When a reliever converts to a starter, though each pitcher is different, there are several things to look for:
1. He will lose about 1-3 MPH on his fastball
2. His strikeout rate will decline
3. Walk rate basically stays the same – might improve very slightly
4. The more pitches he has and the less he relies on his fastball, the better-equipped he will be – developing an effective change-up is key
5. To be a successful starter, he has to be effective against both left-handed hitters and right-handed hitters
6. ERA will go up (or at least the peripheral stats that predict ERA)
Let’s break down each converted starter, comparing their 2011 stats to recent converted starters (from relievers) RHP Alexi Ogando and LHP CJ Wilson to see how their conversions to starting went in the first year (2011 for Ogando, 2010 for Wilson).
Bard = 97.3 MPH
Feliz = 96.3 MPH
Sale = 95.3 MPH
CJ Wilson as RP = 93.4 MPH
CJ Wilson as SP = 90.5 MPH
Alexi Ogando as RP = 96.3 MPH
Alexi Ogando as SP = 95.1 MPH
Historical data from Jeremy Greenhouse suggests the average reliever converting to starter will lose on average 0.7 MPH, however both Wilson and Ogando lost significantly more than that (especially Wilson). Brett Myers, Hong-Chih Kuo, Joba Chamberlain, and Phil Hughes all lost around 2-3 MPH as well, which leads me to believe you should expect at least a little over 1 MPH to be lost in the conversion. It could also have to do with the fact that Wilson went from throwing 74 innings as a reliever in 2009 to throwing 204 innings as a starter in 2010 and he wore down physically in his first season. This would seem to make sense as his velocity regained 0.5 MPH in 2011. Ogando, meanwhile, went from throwing 42 innings in 2010 to throwing 169 innings in 2011… still a huge innings jump, but not as big as Wilson’s. Wilson also didn’t have the initial velocity that Ogando, Sale, Feliz, and Bard had. It could depend on how many innings the 3 of them pitch in 2011 because they will likely wear down towards the end of the season and could see a steeper velocity drop. The bottom line is all 3 possess great velocity on their fastballs and shouldn’t experience any problems related to their fastballs not having enough velocity unless they experience a Joba-like decline. Feliz and especially Bard should both be near the top of starting pitcher fastball velocity in 2012 (with Sale up there too).
Sale = 10 K/9
Bard = 9.1 K/9
Feliz = 7.8 K/9 (9.2 K/9 in 2010)
CJ Wilson as RP = 10.3 K/9 (8 K/9 in 2008)
CJ Wilson as SP = 7.5 K/9 (8.3 K/9 in 2011)
Alexi Ogando as RP = 8.4 K/9
Alexi Ogando as SP = 6.7 K/9
Feliz had the lowest strikeout rate of the 3 by far, but in 2010 he was right with them. He had a skills breakdown in 2011 that saw him walk more and strikeout fewer batters. If he doesn’t recover from that, he will definitely struggle as a starter. If he does recover from that, then he could easily be a very good starter. We know the strikeout rates of all 3 will decline, but what we don’t know is how much they will decline. Looking at the research done by Steve Treder in 2006, historically it looks like we should expect a dip of about 0.6 K/9 – 0.8 K/9. However, once again, looking at our 2 recent examples shows we might have to expect a bigger dip than that. Ogando saw a 1.7 K/9 decline from 2010 to 2011. CJ Wilson looks like he might have been a bit lucky to get 10.3 K/9 as a reliever in 2009 as his previous year’s strikeout rate was only 8 K/9. Likewise, in his first season as a starter, his strikeout rate was only 7.5 K/9 whereas the next season it was 8.3 K/9… so I think we’re seeing a skills dip of around 1.5 K/9 – 2 K/9, not the huge 2.8 K/9 chasm that it appears to be. I’d expect Sale to be around 8.7 K/9, Bard to be around 8 K/9, and Feliz also around 8 K/9 if he turns things around from last season – otherwise I don’t see him lasting as a starter.
Bard = 3 BB/9
Sale = 3.4 BB/9
Feliz = 4.3 BB/9 (2.3 BB/9 in 2010)
CJ Wilson as RP = 3.9 BB/9
CJ Wilson as SP = 4.1 BB/9 (3 BB/9 in 2011)
Alexi Ogando as RP = 3.5 BB/9
Alexi Ogando as SP = 2.3 BB/9
Feliz also had by far the worst walk rate last season – 2 BB/9 higher than 2010! Obviously he must improve upon his control to be successful in 2012. Bard and Sale are both within an acceptable range for control, especially Bard. Bard’s control has improved in each of the last 2 seasons, with his walk rate going from 4 BB/9 to 3.6 BB/9 to 3 BB/9. Even if Sale’s walk rate remains where it is, it will be offset somewhat by his higher strikeout rate. You can see that Wilson’s control actually got a little bit worse when he converted to a starter, but he is the exception, not the rule. Ogando on the other hand improved his control significantly. Part of that could just be gaining experience pitching, as prior to 2010 Ogando was an outfielder! So he not only converted from outfielder to relief pitcher, he then converted from relief pitcher to starting pitcher (and now he’s back to relieving again to start 2012). Despite these two, research has shown that the walk rate basically stays the same on average, so plan for that. If Feliz can show the control he showed in 2010, he could be dominant.
