D-UP Draft Strategy Guide 2025
Want to give yourself the best chance of taking home the $2,000 prize for 1st place? This guide is the resource you need for D-UP drafts.
The D-UP tournament by FastDraft is the largest IDP best ball format of all time but has several wrinkles that make it stand apart from other best ball tournaments:
Each D-UP draft is comprised of 10 teams and 6 rounds for a total player pool of 60 per draft
For reference, Underdog’s most recent Best Ball Mania drafts used 12 teams and 18 rounds for a player pool of 216 per draft
D-UP drafts are positionless, so you can build a team of six linebackers, six defensive linemen, six defensive backs, or any combination thereof (there are 28 variations total)
D-UP Drafts require you to boost four of your players at rates of 2x, 1.75x, 1.5x, and 1.25x
While we will be able to analyze the winning and other successful rosters this time next year as we prepare for the 2026 iteration of the tournament, how should we account for these nuances to best position our teams for success in 2025? This strategy guide is designed to help you approach these drafts by utilizing past performance data in the D-UP scoring system, historical ADP success from The IDP Show best ball drafts, and strategies in other tournaments to generate approaches to overall strategy, roster construction, and boost utilization.
If you are looking for the results of this research, here are my top 10 data-driven insights for D-UP 2025:
Do not commonly reach beyond 5 ADP.
Only employ a 6th-round “wildcard” strategy if the board has given you value in rounds 1-5.
You can build an excellent roster without trying to find the next Zach Baun or Andrew Van Ginkel.
Do not overthink your selections for the sake of uniqueness.
Linebackers would have been the foundation of most successful D-UP rosters in 2024.
Defensive linemen had slightly more upside than defensive backs, but there were more viable DBs than DL in 2024.
The market is often right on the top 8-10 DL, but is overall much better at identifying high-scoring LB.
Boost your best players.
When you have boost coin flips, go with the player who should have more up and down weeks (variance).
Boosts on “ceiling” players will be more effective when you also have good floor players on your roster.
How should your general strategy be different?
One of the best minds in best ball tournaments is Hayden Winks, whose analysis of past Best Ball Mania drafts has yielded many valuable insights for managers in these contests. While many of the insights generated from his data analyses deal with roster construction strategies through various rounds (which assumes you are required to draft/start all positions), many of his cautionary tales of what not to do have broader applicability to something like the new D-UP contest.
Two specific insights from Hayden’s data last year seem to be equally, if not more, applicable to D-UP. First, reaching for “your guys” happens all the time across draft formats. I am guilty of it as well, but reaching in best ball is a losing strategy that should be more acute in small-format best ball like D-UP.
While the ADP for D-UP is still forming, it does give a good sense of how drafts have been playing out. Winks’ article suggests not reaching more than 5 spots to get a player you are targeting and highlights the scoring hole managers typically dig for themselves. Remember that this adage comes from a contest with many more players selected; in the D-UP contest with a smaller player pool, the risk of ADP reaches grows significantly.
A second finding from Best Ball Mania draft analyses is that teams with higher negative average ADP value were consistently outperformed by those closer to even or even slightly positive. It is noteworthy that teams with the most ADP value had the widest range of outcomes, with some representing the highest scores in the tournament but others falling below the average team score. Applying this to D-UP, the reduced player pool should compress the overall range of average ADP value and potentially steepen the curve for those who can extract value in their drafts. But going only with value creates a high-risk range of outcomes.
These recommendations and past analyses create our first two general tips:
Reaching beyond 5 spots in ADP is fine on occasion, but should not be a core part of your D-UP strategy.
Teams that have scooped up ADP value in Rounds 1-5 can afford to take a reasonable wildcard in Round 6 while teams with aggressive builds should look for ADP value in their final selection.
Moving beyond ADP, another frequent topic of conversation in Best Ball tournaments is the “uniqueness factor” in that you may feel the need to construct a roster that no one else in the tournament is likely to have.
