Alexander Smith and the Terrible, Horrible, No Good, Very Bad Fantasy Season (Plus Five Other Quarterbacks Evaluated)

Alex Smith was second only to Russell Wilson in fantasy points during the fantasy-relevant portion of the 2017 season, but his fantasy value was practically zero. Matt Ryan’s value was substantially less than zero. I’ll explain what I mean in this article, which also looks at four other quarterbacks’ 2017 seasons.

I have already written about the fantasy value of the top two fantasy QBs of 2017, meaning the two who scored the most fantasy points over the season: Russell Wilson and Cam Newton. As described in that article, I measure a player’s fantasy value in Fantasy Value Points (FVP) in order to understand what factors impact a player’s fantasy value. FVP represents the sum, on weeks in which the player is a projected fantasy starter, of the average fantasy point differences between the player and members of a comparison set representing likely next best options.

Today we look at six more of 2017’s top quarterbacks. Four of them were top quarterbacks in the sense of being drafted high: Tom Brady, Aaron Rodgers, Matt Ryan, and Drew Brees. The other two, Alex Smith and Carson Wentz, were not top-drafted quarterbacks but are top quarterbacks because they had very successful seasons.

Brady: The Number 2 QB

In the Russell-Newton article, I argued that Russell Wilson was worth 90.42 FVP and Cam Newton was worth 40.84 FVP in 2017, and I teased that although Newton was the number two quarterback in overall fantasy points, he was only the third highest quarterback in FVP. The second highest FVP total belonged to Tom Brady, at 52.47, despite the fact that Newton outscored Brady in fantasy points, 288.38 to 280.28, in weeks 1-16 (FVP disregards week 17 scoring).

The reason for this apparent anomaly is that in one week – week 4 of the 2017 season – After poor performances in weeks 1-3, Newton was projected as the QB 22 in the fantasydata.com projections. The FVP formula penalizes players for low weekly projections to reflect the fact that low-projected players usually sit on the bench, and you don’t score fantasy points on the bench. Week 4 happened to be one of Cam’s best weeks of the season, 33.04 fantasy points, but because of the bench penalty, his FVP for that week was only 3.82.

The weekly FVP number is derived by (1) examining the comparison class of presumed replacements, which in the case of high-ADP quarterbacks like Newton and Brady, is the set of quarterbacks projected to finish between QB13 and QB24 for the week; (2) for each member of the comparison class,  determining whether the quarterback we are evaluating is projected higher than the member of the comparison class; (3) if so, taking the difference in fantasy points for that week between our quarterback and the member of the comparison class; (4) if our quarterback  is not projected higher, assigning the value 0 to reflect the fact that our QB ought to be benched that week; (5) averaging these values over the entire comparison class.

In Newton’s case, since he was the projected QB22, step 4 has us assigning him 0 points versus QBs 13 through 22. Step 3 has us taking the difference between his production and those of the projected QBs 23 and 24, who were Ben Roethlisberger and Derek Carr. Newton scored 33.04 fantasy points, Roethlisberger scored 10.54 and Carr scored 9.72, so Newton gets credit for (33.04 – 10.54) + (33.04 – 9.72)  = 45.82 points. To get the average over the entire comparison class, we divide by 12, yielding 3.82.

(You can support the Cam Newton Foundation here)

In comparison, Brady’s Week 4 FVP was 1.42, despite his much lower fantasy point total of 20.48, because he was projected as the number one QB for the week. That means he was projected to start vs. all members of the comparison class. The QBs in the comparison class scored a total of 228.78 fantasy points that week, an average of 19.06. Brady’s FVP total for the week is the average difference between his point total and the comparison class’s, which is (20.48 x 12 – 228.78)/12 = (245.76 – 228.78)/12 = 16.98/12 = 1.42.

We can see that if Newton had been projected as a top-12 quarterback, his FVP for the week would have been (33.04 x 12 – 228.78)/12 = (396.48 – 228.78)/12 = 167.7/12 = 13.98. He missed out on about 10 FVP because of his low projection in week 4.