Pitches and Usage
Sale = Fastball (51.9%) Slider (36.4%) Change (11.7%)
Feliz = Fastball (79.8%) Curve (11.1%) Slider (4.8%) Change (4.3%)
Bard = Fastball (67.9%) Slider (24.7%) Change (7.4%)
CJ Wilson as RP = Fastball (70%) Slider (18.5%) Cutter (5.6%) Change (5.3%)
CJ Wilson as SP = Fastball (49.2%) Cutter (18.6%) Slider (12.1%) Change (11.7%) Curve (8.5%)
Alexi Ogando as RP = Fastball (64.3%) Slider (30.9%) Change (4.8%)
Alexi Ogando as SP = Fastball (67.4%) Slider (27.7%) Change (4.8%)
Feliz is the only one of the 3 to throw 4 different pitches, but considering he relies heavily upon his fastball and barely uses his slider and change-up I’m not sure that gives him a leg-up on the other two. Sale looks like he’ll be able to best handle the conversion to starter as he has 3 good pitches he’s confident in and doesn’t rely on his fastball too much. He has a killer slider that helps him rack up the strikeouts. Bard also has a great slider and an even better fastball than Sale, but he’ll have to use his fastball less and change-up more to be successful as a starter. Historically, Greenhouse found that relievers use their fastball about 3% more often as relievers than starters. Once again, our two recent examples don’t agree with these findings at all. Wilson threw his 4-seam fastball about 20% less as a starter, however he threw his cut fastball 13% more often, so he basically just changed one fastball for another. He also added a curveball to his repertoire, threw his change-up more and slider less. He had to make adjustments with only a 90-91 MPH fastball as a starter. On the other side, Ogando actually used his fastball 3% MORE as a starter as it still sat at 95 MPH and he had a lot of confidence in it and not much confidence in his change-up. It doesn’t appear that Sale or Bard have any problems throwing a change-up and can probably learn to throw it more as a starter. Feliz on the other hand doesn’t use his enough, though he does also have a curveball, so that helps keep hitters off-balance.
None of Sale, Feliz, or Bard have splits issues, and neither did Ogando or Wilson when they were relievers. They’re all effective against both left-handed hitters and right-handed hitters.
Predictive ERA (Using xFIP)
Sale = 3.00 xFIP
Bard = 3.05 xFIP
Feliz = 4.27 xFIP (3.50 xFIP in 2010)
CJ Wilson as RP = 3.22 xFIP
CJ Wilson as SP = 4.06 xFIP
Alexi Ogando as RP = 3.69 xFIP
Alexi Ogando as SP = 3.94 xFIP
Again, Feliz will not be a good starter in 2012 if he repeats last year’s skills, but even his 2010 form didn’t stand up to Sale and Bard as his xFIP was a half run higher than both. Greenhouse’s study shows that we should expect an uptick in ERA of somewhere between 0.2 – 0.4. This was true for Ogando, but not for Wilson whose xFIP increased by more than 0.8. However, I believe Ogando is a more similar type of pitcher with his 95-96 MPH gas than Wilson (plus Wilson’s ERA only increased by about 0.5). I’d expect Sale and Bard to both have an ERA in the low to mid 3s and Feliz to be in the mid to high 3s. I’m thinking around 3.2 ERA for Sale, 3.5 ERA for Bard, and 3.8 ERA for Feliz.
What Feliz does have going for him is that before he was converted to a reliever, he was a very good starter in the minors with over 50 starts under his belt. If he can go back to pitching like the 2010 form of Neftali Feliz, then he should have no problem being successful in 2012. Bard had a failed attempt in 22 starts in 2007, but that shouldn’t be weighed against him as he has completely transformed his pitching style since then. Sale hasn’t started since 2009 at Florida Gulf Coast University, but he still looks like he’s best-equipped to make the conversion back. He’s been compared to Josh Johnson, Jon Lester, and, interestingly enough, CJ Wilson – so that’s pretty elite company. For all 3 of these pitchers, their biggest problem aside from translating their relieving skills into starting skills will be how many innings they pitch. Although CJ Wilson pitched over 200 innings in his first season as a starter, as we saw with Ogando, it’s difficult to jump 150 innings and pitch 200+ innings in your first year as a starter. Expect about 160-175 innings and anything more is gravy. Despite this, I’d still expect Sale (and possibly Bard) to be in the top 40 starters for 2012 (and I’d certainly draft him first out of these 3). See the “Statistics” link below to look at the stats that I used to gather this information comparing the 5 converted starters as well as the articles that I referenced throughout this piece. Sale, Bard, and (especially) Feliz are all risky, but they all have ace upside with their stuff as well. We’ll see how well it translates.