Our Lineup Optimizer is an incredibly powerful tool inside our Player Models, particularly when creating a large number of lineups. However, it’s just that — a tool. We still have to make decisions. Otherwise, everyone would have the same 150 lineups in each contest.
This is intended more as a teaching piece than a step-by-step guide. These rules are suggestions that can show you how to translate your read on this game into lineups.
Additionally, this piece focuses on DraftKings Showdown contests, but similar strategies are useful on FanDuel.
General Theory
The two most important factors in building +EV Showdown lineups are correlations and uniqueness. Building correlations into our lineups are rather obvious. Given that we’re only building around one game, every event on the slate will impact every player. Each of our lineups should be built around a specific scenario while trying to find the best way to profit if that scenario happens.
Some correlations are fairly straightforward. Matthew Stafford and Cooper Kupp are likely to have big games together, for example. Others are less so. This is an older article, but the Undervalued Correlations need to be considered here—namely, those across teams. Quarterbacks correlate with the opposing quarterback at a higher rate than any position other than their own top receiver.
The other factor, uniqueness, differs dramatically based on the type of contest you’re playing. When building for the 98,000 entry “Showdown Rush,” for example, finding unique lineups is very difficult. However, having a 1% chance at getting first place to yourself generate $2,000 in expected value. A 2% chance at a five-way chop is only “worth” $1,400. (Those percentages are way, way, higher than anybody’s actual expectation, but chosen for illustration purposes.)
In super large field contests, building some (intelligently) “anti-correlated” lineups has some value as well. The field is fairly sharp about which players and positions correlate. Therefore, going in the other direction can help boost your chances of a unique lineup — and your expected ROI. It’s a balancing act between the two critical factors that need to be adjusted based on your goals.
Generally, finding contrarian lineups is more valuable than “better” lineups in terms of our profit over time in large contests. This is less the case in smaller contests but still important given the top-heavy nature of most tournament payouts.
Team Allocation
It’s still profitable to build around “onslaught” style lineups in showdown contests. The field feels far more comfortable with “balanced” builds with three members from either team. We’re going to take that a step further and build specifically around the Bengals. Four members of the underdog team is the best long-term Showdown strategy, as the field overrates the likelihood of the favorite winning.
We’ll also stretch that out to include lineups with five Bengals. For a five-Bengals lineup to work, it would almost certainly be a large Bengals victory. That isn’t very likely — but it’s likelier than the field will be building for.
To do this, we just need to click the “teams” tab in our optimizer, then select the minimum and maximum members of either team to include. I’ve set both the Rams and the Bengals limits, but that’s not actually necessary. Selecting for either team (when only two teams are available) is sufficient.
Player Groups
Now that we’ve settled in on a Bengals-heavy approach let’s set some player groups. Starting with the Rams, let’s consider what has to happen for the Bengals to have four (or five) players in the optimal lineup. It probably means the Bengals get out to a big lead early on, forcing the Rams to pass.
Los Angeles has a fairly concentrated passing attack. Since Odell Beckham jr. joined the team, he and Cooper Kupp have accounted for 44% of the Rams targets and two-thirds of the passing touchdowns thrown by Matthew Stafford. I want to have at least one of them if the Rams are trailing. I’m fine with either of them in the captain spot as well.
On the Bengals side, it’s likely that the Bengals built their lead through the air. Their attack focuses primarily on three players — Tyler Boyd, Tee Higgins, and Ja’Marr Chase. It’s unlikely all three of them have a big game, but one or two of them should.
I also want to keep open the possibility that the Bengals score an early rushing touchdown or run the ball effectively from the front. Therefore, we’ll have a rule to include one of the Joe’s — quarterback Joe Burrow or running back Joe Mixon as well.
Taken together, these groups should account for three to five of our lineup spots, leaving the optimizer to fill in the rest.
Player Correlations
As discussed, correlating lineups is one of the most important factors in showdown slates. We’ll be avoiding some of the obvious correlations (for example, pair captain quarterback with two receivers) in the interest of staying somewhat unique. Instead, let’s try to find some less common groupings that give us an edge. These aren’t intended to be used for all of your lineups, but mixing them into a certain percentage can be valuable. They each have a specific scenario in mind and would likely be profitable if that scenario played out.
First, we’ll start by making a rule that lineups including the Bengals defense, must also include Matthew Stafford. These two players negatively correlate but can pay off together in specific situations. Namely, a Stafford pick-six. In a way, throwing an interception for a touchdown is a great thing for quarterbacks from a fantasy perspective. It gets your team the ball back immediately, but in a better game script than when you had it before.
The Bengals defense probably needs to score a touchdown to end up in the optimal lineup, so pairing them with Stafford helps differentiate lineups while still making sense for a specific scenario. This also forces our lineups to include four Bengals and two Rams, given our wide receiver rule above. Therefore, Stafford will still be paired with one of Kupp or Beckham when we include the Bengals defense.
On the Bengals side, we’ll be taking another anti-correlated — but logical — approach. If one of the Bengals star wideouts (Higgins or Chase) ends up as the optimal Captain, it’s likely that he drew an inordinate share of the Bengals’ targets. That means we’ve already accounted for a fairly big chunk of Burrow’s scoring. Therefore, the rule will be to exclude Burrow from lineups with Higgins or Chase at Captain.
The field generally pairs Captain wideouts with their quarterback — for a good reason. However, if either player scores a couple of early scores, the Bengals are likely to pass less overall, making it hard for Burrow to be a good value (even with two touchdowns under his belt). This has the added benefit of forcing Mixon into these lineups, given our “at least one of Mixon/Burrow” rule above.
I wouldn’t utilize either of these correlations for the bulk of my lineups but toy around with using them — or other correlations like them — for a small sample and see what the optimizer produces.
General Rules
Beyond the slate-specific rules discussed above, we can set some other rules to give us an edge. My favorite for Showdowns is the maximum salary cap setting. Simply lowering this from 100% gives us a huge edge in building unique lineups — leaving salary on the table is a great way to get unique builds.
We can also control the maximum exposure to any one player — customizable by position (including Captain). This also helps to diversify lineups, as it forces the optimizer away from the best Pts/Sal (or Ceiling/Sal) projected players. Setting the “bounce” also helps with this. The bounce cuts a player’s projections by a given percentage each time they appear in a lineup. They recover each time they aren’t, but this helps to diversify.
Additionally, we can control which players we allow in the Captain spot. Generally speaking, it’s OK to eliminate kickers and defenses from captain consideration. I’d leave the defenses as an option here, though. Stafford has a penchant for throwing touchdowns to his opponents, leading the league in pick-sixes the past two seasons. The Bengals offensive line is also a weak spot. They rank 31st in adjusted sack rate. That could help the Rams defensive unit rack up points.
While this was focused on the Bengals winning the game, we could tweak them based on how we think this game will play out. (Or give ourselves exposure to a variety of scenarios.) Besides shifting our exposure to four (or five) Rams, we’d want to focus more heavily on the Rams rushing attack and the Bengals passing attack.
Regardless, experiment with various rules and settings and see what you can come up with.
Good luck in the Showdown streets!
