Identifying 2023 TD Regression Candidates: TEs (Fantasy Football)
As the tight end position has evolved over the last decade, our strategy for drafting them in fantasy football has changed as well. Players like Travis Kelce and Mark Andrews are now being selected earlier in fantasy drafts because of the high floor and upside that they provide each season. Not only do they have the opportunity to finish with double-digit touchdowns, but they are also likely to finish with over 100 targets. So what happens when you miss out on drafting Kelce, Andrews, or even T.J. Hockenson this season? One alternative is to bank on a late-round option. The risk with these players is that they tend to be very touchdown-dependent, offering plenty of volatility from week to week. Fortunately, we can use the metric “Expected Touchdowns (xTD)” to identify the most volatile tight ends in fantasy football, highlighting which players to target and fade in your leagues.
If you are new to this series, be sure to check out the first two articles breaking down touchdown regression candidates at the running back and wide receiver positions.
Play-by-play data used for this metric was provided by nflfastR.
Calculating Expected Touchdowns
To calculate Expected Touchdowns, I created an xTD model that uses historical play-by-play data to determine the likelihood of a player scoring on any given play. This is based on a variety of variables such as the down, distance to the goal, and the type of opportunity (rush attempt, pass attempt, or target). For targets and pass attempts, I also factor in the depth and direction of the throw. Once we account for each of these variables, my model will calculate an xTD value for each opportunity on a scale from 0 to 1.
To better understand xTD for tight ends, we can take a look at the following play by Travis Kelce against the Las Vegas Raiders in Week 5:
- Type of Opportunity: Target
- Yards to Go: 4 Yards
- Yardline: Raiders’ 4-Yard Line
- Down: 3rd Down
- Air Yards: 3 Yards
Given that this play was in a goal-to-go situation, the likelihood of scoring was relatively high. As a result, my Expected Touchdowns model arrived at a value of 0.2321. In other words, historically, 23% of plays in this specific situation have resulted in a touchdown. To no surprise, Kelce scored on this play. As a result, his Touchdowns Over Expected (or TDOE) is calculated as follows:
- Touchdowns Scored: 1.00
- Expected Touchdowns: 0.2321
- Touchdowns Over Expected: +0.7679
As you can see above, TDOE is the difference between the touchdowns scored and their xTD. By contrast, if Kelce failed to find the end zone, his TDOE would be -0.2321 on this specific play. If we run this same analysis for every one of Kelce’s opportunities against the Raiders, we arrive at an xTD value of 1.31. In other words, based on the location and depth of his opportunities, Kelce should have scored at least one touchdown in this game. Because he scored four times, his TDOE for the week was at +2.69.
Naturally, Expected Touchdowns is a measure of opportunity and usage, as players that command more targets will have more opportunities to find the end zone. Touchdowns Over Expected, on the other hand, is a measure of efficiency, which is both volatile and prone to regression. The assumption is that TDOE will typically regress closer to zero over time. In fact, after a big Week 5 performance, Kelce regressed heavily to the mean (-0.26 TDOE) when he failed to score against the Buffalo Bills in Week 6, despite receiving more opportunities.
Interpreting xTD and TDOE
When we run the xTD model on every opportunity for every tight end in the league, we get the chart above. The x-axis represents a player’s Expected Touchdowns (xTD), while the y-axis shows the actual number of touchdowns scored. The gap between a player and the trend line is their Touchdowns Over (or Under) Expected. The expectation is that if a player receives similar usage next season, they are likely to regress to the line and score much closer to their expected touchdown value.
Based on this information, how likely is it for a player to regress to the mean? Historically, tight ends who scored above their xTD value experienced an 84.2% decline in Touchdowns Over Expected the following season. On the other hand, tight ends who scored below their xTD value were about 98% more efficient in their next campaign. In short, regardless of where a player is located on this chart, they are likely to regress toward their Expected Touchdown value the following year.
Negative Regression Candidates
George Kittle was by far the most efficient touchdown scorer at his position, finishing with +5.6 touchdowns over expected. While his efficiency was truly elite, historical data would indicate that Kittle is bound for regression this upcoming season. Per my touchdown model, we have only seen 12 tight ends since 2010 finish with at least five touchdowns over expected in a single season. Of those 12, ten would regress to the mean, averaging only +1.3 touchdowns above expected the following year. In other words, Kittle’s efficiency is almost guaranteed to come back down to earth. As for his usage and volume, Kittle averaged the lowest target and air yards share of his career since his rookie season, likely driven by the emergence of Brandon Aiyuk. And while his numbers did improve with Brock Purdy under center, Kittle’s target share of 20% was still a significant decline from his average over the last four years (25.3%). Make no mistake, Kittle will remain a TE1 this season. However, he could drop out of the top tier if his volume does not improve and his efficiency regresses down to average.
