A few weeks ago, I wrote the first part of this series on predicting multi-TD games. Read the full breakdown to see all of the data, but here are the main takeaways:
• Emphasize volume over price or matchup for running backs.
• Emphasize talent over all else for wide receivers.
• While high DK Bargain Ratings typically lead to value, high FD Bargain Ratings are a better indicator of multi-TD upside.
• Focus on matchups more with tight ends than running backs or wide receivers.
• Potentially fade running backs and wide receivers with great matchups. Or at least don’t be scared to roster those positions even when they don’t have great matchups.
In Part 2, we’re going to look at Vegas data from multi-TD games from the last couple of years and see what trends we can find that could potentially lead us to these rare occurrences — Black Swans perhaps? — in 2016. Specifically, we will look at the following data for every multi-TD instance from our sample:
• Implied point total
• Opponent implied point total
• Spread
• Absolute value of the spread
• Spread percentage
• Game total (over/under)
• Line movement
• Absolute value of the line movement
Of course, the information from the multi-TD games is only useful when compared to baselines. Let’s establish those first. Here were the averages across all 2015 games, as a reference point for a typical NFL year:
| Implied Points | Opp Implied Points | Spread | Spread AV | Spread% | Total | Movement | Movement AV | |
| NFL | 22.58 | 22.58 | 0.00 | 4.86 | 50.00 | 45.11 | -0.08 | 0.89 |
And now, here’s how the total sample and each individual position compared (click here for a color-formatted table):
| Implied Points | Opp Implied Points | Spread | Spread AV | Spread% | Total | Movement | Movement AV | |
| NFL | 22.58 | 22.58 | 0.00 | 4.86 | 50.00 | 45.11 | -0.08 | 0.89 |
| Sample | 24.00 | 22.34 | -1.69 | 5.12 | 53.85 | 46.32 | 0.06 | 0.89 |
| RB | 24.22 | 22.03 | -2.27 | 5.08 | 54.88 | 46.29 | 0.01 | 0.85 |
| WR | 23.77 | 22.60 | -1.17 | 4.92 | 52.34 | 46.32 | 0.09 | 0.92 |
| TE | 24.08 | 22.36 | -1.72 | 5.81 | 55.65 | 46.39 | 0.11 | 0.89 |
Implied Points Totals
Right away, we see very intuitive results: The RBs with multi-TD games came from 1) high-scoring offenses 2) that were favored 3) by quite a bit. I know . . . you really collected all of this data to come to that conclusion?
Yes, and here’s why: Our Trends tool is amazing. It helps us find hidden value and edges that can’t be found anywhere else. For example, we see that RBs projected for at least five DraftKings points and on teams that are at least -3 favorites and implied for at least 24 points have historically done well:
However, that doesn’t necessarily tell us about their multi-TD upside — the kind of upside we want in the DK Millionaire Maker. The data in the table does. And sometimes that data will match up with that traditional value that we know — target RB favorites on high-scoring offenses — but sometimes the data isn’t so intuitive.
For example, TEs in the multi-TD sample were typically from favorites, as shown by their average -1.72 spread. That was the case for both RBs and WRs, too: The RBs were the heaviest favorites with an average spread of -2.27. However, even though RBs had a much larger spread on average, TEs actually had the largest absolute value spread.
This means that TEs had a much wider distribution of games: Even though their average spread fell in line with the overall sample, their high absolute value suggests that some of the TEs in the multi-TD sample came from huge favorites and some came from huge dogs. For example, Delanie Walker had a multi-TD game in Week 15 last year, a game in which the Titans were +14.5 dogs and ended up losing 33-16.
This falls in line with what we know about the distribution of fantasy points for different positions. To show this, I took the top-350 fantasy players last year and charted the average yards (rushing plus receiving) per touchdown for each position:
This shows that WRs were less dependent on touchdowns for fantasy production than RBs and way less than TEs.
And all of this connects to the absolute value spread. Tight ends are touchdown-dependent, and touchdowns can be scored whether a team has positive or negative game flow. Running backs and wide receivers, however, are very dependent on game flow for their production. Put simply: Tight ends need red-zone trips, whereas running backs need rush volume and wide receivers need targets. For a tight end to be successful, he basically needs one target at the goal line, and that can happen regardless of game flow. But running backs and wide receivers need repeated touches and targets all over the field. They normally don’t get those unless the game is flowing their way.
Other Vegas Metrics
Tight ends in our multi-TD sample had the highest average total (over/under), which makes sense, as tight ends can flourish as red zone threats in high total games, regardless of whether they’re on the positive end of those games. As you’d probably expect, all positions were above league average when it comes to game totals.
The line movement data here is interesting: Overall, lines in 2015 moved on average -0.08 points. The reason that number isn’t zero is because bets come in on the game total, spread, and moneyline. In a vacuum, if a team’s spread moves by -0.5 points, you’d think that the opposition would move +0.5 as a result. However, if the overall game total moves down by 3.5 points — because of unexpected weather concerns, for example — then both teams would see negative movement. A line moving for a single team doesn’t necessarily inform us about their opposition.
Anyway, what the line movement data does suggest is that the public was more on these multi-TD game instances than they were on a regular NFL game. As you can see in the table above, I also looked at the absolute value of line movement to see which positions saw the most extreme movement regardless of direction. This seems to be the theme of Part 2, as tight ends saw the heaviest movement in the positive direction but their absolute value was right in line with league average. Again, this suggests that there was extreme movement both ways — positively and negatively — as the sample would need to contain instances of strong negative movement in order for the absolute value to be ‘average’ and the overall average movement to be positive.
We’ll end Part 2 here — Part 3 will look at more traditional positional metrics and will be out next week — but I think that we came away with some interesting findings, especially from the tight end position. Although it is still optimal to target RBs and even WRs to a degree from favorites in high-total games, you can likely find contrarian value in TEs by targeting the extremes: Tight ends who are either big favorites or big dogs. And I’ll finish with good news: You don’t have to forgo ‘value’ to roster these guys.
Hello, Antonio Gates in Week 1!
