After two weeks of identifying underowned pieces who could help you solidify your stretch run, I figure it is time to do the opposite. Below are three pitchers that you have likely been relying on for some time now, but may not want to do so now that it is all on the line. Their stats are pretty on the surface, but regression seems likely when their underlying metrics are considered.

Regression is an awkward word this time of year. Even if the metrics suggest it, there is no guarantee these pitchers will come crashing down to Earth before the season ends. Therefore, this column could also be considered the beginning of next year's draft prep, or advice to owners in keeper leagues looking to 2017 already. Regardless, let's look at J.A. Happ, Rick Porcello, and Felix Hernandez in further detail.

Editor's Note: Purchase a full season NFL Premium Pass (including DFS Premium), and also get MLB Premium + DFS for free through the playoffs. Premium DFS research, lineup picks, expert lineups, tools and more. You can see screenshots of our NFL Premium and MLB Premium and DFS tools. What are you waiting for?

The Fantasy Jury Is Out

J.A. Happ (SP, TOR)

Happ has rewarded owners who believed in his breakout last year, going 17-4 with a 3.34 ERA so far this season. However, his xFIP is an unsightly 4.06, and accumulating a bunch of wins is not predictive of anything going forward. The peripherals suggest that Happ should be a streaming option at most in standard formats.

First of all, Happ appears to have been fortunate this season. He is stranding 80% of the runners who reach against him, a higher figure than the league's 72% norm. Higher strand rates are sustainable for pitchers with strikeout stuff, but Happ's 21.6% K% does not suggest he has it. He also has a .277 BABIP against, significantly south of his career .290 mark. Add in the wins, and Lady Luck has smiled upon Happ so far.

Happ's current strikeout rate is league average, and his arsenal suggests that owners looking for more are bound to be disappointed. By SwStr%, Happ's change is his best pitch at 12.6%. He only throws it 6.6% of the time, however, and its chase rate of 23.3% is very meh. His 2-seamer offers a much higher chase rate (38%), but its 8.9% SwStr% is still unlikely to produce Ks with any frequency. The slider (9.8% SwStr%) and curve (9.4%) are also unimpressive for strikeout purposes. Happ's heat generates whiffs a well above average 10.2% of the time, but the Ks will not come until Happ has at least one real weapon to complement it.

It doesn't appear as though Happ is trying to develop that weapon. Happ is throwing more 2-seamers this year, going from 9.4% last year to 15.4% this. He is featuring fewer curves (10.8% to 7.5%) and changeups (9.4% to 6.6%) to make room for the 2-seamer. Put another way, he is throwing more of his worst pitch by SwStr% instead of the two offerings most likely to develop into a K pitch. Even if luck continues to help Happ, his lack of K upside limits his fantasy appeal.

Verdict: Chump

Rick Porcello (SP, BOS)

After last season's disaster, Porcello has once again firmly entrenched himself in the minds of fantasy owners with a 19-3 record and 3.23 ERA. Like Happ, his 3.93 xFIP is much less impressive. Also like Happ, his 20.6% K% doesn't really move the needle for fantasy purposes. His success to date has to do with Boston's strong offense and a BABIP of .268. Day to day offense can be fickle, and the BABIP doesn't feel sustainable to me.

One of the reasons Porcello's BABIP against is down is a lowish LD% of 19.1%, the first time since 2011 Porcello has posted a below league average mark. That will probably not become the new norm. His FB% is also up, from 32.5% last year to 36.5% this. This is a path to a sustainable decrease in BABIP, but it is playing with fire. Porcello's 10.1% HR/FB is fine, but he posted a 14.5% rate last year and has a career rate of 11.5%. More fly balls are probably a bad thing for Porcello long term.

Maybe you're thinking Porcello has changed his pitch mix to decrease his BABIP. That is also not the case, as the biggest repertoire change is more 2-seamers (29.8% to 36.3%) at the expense of fastballs (35.5% to 23.1%). PITCHf/x has a hard time distinguishing between the two, and most pitchers legitimately throw both sometimes. As a result, it is the least consequential of the possible pitching mix changes.

