-bats, or one home run for every 18.6 at-bats. The year before, he hit 12 home runs in just 198 at-bats, or one home run every 16.5 at-bats. Those who think he had a power explosion last year are wrong. He actually had a small regression!
18. Based on AB/HR from last year, who could be this year's Ben Zobrist if he gets playing time? To give some comparison, Albert Pujols has an AB/HR of 12.09. Small sample sizes, but Randy Ruiz had an 11.5.
19. Kyle Blanks had a 14.5.
20. Ryan Raburn had a 16.31.
21. On Twitter, the number of followers of the cast of "Jersey Shore":
22. I really liked Ronnie and Sammi. Apparently, I'm nearly alone.
23. Last season, Carl Crawford had 44 stolen bases before the All-Star break. He had just 16 after the break despite getting on base at almost the exact same rate.
24. Over his career, when he plays third base, Chone Figgins is a .298 hitter. When he plays second base, he's a .268 hitter.
25. More Figgins: Some of this has to do with personnel, of course, but last year, the Angels were third overall in MLB in stolen bases. The Mariners? 16th.
26. According to katron.org, seven of Curtis Granderson's doubles or fly outs at Comerica Park last year would have been home runs at Yankee Stadium.
27. Conversely, eight doubles or fly outs off of Edwin Jackson at Comerica last year would have been home runs at Chase Field.
29. More from Gregory, this time about Chris Davis: For most of last season, Davis served more as a fantasy punch line than a sleeper, striking out once every 2.79 plate appearances, worst in the majors. But Davis turned a corner in his plate discipline in the season's final month, batting .318 with an OPS approaching .900. And most importantly, he improved that strikeout rate to once every 4.2 plate appearances.
30. If you would like an Ocean View Balcony Suite on Deck 7 for the Rick Springfield Cruise, it will cost you $2,849.
31. In 18 games (15 as a starter) away from Coors Field last year, Jason Hammel had a 3.13 ERA.
32. Hammel's total ERA was 4.33, but his FIP (fielding independent pitching -- basically, his ERA based only on the things a pitcher can control, such as home runs, walks and strikeouts) was 3.67, good for top-20 in the majors last season.
33. Last year, no pitcher had a worse run support average than Aaron Harang. His strikeout-to-walk rate, however, was 18th-best.
35. If I have to use the restroom at work, I always go to a different floor than the one I work on.
36. Shout-out to Jason Vida of ESPN Stats & Information for this one, as a lot has been made about Troy Tulowitzki's speed and whether it will continue this year: In the Rockies' first 46 games of the 2009 season (when Clint Hurdle was managing the team), Colorado had 45 stolen base attempts (11th-best in MLB). In that span, Tulo played in 43 games, stealing four bases and being caught stealing three times (0.16 SB attempts/game).
After Jim Tracy took over on May 29, the Rockies had 116 stolen base attempts (76 SB, 40 CS) in their final 116 games of the season. That rate of one stolen base attempt per game ranked seventh in the majors in that span. With Tracy as manager, Tulo played in 108 games, stealing 16 bases and being caught stealing eight times (0.22 SB attempts/game). So although the Rockies didn't run at a much higher rate once Tracy took over, Tulo's SB attempts per game clearly increased with Tracy.
37. I don't know if it's because I used to write scripted television and movies for such a long time, but I find that I rarely watch scripted shows these days. If it's not sports or "SportsCenter," "Baseball Tonight," etc., I will watch "The Daily Show." That's my only must-watch. I also really like "The Colbert Report," and I admit to loving the mindless reality of shows like "American Idol," "Survivor" and "Celebrity Apprentice."
38. As Jay Jaffe noted in the TMI blog (a fantastic must-read every day, incidentally) on ESPN.com, Clayton Kershaw added a slider to his repertoire in early June of last year. From June 1 on, Kershaw had a 2.15 ERA and a 120/56 K/BB ratio in 109 innings pitched. Hitters hit .198 against him. Of pitchers that pitched at least 100 innings, only Felix Hernandez (2.07) had a lower ERA. Kershaw's 120 strikeouts, however, were tied for just 25th. That's probably because
39. Of 31 games he pitched last year, Clayton Kershaw pitched more than seven innings only once. He pitched more than six innings only seven times.
40. What do Alex Rodriguez, Hanley Ramirez, Matt Holliday and David Wright have in common? Over the past three years, they have stolen at least 50 total bases, hit .287 or better, and have more home runs and RBIs than Torii Hunter's 71/275. They are the only ones who can claim that.
41. Torii Hunter is currently going 84th overall and 23rd among outfielders in ESPN standard drafts.
42. I was thinking of making this 100 stats, but my intro was so long and I didn't want to reuse stats from my "Love/Hate." So you get 50. Well, 40 real ones. It's still almost 4,000 words. I don't like writing this long generally, but then I also don't like cutting.
43. Last season, of the four catchers the Yankees used, Jose Molina had the lowest catcher ERA (3.31). Jorge Posada had the highest (5.02). In fact, of catchers in major league baseball who caught at least 350 innings last year, only Rob Johnson of Seattle (3.22) had a lower ERA than Molina.
44. Jose Molina is now on the Blue Jays.
45. My favorite non-ESPN iPhone app is Shazam. Thing is awesome. I just used it to identify and then purchase "Tarzan Boy" by Baltimora. Don't judge.
46. After coming over to the Mets last year, Jeff Francoeur had 308 plate appearances. He hit .311 with 10 home runs and 41 RBIs.
47. According to this very helpful chart, the average number of home runs needed to win that category in an ESPN standard 5x5 league last season was 326. You needed 224 stolen bases, or 102 fewer, to earn the same number of points in steals. In other words, you needed more power than speed. Also, you need to have a team batting average of .279 to finish fifth. I made this point in Love/Hate, but I think it's important, so I'm making it again.
48. Last season, there were 25 outfielders who stole at least 20 bases. All of them hit at least .279. And there were only 15 outfielders who hit at least 20 home runs and also hit at least .279.
49. In 2008, Nick Markakis had 595 at-bats. He walked 99 times. In 2009, he had 642 at-bats, or 47 more. He walked only 56 times. So which is the real Markakis, the one with walk rates of .08, .09 and .08 in Years 1, 2 and 4 of his career, or the supposed "breakout" Markakis of 2008 who had the .14 rate?
50. I'm Syracuse Orange, Class of '92. I picked Kansas to win it all in my Tournament Challenge Bracket. My thinking: I do this all the time. The homer in me wanted to pick Syracuse, but I didn't want to jinx the Orange. There is no statistical basis to think that this will have any effect, but I don't care. Stats are overrated.
Matthew Berry -- The Talented Mr. Roto -- also bought "Naughty Naughty" by John Parr. Seriously, don't judge. He is also the creator of RotoPass.com, a Web site that combines a bunch of well-known fantasy sites, including ESPN Insider, for one low price. Use promo code ESPN for 10 percent off. Cyberstalk the TMR | Be his Cyberfriend
Matthew Berry rants on the impossible task that is predicting future results based on past-performance measurements, and gives 50 facts that may or may not have relevance in 2010.