Is Chris Archer Rosterable?

Introduction

I made the following controversial proclamation earlier this year – that Chris Archer is highly overvalued by fantasy owners. I go even further to say that in all but deep mixed leagues and mono leagues – continually rostering the right hander would be a poor use of fantasy resources.

My contention with Archer’s value stems from his ratio stats:

I further clarified what I meant by “almost no value in most roto formats” with the following tweet:

In 12-team mixed rotisserie leagues, if Chris Archer was available for pickup on my waiver wire – I may or may not bid on him depending upon the construct of my team and Archer’s forthcoming weekly matchup. I certainly would not spend much of my FAAB budget on him, as I wouldn’t intend to roster him for the entire duration of the season.

The response that was generated on Twitter from this statement was fascinating. On one hand, there were people of the mindset that I came from outer space. How could I not spend 50% of my FAAB budget to scoop up Archer if available??? Then, there were those with the total opposite school of thought – who uttered their complete disdain for Archer. They cited that Chris had been and forever will be a disappointing fantasy asset.

I found this bifurcated reaction fascinating because Chris Archer has simultaneously developed both a good and bad [fantasy] name. He indeed is a polarizing fantasy asset.

But let’s take a deep breath now. Let’s take the ‘name value’ out of the discussion and focus purely on the statistical numbers at hand.

Before continuing, I want to point out that my vantage point is an in-season reflection, and not simply a pre-season observation. Certainly, it isn’t all that difficult to make the case to pass over Archer during the pre-season drafts. Archer was selected as the 47^th most valuable pitcher (starter or reliever) in the NFBC during the month of March, according to his average draft position. All it took for anyone to uniformly pass on Archer was to have ranked him as the 60^th pitcher or lower.

The above is quite obvious, and not unique to Archer. For example, the ATC Projections found that Mike Clevinger was overvalued at the draft table this season. Although I did project Clevinger to earn a high rotisserie value in 2019, others simply projected him significantly higher. I predictably ended up drafting zero shares of him.

Today I am more concerned with the consideration of picking up / rostering a player with the statistical skill set of Archer during the season.

The Archer Line

Below is the full view of Archer’s 5×5 scoring category statistics over the past few seasons:

The ERA Trend does not look especially positive for Chris. Admittedly, a good part of the damage in 2019 came in his most recent outing against the Dodgers (6 ER in 4 IP) – and perhaps he was playing hurt.

His WHIP is looking absolutely poor lately; Archer’s walk rate has ballooned to 11% (from 8% in 2018). His FIP this season is even more ominous, at 4.82 – and his average fastball has lost 2 MPH from where it was last year. His fastball is one of his least effective pitches, yet he is throwing it a bit more this season. His HR/FB has risen to 15%, and that may even be a tad lucky in the current 2019 baseball environment.

No matter how one had previously valued Archer for 2019, it is clear that expectations for him need to be tempered from where they were pre-season. Below is a chart of how the FanGraphs projections systems have adjusted their views:

Wow – THE BAT seems to have totally lost confidence in Chris Archer going forward. His rest of season projected ERA trumps his already high 4.33 ERA to date. Since I’m all about averages – for now, let’s assume that Archer’s true ERA talent has ticked upwards about a quarter of a run. Archer is essentially [according to the average projection] a 4.00+ ERA starting pitcher.

What about Archer’s other 5×5 projected rest of season stats?

Here we have a bit more agreement going forward. The FanGraphs projection systems estimate that Archer will have a mid-to-upper 9’s K/9 rate, roughly a 1.30 WHIP, and about 7 more wins on the season.

Let’s now utilize the 4.06 ERA and 1.31 from above, and then scale the rest of his going forward counting metrics to a full season level. His full season projected line (based on ROS average projections) would look like this:

We will call this the Archer Line.

Now ATC of course would be a much more complicated amalgamation of projection systems, but for now let’s run with this simple average projected line for Archer.

Question

The question is …

How valuable is the Archer line?

How valuable is a sub 10-win pitcher with ~170 Ks, a 1.30+ WHIP and a 4.00+ ERA? To try and answer this question, let’s look at a valuation of the final 2018 season stats.

Method Assumptions

• 2018 Final Stats
• 5×5 Scoring Categories
• 15 Teams, NFBC-style positional requirements, with the condition that teams must roster at least 5 starting pitchers and 2 relief pitchers.
• My Z-Score method of valuating full season auction dollars, letting the Z-Score formulae set the H/P split.

Although teams in the NFBC traditionally do not have restrictions on the 9 pitchers, I will use 5 SP / 2 RP / 2 Any P to better reflect that in practice, teams use a minimum number of starters and relievers.

Further, to make sure that my valuation method wasn’t biased, I performed the exercise using the FanGraphs auction calculator (not shown). Results were typically very close to each other – values were no more than about $1-2 in either direction.

By the league assumptions above, the top 135 pitchers (9 pitchers * 15 teams) are worth $1 or more. All other pitchers will be worth less than $1. If we figure that each team may roster 4 additional pitchers to reflect the bench, there would be 195 pitchers worth observing. The 195 pitchers reflect those which ended up earning a full season value of at least negative two dollars. 195 pitchers should be sufficient for us to study the pitchers of “value” in shallow mixed leagues.