Aaron Rodgers: The Injury Difference

Aaron Rodgers was injured at the beginning of week 6, which led him to miss all of the rest of the season except week 15. He finished the season with only 15.64 FVP in 7 games, an average of 2.23 FVP per game. This projects to 33.45 over a full 15 games (remember, week 17 doesn’t count, and there’s one bye week in weeks 1-16). But if he hadn’t been injured, we would project him as the QB2 with 69.9 FVP, ahead of Tom Brady and behind only Russell Wilson. The reason for this large discrepancy is the double-whammy that his seasonal FVP suffered as the result of being injured early in the game.

First, his seasonal total suffered from his -12.32 FVP performance in week 6. Since the negative total was due to injury, our projection would restore that amount, giving him a projected total of 15.64 + 12.32 = 27.96. Second, for the purpose of averaging, we would reduce the denominator from 7 to 6, giving us a respectable per-game average FVP of 4.66, corresponding to a projection of 69.9 over a 15-game season. The injury therefore caused Rodgers to lose a projected 36 FVP points on the season (the 66.9 projection if we factor out the injury,  vs. the 33.45 projection if we don’t), over and above the 8 games he missed.

Matt Ryan’s Sub-Replacement Season

Matt Ryan’s was drafted as the QB5 before the 2017 season after his highly successful 2016. He disappointed, scoring only 228.1 fantasy points. The FVP formula demonstrates just how bad his performance was: his season total was -23.31. A negative FVP means he had negative value compared to the projected replacements. So fantasy player would have been better off playing a replacement-level QB than Matt Ryan in 2017.

Primarily, Ryan’s low FVP was due to not scoring very many fantasy points. His 228.1 total is an average of 14.26 points per game. But this total includes his second best week for fantasy points, which was week 17. Because FVP excludes week 17 points, his 18.08 points that week don’t count. His total fantasy points for the fantasy-relevant weeks was 210.02, an average of just 14 points per game. The average replacement-level QB scored 15.65 fantasy points per week.

Ryan also had the misfortune of having his bye week on a week where the replacement-level QBs had a particularly bad week, averaging 12.74 fantasy points. Compare week 9, Tom Brady’s bye week, where the average replacement-level quarterback scored 16.85 points.

Ryan’s poor performance caused him to be projected as a replacement-level quarterback several times later in the season. Each time this happened, he outperformed the average replacement-level QB. This means that his above-average performances received slight bench discounts, further hurting his season totals. in week 8, he scored 15.86 points as the projected QB13, compared to an average of 14.74 for the replacement level QBs. In week 11, he scored 17 points as the projected QB13, compared to 12.73 average for the replacement-level QBs. And in week 16, he scored 14.22 as the QB14, compared to 14.09 for the replacement-level QBs.

Leaving out these 3 weeks, Ryan’s best performance compared to the replacement-level QBs was week 12, where his 16.78 points beat the replacement-level average of 14.13 points by 2.65 points. In contrast, he twice scored at least 10 points less than the replacement-level average: week 4, where he scored 8.48 points and the replacement-level QBs averaged 19.06 points; and week 13, when he scored 7.02 points compared to the average of 17.06.

Drew Brees’ Bad Season

Drew Brees didn’t have a bad year in real football, but the Saints’ pivot towards the running back meant he had a low-scoring fantasy season. There is not much to say except that he did not score many touchdowns, and therefore did not score many fantasy points. As a result, he did barely better than the weekly QBs 13 through 24, and ended up with only 11.15 FVP on the season.

Carson Wentz’s Breakout

Carson Wentz had an outstanding fantasy season for a quarterback who was not drafted as a QB1. Because he was drafted as a QB2, we set his comparison class as the QBs 1 through 12. That means he only got credit for the differential between his fantasy points and the points scored by QBs projected lower than him but higher than QB13 that week. It helped that he got off to a hot start and was projected as QB 6 or higher seven times.

