OK. I’ll admit it. I’m addicted. I must be. It’s the only reasonable explanation. I must be addicted to pitching. And I know I’m not alone. The evidence? How about Barry Zito for seven years at $126 million? A four-year, $40 million contract for Ted Lilly? Five years, and $55 million, in a deal for Gil Meche? Jason Marquis for $20 million? Adam Eaton for $24 million? Should I go on? I may be addicted, but, clearly, I’m not the only one in need of a support session at Pitchers Anonymous.
What makes Major League general managers chase pitching? Did you ever hear the phrase, ‘pitching wins pennants’? Just how important are pitchers when it comes to a baseball club’s pursuit of that most valuable of commodities, wins? Consider this bit of simple logic … the fewer guys a pitcher puts on base, the fewer runs he allows to score. The fewer runs allowed, the better chance of piling up wins. This all sounds incredibly obvious, but is it supported by fact? Let’s look at a little bit of data and see what we can find.
I gathered up the 2006 team pitching statistics for all 30 Major League ball clubs. So. The fewer guys a pitcher puts on base, the fewer runs he allows to score … or … the lower the WHIP, the lower the ERA? What does the data say? I ran some simple calculations in Excel to obtain a correlation of team WHIP and ERA. A correlation coefficient will identify whether or not there is a meaningful relationship between two sets of data. The coefficient will range from -1.0 to 1.0. If it is between 0.0 and 1.0, we would say we have a positive relationship of some magnitude. Meaning, both statistics move in the same direction. In this case, as WHIP goes down, so does ERA. If it is between 0.0 and -1.0, we would say we have a negative relationship of some strength. Meaning, the statistics move in the opposite direction. In this case, as WHIP goes down, ERA goes up.
The 2006 team statistics yielded a correlation of 0.87 for WHIP and ERA. As we thought, an incredibly obvious result, but a pretty strong positive relationship, nonetheless. It’s safe to say, as WHIP goes down, so does ERA. But, also, we could say as WHIP goes up, so does ERA. Interesting. Two closely related categories. When we dig further, we find a correlation of ERA to Wins of -0.75. And a correlation of WHIP to Wins of -0.81. Two additional reasonably strong correlations, this time negative, providing yet another expected result. As ERA and WHIP go down, Wins go up, or, alternatively, as ERA and WHIP go up, wins go down. Clearly, in total, we have three closely related statistics. Could they also, perhaps, be closely related fantasy categories as well? Knowing the answer to this question would bring value to us fantasy GMs.
Let me ask you this. How many of you go into a draft loaded with good intentions. You have the “perfect plan” to bulk up on wall-banging sluggers. ‘No more superstar pitching staffs for me’, you claim. Yet, the annual ritual begins. The selection frenzy starts. The dust settles, and a list of pitchers that would make the 90’s-vintage Atlanta Braves drool comes to rest on your team.
In the aftermath, what makes us hang our heads in shame? We’ve all heard it. We all believe it. Other than a few exceptions, like Johan Santana or Chris Carpenter. Or maybe … a BIG maybe … hurlers such as Roy Oswalt, Brandon Webb, Jake Peavy, Carlos Zambrano or Roy Halladay, pitchers are unreliable. Their performances bounce around too much from year to year. They come with too much injury risk. You can’t count on them. You can’t trust them. Yet, we find ourselves sadly bleeding auction dollars and draft picks in futile attempts to corner the arms market. Are we driving our teams to ruin?
Maybe not. Why? How do the aforementioned “safe” pitchers fare in the same correlational analysis we performed earlier? I examined the career stats of our seven “aces”. With regard to the relationship between WHIP and ERA, I think it is fair to say that the individual data supports our earlier observations. The correlation coefficients for these “front-line” starters ranged from a reasonably strong 0.75 to an astounding 0.99 suggesting that, as WHIP goes down, so does ERA … or vice versa. To satisfy anyone’s curiosity, which pitcher had an almost perfect correlation between WHIP and ERA? Carlos Zambrano wields the lofty 0.99 relationship.
Unfortunately, however, things become a bit clouded. Our next step was to compare ERA and WHIP to Wins. On the average, these seven studs carry about a -0.49 relationship between ERA and WHIP to Wins. This relationship is in the direction that we would expect, in that as ERA and WHIP go down, Wins go up, but it is not very strong. What happened? Well, five of our pitchers followed pattern. Looking at the relationship between WHIP and Wins, for Santana (-0.92), Carpenter (-0.91), Zambrano (-0.82), Peavy (-0.75) and Webb (-0.74), we receive the expected results. Halladay and Oswalt, however, were the exceptions. The correlation of Roy Halladay’s WHIP to Wins was -0.26. The right direction but of negligible strength. The gross aberration was Roy Oswalt. The relationship of his WHIP as correlated to Wins was 0.70. A positive correlation! Meaning, as his WHIP goes up, his Wins go up as well! How can this be? Looking back on his career stats, his two highest WHIPs were logged in 2004 and 2005 at 1.24 and 1.21, respectively. These were also his best seasons for wins with 20 in each of those years. In the end, even though in five of seven cases the expected relationship was supported, I would judge this part of the analysis to be somewhat inconclusive.
What does this all tell us? First, I think deeper study is probably merited. There does seem to be an interrelationship between the three categories, but I think it is also fair to say that, to gain an even better understanding of the meaningfulness of the individual pitcher data, we should examine a player population broader than our seven “aces”.
Further, my contention has always been that the pitching categories in fantasy are so interrelated that it was unwise to deprioritize pitching to too large of a degree at the draft and during the season. I’ve always believed that this leads to a GUARANTEE of poor or mediocre performance in all three of the aforementioned pitching categories. A fantasy team might be able to survive a hit in one category, but can it overcome poor performance in three? Probably not. Nevertheless, I think it would be profitable to try to prove this out by performing these same correlational studies on actual year-end fantasy league results. In fact, we could add another facet in examining whether or not superior performance in the pitching categories actually correlates with winning a league.
For me, I think it’s a choice of which risk you want to take. Do you want to avoid the unreliability risk inherent in pitchers and draft them in the middle to later rounds? If you do, you may set yourself up for obtaining a stable of hurlers who hurt your WHIP and, subsequently, your ERA and Wins. Of course, one answer to this is to draft a healthy complement of middle relievers. But, then, will you suffer in Wins and Strikeouts?
Alternatively, do you want to roll the dice with regard to individual injury and performance risk and draft pitchers relatively early in the hopes of racking up points in three categories that clearly have some level of interdependency? The thought being that superior performance in these three categories, coupled with mediocre to good results in the rest, will ensure that you stay in the pennant race until the end. With regard to my “addiction”, I usually choose this latter course. Sadly, however, I always seem to go just a little bit overboard.
I recently finished what has become my first official draft of the 2007 season. The results? I couldn’t stop myself. Chris Carpenter, Brandon Webb, Aaron Harang, Matt Cain, A.J. Burnett, Scott Olsen, Javier Vazquez. Oh, no! It has happened again! “Hello? Pitchers Anonymous? I need your help …”
I welcome your comments @bob@fantasybaseballmafia.com
by Bob Sikon