Julio Teheran was tagged for 6 earned runs across 6 1/3 IP on Tuesday. The Red Sox hit 4 doubles, a triple and a home run. Brock Holt completed ¾ of his cycle before Teheran’s exit as part of 13 hits the Red Sox tallied off the Braves’ starter. The pitcher once seen as the future staff ace for the Braves has seen his share of disappointment this season but, while ERA estimators foretold of a step back for Teheran in 2014, it hardly looked like he was headed for collapse. Upon further review, fixing Teheran may not be as as simple as ditching a specific pitch or tweaking one element of his approach.

Entering the season, plenty of questions surrounded Teheran. While he was effective at keeping runs off the board in 2014, finishing 19^th among all qualified starters with a 2.89 ERA in 221 IP, better ERA predictors, like xFIP and cFIP, indicated he might not be a top 20 starter going forward. Teheran finished 46^th among qualified starters with a 3.70 xFIP and tied for 53^rd among pitchers with more than 100 IP with a slightly better than average 99 cFIP (cFIP is scaled so that 100 is average and lower numbers indicate more successful pitchers). While these weren’t bad results, they certainly did not point to another season in which Teheran would finish in the top 20 in ERA. Teheran was also playing with fire by allowing a high percentage of hard-hit balls: 32.1% of balls hit off of Teheran in 2014 were considered “hard contact”, 19^th worst among starters.

In addition to the mediocre ERA indicators, Teheran’s average fastball velocity slipped nearly a full mile per hour last season from 92.1 in 2013 to 91.3 according to Fangraphs’ PITCHf/x data. His first-pitch strike percentage also dropped from 65.4% in 2013 to 60.3% last season. First-pitch strike throwing and changes in fastball velocity can affect runs allowed but Teheran was able to escape mostly unscathed last season. This season is a different story.

Including the 6 runs allowed in Boston, Teheran’s first two and a half months of 2015 have been disastrous. His 5.07 ERA, 5.05 FIP, and 4.23 xFIP rank in the worst quartile of the 105 qualified starting pitchers on Fangraphs’ leaderboards. His 6.08 DRA ranks in the worst 20 starting pitchers with 30 or more IP and his 114 cFIP does not paint an optimistic picture. By any measure of runs allowed or predictive stats, Teheran looks like a broken man. Home runs account for a significant portion of Teheran’s woes as his 17.1 HR/FB% ranks 6^th worst in baseball, but what has caused Teheran to collapse?

Chart courtesy of BrooksBaseball.net

Looking at Teheran’s pitch selection in 2015, he has continued to primarily mix his four- and two-seam fastballs and use his slider as his primary breaking pitch, mixing in a curveball and changeup. Looking at BrooksBaseball.net’s pitch usage chart, Teheran has steadily increased his use of his four-seamer at the expense of his two-seamer (classified as a sinker by BrooksBaseball.net) and slider. His curveball and changeup usage has increased slightly as the season has progressed but we’d need to see more data to know if that is more than just situational usage.

The increase in four-seam fastballs could have something to do with Teheran paying attention to outcomes. According to Fangraphs, only his four seam fastball has provided value this season, with his other four pitches providing negative or negligible value. This month, Teheran is actually throwing his four-seam fastball harder than he has for years, averaging nearly 93 mph. Looking at the pitches themselves paints a muddy picture, however, with no clear link between a single pitch and his struggles.

Above is a table of pitch velocities and movement taken from BaseballProspectus.com’s PITCHf/x leaderboard (Teheran’s two-seam fastball is considered a sinker by its standards). Teheran’s ability to generate whiffs on his sinking two-seamer has dropped significantly this season, but it’s hard to tell where that additional contact is coming from by just looking at the data. His two-seamer is actually averaging over a mile per hour faster than last season and the horizontal and vertical movement are slightly higher than last season. The missing whiff rate on swings against the two-seamer may be worth further investigation. Also notable is the drop in whiff rate against Teheran’s changeup. Like the two-seamer, the velocity on the change has increased slightly, and the horizontal movement has increased. Teheran has lost a little vertical movement off his changeup (5.53 in 2014 vs. 4.67 in 2015), but what PITCHf/x’s movement stats actually tell us is the position of the ball against where it would be expected to end up absent any spin effects. Would less than an inch difference in that perceived final placement really lead to a whiff rate less than half of last season’s percentage?

Maybe Teheran’s struggles are explained by something simpler than the effectiveness of individual pitches? In 2013, Teheran was among the league leaders in pitches thrown for strikes, with over 53% of his pitches in the zone. Last season, it slipped to just over 52% and this season he’s throwing only 49.2% of pitches in the zone. Teheran’s walk-rate has exploded to 8.6% of batters faced. Teheran doesn’t strike out batters at high rates like Trevor Bauer or Tyson Ross, so he can’t rely on stranding runners once they get on. In addition to the walk rate, the earlier mentioned poor home run rate means a lot of those runners are scoring on balls out of play so Teheran’s respectable-at-a-glance .299 BABIP doesn’t mean much. Walks are a problem, home runs are a problem, so what can Teheran start to do differently to get on track?

So far this season, Teheran is the fourth worst starting pitcher at throwing strikes on the first pitch, with just 55.8% of plate appearances beginning 0-1 or with the batter putting the ball in play. That represents a nearly 10% slide from his breakout rookie season. On Tuesday, he threw 17 of 29 first pitches for a strike, slightly better than his season average, but one of those strikes was hit for a home run by Brock Holt in the bottom of the 7^th, just one batter before Fredi Gonzales pulled the hook on Teheran’s night. If getting ahead in the count used to be a key to Teheran maintaining his ERA it appears he dropped that key on the ground and then accidentally kicked it in the gutter.

Surprisingly, Teheran doesn’t pitch from behind in the count as often as his walk rate and first pitch strike percentage would lead you to believe. According to BaseballSavant.com’s PITCHf/x leaderboards, Teheran ranks 137^th among pitchers who have faced 50 or more batters this season in percentage of pitches thrown while behind in the count. What it does tell us is that it appears Teheran isn’t as good at pitching while behind in the count as other pitchers, like Max Scherzer, who has thrown the highest percentage of pitches while behind in the count.

If Teheran’s control is declining, falling behind in counts certainly isn’t a situation Teheran wants to find himself in. It’s possible that Teheran could lean more on his four-seam fastball, his one valuable pitch so far this season. It’s not certain that throwing more fastballs will turn Teheran into Bartolo Colon, though, especially with his control issues. What is also certain is that Teheran is a mess and there’s no clear way out for the starter as things stand.