Replacements – Following Up

In our last post, we discussed the problem of placing a value or ranking on a pitcher like Clayton Kershaw, someone nominally expected to be a top starter but is projected to have fewer starts than other top starters. How do we incorporate the value from his replacements when drafting him? My conclusion was that we were not effectively doing so already, but there are a few related points from that discussion that I still wanted to follow up with this post.


I might have overrated replacement value of the starters, simply because of the timing issue. In most season-long leagues, we make our roster moves once a week. An injury doesn’t conveniently occur in time for us to replace that starter in time for his next start – and in fact in many cases he makes an abbreviated start before leaving with an injury. Many times it can be an Ivan Nova special – where the pitcher tries to gut it out with an injury and gets blown up because he’s at less than full capacity. In other instances the team just might opt to skip a start or push that starter back 2-3 days – enough for you not to get the full benefit of that in your weekly lineup.

Liss and I discussed that concept on the air and concluded that instead of using our replacement level to bump Kershaw up to 32-33 starts, it should be closer to 30 starts. Thus, instead of adding eight starts from our proverbial 4.35 ERA, 1.36 WHIP pitcher and getting the strikeouts and win opportunities that he provides, we’re going to lower that to four-to-five additional starts.

Liss also raised the possibility that I’m underrating the value of the pitchers available – that there is a “survivor’s benefit” from the remaining talent pool. The idea here is that we’re working just with a set of projections, and not the actual pitchers that are still getting regular turns in the rotation for their teams at the time. The teams in real life will move on from those replacement level pitchers not living up to those expectations, and of course many ¬†others will emerge that are better than those projections. But I’d push back against that changing our level too much – because that presumes that we’ll do better than our opponents in finding those guys, and that the timing in finding those guys will work in our favor.

Another issue not addressed in the first article is whether your league uses DL slots. ¬†It seems pretty obvious if you think about it, but if you use DL slots instead of having to park your injured players on reserve (as is the case in the NFBC), the value of the replacements available will be lower. If you’ve ever played in a deep league that also uses a DL list and have tried to find replacements, you know what I’m referring to.

This also seems obvious, but the replacement issue applies more when applied to better pitchers in the original. You’re starting with much better ratios when you lose a pitcher like Kershaw than, say for example, Chase Anderson. Anderson vs. Jeff Samardzija in many cases would be part of your regular start/sit decision process to begin with, so it’s less important to account for replacing Anderson’s innings missed due to injury.

RotoWire has the best daily fantasy football tools on the web.
Try Our NFL Lineup Optimizer Now

Clayton Kershaw was the inspiration for this topic, but who else might merit a slight upgrade based on replacement value? Here are a few prominent starters with fewer than 30 projected starts (on RotoWire, of course):

  • Noah Syndergaard (29) – Thor made just 25 starts last season, missing time most notably due to contracting hand, foot and mouth disease, but also due to a strained right index finger. Luckily neither injury was related to previous elbow and shoulder woes. Unfortunately, he still lost about one mph off his fastball and struck out one less batter per nine innings. Yet his fastball remains elite and the strikeout rate is still great, if not elite.
  • Stephen Strasburg (29) – Given that Strasburg hasn’t made 30+ starts since 2014, his presence here hardly is a surprise, and 29 projected starts might be generous.
  • James Paxton (27) – Paxton threw 160.1 innings in 28 starts, both career highs for him in the majors (in 2016 between the minors and majors he combined for 31 starts over 171 innings). He also averages fewer innings per start than some of the other pitchers on this list – should we be giving him a little less credit for replacement value as a result?
  • Charlie Morton (27) – While I didn’t like seeing Morton leave the Astros, the Rays were one of the better alternative landing places for him. They’ve shown how creative that they can be in managing their pitching staff. Unfortunately that creativity can hurt in the form of missed or shortened starts, so I’d be careful not to apply too much replacement value with him.
  • Rich Hill (23) – Hill checks a lot of boxes for our discussion – strong ratios when healthy, but prone to problematic blister issues that force him early from starts, delay his returns from the DL, and pitching for an organization that is quick to push him back, as they have exceptional starting pitching depth. The Dodgers can be as difficult as any team to handicap when comes to managing injuries.

Doing these two blogs is inspiring me to revive the notion of doing Playing-Time Neutral Rankings, starting with the pitchers. I’m playing with the idea of not only upgrading those missing time, but by setting replacement value at 30 starts, perhaps reducing the value of “average” starters that nonetheless are projected to get a full 33 starts.