One of the easiest ways to get ahead of your league heading into the draft is to figure out which players should be considerably better or worse than last year. Of course, accurately projecting which players those are is the hard part.
Some people like to bet on upside young players who should develop. Others prefer to look at players who had hot second halves and expect it to carry over. I'm more fond of the former rather than the latter, but both fall short of good old regression. Certain players were simply worse last year because of bad luck, bad circumstances, or uncharacteristic performance. A good portion of those guys will be better in 2015.
When I started looking for regression candidates among pitchers, I started with some obvious factors. BABIP against is one of the most obvious and strand rate is another that's generally accepted. As I'm sure you know, BABIP is the opposing hitters' batting average on balls in play. Pitchers have very little control over this. Strand rate is how many opposing base runners are prevented from scoring. Some believe pitchers have more control over this one, but there's still a lot of luck involved.
The one factor I added this year was HR/FB rate. This is the percentage of fly balls allowed that leave the park. I wasn't 100 percent certain when I started that this was actually as regression prone as the other two, so I took a look at the last three years of data. From 2012-14, 25 pitchers allowed a 13 percent HR/FB rate and pitched at least 100 innings the following season. Of those pitchers, 22 posted a lower rate the following year and 15 posted a better ERA. That's good enough for me and leads us to our first regression candidate.
Rick Porcello is an interesting case in that he checks all the boxes of what I was looking for. His strand rate was 5.4 percent below average. His HR/FB rate was 3.1 percent above average. His BABIP was 36 points higher than league average. What does it all mean for a pitcher with a 4.92 ERA in 2015 and a career 4.39 mark?
If you normalize Porcello's BABIP, he would have given up 19 less hits over the season. That's not insignificant. If you normalize his HR/FB rate, that's six less home runs. Normalizing his strand rate would also have a significant effect. The problem with Porcello is he's never really been normal in that regard.
His career strand rate of 69.6 percent suggests he may not be a very good pitcher with runners on base. Further examination reveals that in his career hitters have an .806 OPS off him with runners on base compared to a .719 mark with no one on. Those numbers have improved a little over his career, but I don't think we can expect him to be completely average when it comes to preventing base runners from scoring.
It's also plausible that the Red Sox defense that contributed to his atrocious .332 BABIP against won't be much better in 2016. Eduardo Rodriguez was the only Sox pitcher in 2015 to make 18 or more starts and have an ERA within a half point of his FIP. In other words, with Porcello's ability with runners on, the Red Sox defense and ballpark, he's going to be prone to looking like a regression candidate most years.
With all of that said, I'd expect a much better performance from Porcello in 2016 and not just because of how good he was down the stretch of last year. Although, as Al Melchior tweeted, there is reason to hope that his turnaround was skill based and not just random.
In 2016, I would expect 190-200 innings with an ERA below 4 and a WHIP in the 1.25 range. I would expect his strikeout rate to dip to around 7 K/9 and his wins to rebound to the 12-15 range. If you're looking for a comp from 2015, I'd think somewhere between Jordan Zimmermann and Julio Teheran. If he reaches those marks, he's likely a top 40 pitcher, but he won't be drafted like it because of limited upside.
If Porcello is the poster boy for positive regression, Marco Estrada is the other side of that coin. Estrada had a career year, throwing a career best 181 innings while winning 13 games and posting a 3.13 ERA. Unfortunately, you should believe none of it.
Estrada will always be a low BABIP guy with his fly ball tendencies, but the .216 mark he posted last year was the lowest by a qualified starter since 1988. There is no chance he's repeating that, or likely getting close. Maybe more importantly, it's very difficult to see him matching that HR/FB rate in Toronto.
Estrada is a guy who was a streamable option at times last year, but with the expected regression he won't be worth a roster spot. He had a 4.40 FIP in 2014 and his xFIP was nearly a half run worse. His strikeout rate fell to a career-worst 6.5 batters per 9.
The other thing that doesn't get considered is that Estrada's low BABIP led to his high(er) innings total, as did his 79.2 percent strand rate. I wouldn't be surprised to see Estrada fall to around 150 innings if he's even able to maintain his spot in the rotation.
King Felix was more peasant than royalty in the second half, posting a 4.48 ERA and allowing a .785 OPS after the All Star Break. That, along with his 2,200 career innings, has many people preparing for a further decline in 2016. I am not one of those people. Hernandez doesn't fit all of the markers for a bounce back season, as his BABIP and strand rate were both better than league average. However, the one thing that was out of line was way out of line.
Hernandez gave up more than a HR per nine innings for the first time since 2006 largely because of a 15.3 percent HR/FB rate. As you might expect, his HR rate ballooned in the second half, as did his hard contact rate.
The problem with second half stats like these is they just aren't as predictive as full season stats, or even three year averages. Sure, there are absolutely instances where pitchers decline in the second half after a ton of innings and are never the same. I just don't feel like that's near as likely with Hernandez because his home run rate regressing and him once again being one of the best pitchers in baseball in 2016.
I would expect his HR/FB rate to be around 10 percent next year and his ERA to fall back into the lower threes. It may be a stretch to see Hernandez as a top-five pitcher, but that's more about everyone else than him. He should be a quality ace in the 2016.
It might be hard to think about a pitcher with 274 Ks as a regression candidate, but everything about Sale's 2015 was a bit wonky. Some people might look at Sale's 2014 and say that 2015 was his regression. He was coming off of a year where he outperformed his FIP by .40 largely due to an 81.5 percent strand rate and a 7.5 percent HR/FB ratio. Both of those number normalized, but it's his BABIP in 2015 that really caused him some troubles.
Sale had pitched more than 600 innings with a BABIP around .290 heading into 2015 and then everything fell apart. When you look at the "why," it's not exactly clear. Sale's ground ball and fly ball numbers are similar to his career norms, while he saw a very slight bump in line drive rate. He induced more soft contact than he had since 2011. The one area where Sale saw a major disturbance was infield hits, where he gave up a career high.
All of this adds up to Sale posting his lowest xFIP as a starter (2.60) in the same year that he posted his highest ERA (3.41). There should be little doubt that if Sale pitches like he did last year he's more likely to win the Cy Young than end up with with an ERA pushing 3.5.
The Indians: Specifically, Carlos Carrasco was fantastic last year, posting a 2.66 xFIP that was fourth in all of baseball. He and Corey Kluber should post better counting numbers than they did in 2015 and there's a chance Danny Salazar joins them, challenging the Mets for the best rotation in Fantasy baseball.
Hector Santiago:Santiago is on the opposite side, riding a .252 BABIP and a strand rate near 80 percent to a sub-4 ERA. I'm not sure he'll be as bad as I expect Marco Estrada to be, but I can't imagine Santiago approaching last year's numbers unless he does something very different.