Wentz scored no points in weeks 1-3, 6, 10 (his bye), and 15-16 (his injury). In the other  weeks, he scored 15.08 points as the projected QB8, good for -0.9 FVP (week 4), 27.26 as the projected QB8 (3.15 FVP) (week 5), 31.02 as the projected QB6 (6.67 FVP) (week 7), 17.14 as the projected QB3 (1.85 FVP) (week 8), 24.76 as the projected QB11 (0.79 FVP) (week 9), 20.02 as the projected QB3 (3.11 FVP) (week 11), 23.98 as the projected QB5 (4.41 FVP) (week 12), 16.92 as the projected QB4 (2.23 FVP) (week 13), and 27.24 as the projected QB3 (9.66 FVP) (week 14). His total was therefore 30.97. Extending the average over 15 games would have been good for 35.7 points, but misses the fact that his bench discount was concentrated in the first half of the season, and he was routinely projected as a top-6 QB starting week 7, during which time he averaged about 21 points per game. Those seven games accounted for almost all of his seasonal FVP, and he averaged over 4 FVP/game during that time. If his performance in weeks 15 and 16 would have kept this pace, he would have made a run at Cam Newton’s season total of 40.84 FVP.

Alex Smith: Bad Timing

Alex Smith was the number two quarterback during the fantasy-relevant season, weeks 1 to 16, scoring 295.18 fantasy points. He fell to fourth behind Brady and Newton in week 17, when he was benched in favor of Patrick Mahomes. Nevertheless, his total FVP on the season was 0.79, meaning he was basically indistinguishable from his projected replacements. Why such a large anomaly? Partly because his comparison class was the top 12 QBs, and he wasn’t a top projected QB often; but largely because his timing was terrible, and he saved his best weeks for when he wasn’t supposed to be a starter, and his worst weeks for when he was supposed to be a starter.

As a result of low projections, Smith got 0 FVP points in weeks 1-4, 8, 13 and 15 because he was not viewed as a starter those weeks. But his three top fantasy performances came in these weeks: 37.64 in week 15, 31.02 in week 1, and 27.32 in week 4. These 95.98 points account for 32.5% of his seasonal points, but the FVP formula gave him no credit for these games because of the bench discount.

Smith’s other 20+ point games were in week 5, where he was the projected QB9 and scored 26.86 fantasy points, and week 7, where he was the projected QB7 and scored 25.68 points. These discounted performances added up to a paltry 6.99 FVP. So in all, Smith’s top 5 games, in which he scored 148.52 points – more than half his season total in 1/3 of the games – added up to only 6.99 fantasy value points. In comparison, Russell Wilson exceeded this FVP total in a single game eight times last season, and doubled it three times in a single game. Timing matters.

Smith’s entire FVP from his top five games was almost undone by his one worst game, week 11 in which he scored only 7.9 points. While this is very good for a worst game, it happened to be one of the very few times in the season that Smith was highly projected at QB4, with the result that the negative value of the game was weighted quite heavily. It also happened that out of the quarterbacks in the comparison class that he was projected to start over, several had strong weeks: Brees had 21.6 fantasy points, Cousins had 25.98, and Roethlisberger had 28.96. The average QB he was compared to outscored him by an average of about 9 points, translating into an FVP of -5.98.

Smith’s second worse game was week 14, when he was projected as QB4 again. He scored 11.12 points for a negative FVP of -2.43.

Summarizing the results for Alex Smith so far, his top five games (all of his 20+ point games) netted him 6.99 fantasy points, and his bottom two games (both of his 13- point games) netted him -8.41 points. Cumulatively, these come to -1.42 points.

The rest of the season was largely a wash. Thanks to average performances and low projections, his positive and negative performances barely registered on the FVP scale. This includes Smith’s other week as a top projected QB, week 9 where he was ranked QB4 but scored only a slightly above average 18.42 points, good for 0.19 FVP.

Summary

Russell Wilson, to top fantasy scorer last year, was the QB1. Aaron Rodgers had a terrible fantasy season, but is projected to have been the QB2, with 69.9 points, but for his injury. Tom Brady was the actual QB2. Cam Newton was the QB3. Drew Brees had a terrible fantasy season despite a decent QB season. Alex Smith had a replacement-level fantasy season despite being the QB2 in fantasy points during the fantasy-relevant season. Matt Ryan had a sub-replacement level season.

To close out the post, here are the 2017 FVPs for the quarterbacks discussed here.

Russell Wilson: 90.42

Tom Brady: 52.47

Cam Newton: 40.84

Carson Wentz: 30.97

Aaron Rodgers: 15.64

Drew Brees: 11.15

Alex Smith: 0.62

Matt Ryan: -23.31

One thought on “Alexander Smith and the Terrible, Horrible, No Good, Very Bad Fantasy Season (Plus Five Other Quarterbacks Evaluated)”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s