Over at Owner’s Edge by Fanball.com, I wrote a piece about strikeout-to-walk ratios and how they relate to a pitcher’s ERA. I looked at the past two seasons to see how strong a correlation there was between K/BB ratio and ERA. If a strong enough relation existed, I wanted to use that information to see which pitchers stood out as buy-low or sell-high targets based on their K/BB and ERA thus far.
I was happy with the results in terms of the players identified, but some of my statistical conclusions left me a little uneasy, so I went back to the drawing board a bit. This time around, I went five years back and grabbed every qualifying ERA. This data set presented 393 samples with ERAs ranging from 2.27 to 6.47 and K/BB ratios from 8.3 to 1.1. I was comfortable with the depth of this set. In the original piece I used a 2.0 K/BB threshold, but given that 2.0 is the baseline that we generally look for in the fantasy baseball world, I thought it was a bit low for the purposes of what I’m looking to get out of the data.
I bumped it up to 2.5. At 2.0, it’s essentially a coinflip which isn’t surprising considering that it is hardly an elite mark. In fact there were 248 data points of 2.0 or better and it was a 60%/40% split of ERAs +/- 4.00. The worst ERA in the entire study, Eric Milton‘s disgusting 6.47 offering from 2005, actually topped the 2.0 threshold thanks to his sparkling 2.5 BB/9 rate. Moving to 2.5 cuts out the bottom 31 ERAs in the study and 46 of the bottom 50.
Here are the results with the deeper data pool and higher K/BB threshold:
The above charts show that a 2.5+ K/BB ratio is three times more likely to yield a sub-4.00 ERA than not. Within a given season, there will be a group of pitchers whose skills should have netted them a better ERA, but poor defense or simply bad luck plagued them and left their skills unrewarded. The average was eight such starters per season. Given that recent trends have between 80 and 90 ERA qualifiers, it is about 9-10% of starting pitchers that get the short of the stick regardless of skills.
Here are some of the best buy-low opportunities who are also at risk of being part of this year’s batch of unlucky pitchers:
I don’t think you can really buy low on Justin Verlander given how unbelievably hot he has been lately striking out 9, 11, 11 and 13 in his past four starts. However all four of his AL Central counterparts on the list should come at legitimate discounts. I’d target Minnesota’s Kevin Slowey above all. His 0.9 BB/9 is amazing and while it might not hold 100%, he maintained a 1.3 in 160 IP last year so it’s unlikely to jump up too much. Rich Harden, Jon Lester and Jake Peavy won’t be bargain bin pick ups because of their gaudy strikeout totals (and because Harden and Peavy don’t have outrageous ERAs), but if you can get them at any discount, I’d recommend doing so immediately.
The at-risk group has it’s fair share of star power on it, too:
Three-fifths of the New York Yankees rotation is overachieving so far while the remaining two are getting obliterated (A.J. Burnett-5.36, Phil Hughes-7.56). And that over achievement has earned a record just one game above .500. Any regression could be very damaging and quickly push the Yankees to fourth in their division. But I actually expect Sabathia and Chamberlain to get better as we close out May and head into June. Sabathia will up his K-rate while Chamberlain will trim his BB-rate and continue to strikeout a batter per innings.
Jair Jurrjens and Brian Bannister are major red flags. We know what the bottom looks like for Bannister (1.9 K/BB in 183 IP last year led to 5.76 ERA), but Jurrjens flirted with the 2.0 threshold last year and ended up having a pretty successful year. Of course he did go for a sub-3.00 ERA in the first half and then regress heavily with a 4.49 in the second half. At least in 2008 he was straddling the limit with a 1.9 first half and 2.1 second half. I’d sell him instantly. And I’d have never bought Bannister so if you do have him, cash in that lottery ticket as soon as you can because it has an expiration date.
The names on this list that I’m least worried about are: Chad Billingsley, Max Scherzer and Matt Garza because of their strong K-rates of 9.3, 8.4 and 7.9, respectively. Yes Mitchell Boggs is toting an 8.1 K/9, but the last time he reached a mark that high was his final year at the University of Georgia in 2005 so I’m not buying it in the least with just 22 innings of work. As I mentioned earlier, I do think Chamberlain will turn it around, but there is still some risk because he has a nearly 10.0 H/9 rate to go with the gaudy BB-rate. There are concerns that he is trying to save himself to go six or seven innings and it’s causing him to be very hittable in the rare instances that he is actually in zone.
Baseball By Paul 