Cole Kmet was coming off of an inefficient campaign in 2021, failing to find the end zone despite receiving a career-high 94 opportunities. Naturally, we knew that regression was inevitable as Kmet would likely score at least once the following season. Sure enough, his touchdown efficiency swung completely in the other direction, finishing as the TE2 in Touchdowns Over Expected (+3.5) in 2022. After two seasons with very different touchdown totals, I expect Kmet’s efficiency to finish somewhere in the middle this year. The more significant concern would be the volume after the Chicago Bears traded for D.J. Moore earlier this offseason. Moore, who averaged a 24% target share in Carolina, will likely be the WR1 for this offense. And with Darnell Mooney also returning from injury, there may be fewer opportunities available in what should remain a run-heavy scheme. When you combine the concerns about his volume with his inevitable touchdown regression, it seems likely that Kmet’s fantasy production will decline this upcoming season.
Taysom Hill continues to be one of the most intriguing players to project because of his unique usage in the Saints’ offense. Unlike most tight ends, Hill was a true Swiss Army knife, contributing as a rusher, passer, and receiver in 2022. As a result, he finished the season as the TE2 in Expected Touchdowns (8.4), behind only Travis Kelce. In addition to the diversity of his usage, Hill was also one of the most heavily utilized red zone weapons at his position. In fact, about 14% of his opportunities were within the opponent’s 10-yard line. For context, an average tight end receives about 7.5% of his opportunities in that area of the field. So while his touchdown efficiency (+2.6 TDOE) should regress this upcoming season, Hill has the upside to maintain TE1 production, assuming the Saints continue to leverage him in high-value situations.
Positive Regression Candidates
Tyler Higbee might be my favorite late-round option this season. First of all, Higbee is coming off a season in which he set a career-high in target share at 20.9%. To put that into context, the only tight ends that ranked higher than Higbee in that metric were Travis Kelce, Kyle Pitts, and Mark Andrews. Even before Cooper Kupp’s injury, Higbee was already heavily involved, averaging the second-highest target share in the Rams’ offense from Weeks 1 through 9. Secondly, Higbee may also benefit from regression this upcoming season. As you can see above, he severely underperformed, scoring exactly three touchdowns below expected. Had he scored anywhere close to his expected touchdown value (6.0 xTD), he likely would have finished as a low-end TE1 for fantasy managers. And with an even clearer path to targets this season, Higbee is primed to be a value at his 12th-round ADP.
Pat Freiermuth‘s first two seasons are a perfect example of how a player can regress in touchdown efficiency but remain fantasy relevant due to an increase in volume. In his rookie season, Muth was one of the most efficient tight ends, scoring seven touchdowns on only 79 targets (+2.3 TDOE). Unsurprisingly, he regressed significantly in the other direction, scoring the third-lowest TDOE (-2.8) in 2022. Despite his touchdown regression, Freiermuth still finished as the TE13 in points per game as he set career highs in target share (18.8%) and air yards share (18.9%) in his second year. After narrowly missing out on a TE1 season, could 2023 finally be the Muth’s breakout campaign? That will largely depend on George Pickens’ involvement in the offense. However, if last year was any indication, Freiermuth is far and away the second option in this offense, behind only Diontae Johnson. And assuming Kenny Pickett can take a step forward in his development, the Muth should have every opportunity to finish as a TE1 for the first time in his career.
If you are looking to punt the tight end position entirely, Cade Otton could be an option in the 18th or 19th round of fantasy drafts. As you can see above, Otton’s performance left plenty of room for improvement as he scored nearly three touchdowns below expectation in his rookie year. Beyond his inefficiency, the larger concern was the inconsistent volume that kept Otton from truly breaking out. In 16 games, he only averaged a 9.5% target share and a 6.8% air yards share, severely limiting his upside week to week. As we flip the page to 2023, the Buccaneers released Cameron Brate without bringing in any additional competition, moving Otton up the depth chart as the unquestioned TE1. Unfortunately, this offense also lost its QB1 as Tom Brady finally retired after 23 seasons in the league. And while Baker Mayfield is a rather uninspiring replacement, keep in mind that David Njoku was a borderline TE1 in his sophomore year with Mayfield under center. This offers some hope that with enough volume, Otton could be a high-end TE2 for fantasy managers despite the downgrade at the quarterback position.