Note – My reflections on Archer’s value was intended for 12-team mixed leagues. However, I have conducted this study using a 15-team paradigm. [I happen to have this data more readily available, as it is consistent with the NFBC game formats in which I happen to play.] Certainly, if Archer appears to be a fringe roster player in a 15-team league, the same conclusion should be drawn for any formats that are shallower (12-team or less mixed leagues).

Results

To illustrate the value of the Archer line is, let’s see where it would have stacked up in 2018. Let’s take a look at all of the pitchers from the 2018 season who meet the following criteria:

• Top 195 pitchers (Value > $-2)
• ERA > 3.90
• WHIP > 1.20
• K/9 > 6.5
• SV < 5

To study comparable players, I included a stats buffer around all of the ratio components of the Archer line. I included ERAs down to 3.90, and WHIPs to 1.20.

The saves criterion here removes any closers, which we don’t want to consider here (we are looking at starting pitchers only). Rather than using straight strikeouts or innings criteria, I used K/9 to weed out soft tossing/contact pitchers, who wouldn’t compare to Archer [strikeouts are his main value]. I chose to completely ignore wins for this study, as they can be highly variable.

Below are the pitchers from 2018 who meet the above criteria:

There are only 13 corresponding pitchers from 2018. Only one of them, Marco Gonzales, earned a value above replacement (>= $1).

The above pitchers are largely comparable to the Archer line but differ in one or two categories. For example, Gonzales is a good comp ratio-wise, but earned more wins and threw less strikeouts. Castillo is a good comp for total strikeouts and wins but has a higher ERA. Jose Quintana emulates the scoring ratios well, earned a few more wins, but threw less strikeouts per inning. Joey Lucchesi may be the best ratio fit across the board but pitched fewer innings and thus racked up fewer counting stats.

The point here, is that the Archer line (170 K, 9 W, 4.06 ERA, 1.31 WHIP) does indeed stack up with players in this range [perhaps though on the higher side].

These players are not all that valuable.

The depth of value here translates to bench players / streaming options for a 15-team league. In a 12-team (or less) league – these are bench players / waiver wire material. They earned fringe roster value at best.

Perhaps the Z-Score valuation method gives more weight to ERA/WHIP than other valuation methods. But the bottom line is that 170 strikeouts, which is Archer’s main positive contribution to value, is not all that fantastic – and not enough to withstand the ERA/WHIP impact that it manufactures.

Let’s talk a bit about the poor WHIP of the Archer Line.

A 1.31 WHIP is a huge detriment to your fantasy roster. To give you some context of its mediocrity – out of 570 teams in the NFBC main event, only 94 teams currently have a WHIP of at least 1.31. That’s 16.5%, or about one-sixth of all teams. For every starting pitcher that has a 1.30 WHIP on your team, you would need to roster a starter with a 1.10 WHIP simply to balance out to a 1.20 WHIP. Currently, there are only 23 qualified starting pitchers in the Major Leagues who are sporting better than a 1.10 WHIP.

Archer has positives, but the negatives may outweigh the positives in terms of his roto value. And … we aren’t even talking about downside here – we are talking only about expected production.

Stats Replication

I have heard the following argument:

It is hard to find strikeouts on the waiver wire. Although Chris Archer may hurt your team ERA / WHIP, you can manage his ratios by constructing your roster around him.

I don’t really buy this argument. If you happen to roster a number of fantastic ERA/WHIP but low K/9 starters, then Archer may be a nice compliment to them. For example, if you happen to have rostered two of Miles Mikolas, Kyle Hendricks or Sean Manaea, then I’ll agree – your team may be able to withstand a 4.00+ ERA / 1.30+ WHIP.

I have also heard the following suggestion:

What if you used relievers on the cheap to manage the ratios?

Sure, that may work – but of course, you are now using two starting roster slots to manage one player’s ratios. Your team will surely suffer in the counting stats.

I would suggest that the Archer line isn’t necessary to be rostered at all. But, if you desire it, it can be replicated on the waiver wire. It takes more effort, but if you play a combination of 2-start pitchers when matchups are favorable, combined with a high strikeout middle reliever – you can replicate much of the Archer line.

In the example below, I distribute 2/3 of the playing time to 2-start pitchers, and 1/3 of the playing time to low-cost high-K middle relievers. For the 2-start pitchers, I used the replacement level pitchers from above as a statistical guide, but worsened their ERAs. For the middle relievers, I used Chad Green as a rough guide.

The replicated line generates a slightly higher ERA (than Archer), with a few less strikeouts. It will earn roughly the same number of wins and it will match the WHIP. Of course, if you intelligently choose pitcher matchups that will outperform a 4.60 ERA and an 8.5 K/9 in the aggregate (which you may be able to do in a 12-team mixed league), then the replicated line will come even closer to duplicating the Archer line.

For the replicated line – although it does require occupying two roster slots, it only occupies one starting roster slot.  It is an efficient use of a bench slot.

Conclusion

For those who believe the average rest of season projections on FanGraphs, in a 12-team (or smaller) mixed rotisserie league, Chris Archer is not a player that one should spend considerable resources to roster.

If you valuate the projected full season line [159 IP, 4.06 ERA, 1.31 WHIP, 9 W, 170 K] for Archer, you will find it to be below replacement level – and is a line which could effectively be replicated (if desired) on the waiver wire using few fantasy resources.

For those who peg Archer’s ERA much lower (say at 3.75) or conclude that his walk rate will improve – he would be far more valuable. If not, you don’t need to roster Archer. You should deploy any allocated/constrained fantasy capital elsewhere.