Could Boston's defense be boosting Porcello? The Red Sox are 5th in baseball in overall Defensive Runs Saved, so this theory passes muster initially. However, Porcello's batted ball profile suggests that this is probably not the reason for his success either. The majority of his team's defensive prowess is in the outfield, where Mookie Betts (24 DRS) and Jackie Bradley Jr. (9 DRS) collectively comprise nearly all of the team's 35 total DRS. Fly balls against Porcello this year have had a BABIP of .167 against a career average of .153, suggesting that Porcello's BABIP has little to do with Betts or JBJ. It's possible that they are helping with line drives (.651, career .694), but they are usually hits if they aren't hit right at a fielder anyway.

It is his .178 BABIP on ground balls, versus a career mark of .230, that is really driving Porcello's performance. On the Boston infield, Xander Bogaerts has -12 DRS at short while Hanley Ramirez is -5 at first base. They should be hurting, not helping, Porcello on ground balls. If you prefer Ultimate Zone Rating to measure defensive performance, the -7.4 UZR Bogaerts and Ramirez combine for leads to the same conclusion.

It should be noted that pitchers on the same team do not receive identical defensive support. Much like run support, some pitchers get more defensive help than others even with the same personnel behind them. There is currently no way to quantify defensive support by pitcher, but I believe that this type of analysis can still be useful so long as it is not taken as gospel.

BABIP regression will really impact Porcello's fantasy value, as there is almost no chance his strikeout rate is going up. His best pitch for Ks is a slider with an 11% SwStr% and 34.4% chase rate. That would be the fourth option for a true strikeout arm. The 3.5% SwStr% on his 2-seamer is laughably low as well, dooming his 10.9% SwStr% heater to achieve nothing better than his current 20.6% K%. Like Happ before him, Porcello is a questionable fantasy asset if the Red Sox stop handing him so many wins.

Verdict: Chump

Felix Hernandez (SP, SEA)

Felix's surface stats, including a 10-5 record and 3.75 ERA, are the least impressive in this column. Despite this, his 4.41 xFIP, 19.9% K%, and 10.4% BB% are indicative of a far worse campaign. There hasn't been a significant change in the King's pitch selection to blame these troubles on. Sadly, this looks like the beginning of the end of the King's illustrious career.

Felix's stuff is simply not as good as it used to be. His average fastball velocity is down significantly (92.1 to 90.5) relative to last year, reducing the SwStr% on both his heater (5.4% to 4%) and sinker (5.6% to 4.5%). The reduced fastball may also be helping hitters lay off Felix's curve, as its O-Swing% has gone from 43.7% last year to 36.7% this year. As a result, its SwStr% is down from 14.5% last year to 10.7% this. The slider's chase rate has held up, but its SwStr% is also down to 10.9% from 15.1%. Only Felix's vaunted change (17.1% SwStr%, 51% chase) still rates as a strong offering, and it takes at least two to post a strong K rate.

The declining stuff has produced a HR/FB of 15.1%. It may seem ripe for regression, but the King's 15.3% HR/FB over 200 innings last year lends credence to the inflated ratio. The fact that 28.3% of the fly balls Hernandez allows are pulled is also problematic. Worse still, Felix's FB% is up to 29.5% from 26.9% the year before, producing more balls affected by the HR/FB. I see no evidence that Seattle's ace is close to curing his gopheritis.

A .259 BABIP against is the only thing that has kept Hernandez's numbers remotely respectable. Seattle's defense has been terrible this year, posting -21 DRS and -22.7 UZR as a unit, so the BABIP may seem entirely luck driven. There is a little hope if you squint, though. Seattle's infield, led by third baseman Kyle Seager (12 DRS) and second sacker Robinson Cano (10 DRS), is actually above average. Therefore, his .181 BABIP on grounders allowed may be sustainable despite a career rate of .227.

Seattle's overall defense is dragged down by a dreadful outfield, an 11-headed monster that has combined for -22 DRS. This seems to have hurt Felix as you would expect. His .178 BABIP on allowed fly balls is a lot higher than his career .143 mark, and the best hit 15% or so of them aren't even counted in BABIP. The same cannot be said about line drives, as Felix's .585 BABIP against on them is far lower than his .702 career mark.

Felix's 75.5% strand rate almost matches his career rate of 75.2%, but it may regress since Felix has lost the K% that once sustained it. The competent defensive infield and the change's continued viability make Felix a viable streaming option, but he is no longer the must start he used to be.

Verdict: Chump

Live Expert Q&A Chats - Every Weekday @ 1 PM and 6 PM EST (DFS)

Fantasy Baseball Chat Room

[iflychat_embed id="c-55" hide_user_list="yes" hide_popup_chat="no" height="400px"]