In D-UP, there are conflicting pressures around building a unique roster. On the one hand, you do not have to take major risks to build a unique roster with players such as Bobby Wagner, Ivan Pace Jr., Germaine Pratt, Kayvon Thibodeaux, Khalil Mack, and Chop Robinson, who are all currently being drafted in less than 25% of drafts. You can be unique without trying to find the next Zach Baun or Andrew Van Ginkel.
Using those two examples, Baun and AVG would have finished as LB1 and DL1 in 2024 using D-UP scoring. Looking at the 2024 best ball ADP, we can see how unlikely those picks were to succeed given the success of the DL and LB drafted immediately before and after those breakouts (none of the 8 surrounding picks finished in the top 60, and only one was in the top 100).
You can build a unique D-UP roster without having to project breakouts that few see coming.
While the above discussion presents the case for building a unique roster without taking big swings, there is an additional wrinkle that suggests that you do not need to worry about uniqueness in your D-UP roster builds at all.
As we will discuss in detail later, after your draft but before the start of the season, you will select four players on your roster to receive bonus multipliers ranging from 2.0x to 1.25x. There are 360 unique ways to distribute the four bonuses across a pool of six players. Therefore, even if you have the exact same roster as another player, you only have a 0.28% chance of having the exact same bonus allocation.
The bonuses assigned to players significantly reduce the odds of two builds that are exactly the same (even with the same rosters) so don’t overthink your selections.
How should you build your roster?
One of the elements that makes D-UP such a unique contest is that it is positionless, with your top four scorers each week counting for your score. While this adds excitement, it also can induce some anxiety as we have no historical build data to pull from to guide our decision-making.
However, by exploring player performance using 2024 data and D-UP scoring settings, we can gain some valuable insights into the benefits of a variety of builds to get you started on the right track or help you pivot if you are already a veteran D-UP drafter.
With this data in hand, I set out to answer three questions:
Given that the scoring system is similar to what we would call “big play,” what is the value of each position using D-UP scoring?
How successful is a best player available strategy vs some popular configurations of 2025 drafts thus far?
Exploring data from best ball drafts hosted by The IDP Show last season, how accurate are we at drafting certain positions (and how should this inform our 2025 drafts)?
As can be seen in the charts below, LBs dominate early and often in season-long D-UP scoring. Linebackers are the largest share of players in the top 30, 60, or 100 from 2024, ranging from 38-47%. An alternative view of the data suggests that if you hit on an LB pick, you have a good chance of having a viable (top 60) player or, at minimum, a useful (top 100) play.
The data on DL and DB were more surprising. High-end DL represent a larger portion of the top 30 than high-end DB (30% to 23%). This advantage rapidly erodes when we look at the top 60 (30% DB to 27% DL) and the top 100 (37 DBs to 25 DL).
In sum, this suggests that nailing a DL pick is more likely to net you a top 30 scorer than nailing a DB pick, but also that missing on a DL pick is more likely to land you outside of the top 60 or 100 than missing on a DB.
Linebackers would have scored very well in a 2024 D-UP contest, representing a plurality of players in the overall top 30, 60, and 100.
Defensive linemen had slightly more upside than defensive backs, but defensive backs were safer options.
Given these insights, is there a particular configuration of positions (e.g. 4 LB, 1 DL, 1 DB) that is beneficial, or does it make more sense to simply draft the best player available? To explore this question, I created six-player groupings among the top 100 scoring players from 2024 (1 to 6, 2 to 7, through 95-100). I next calculated the top four scoring players from each week for each of these hypothetical “crystal ball” teams, noting the configuration of players and season-long performance.
The distribution of positions in the 95 teams analyzed is presented below, along with the average score within each group. The three most common configurations were:
2 LB/1 DL/3 DB (14)
2 LB/2 DL/2 DB (12)
3 LB/1 DL/2 DB (12)
These findings further suggest the primacy of linebackers in D-UP. In terms of the highest-scoring configurations (occurring at least three times), the top were:
2 LB/3 DL/1 DB (1,183 pts)
4 LB/2 DL/0 DB (1,118 pts)
4 LB/1 DL/1 DB (1,085 pts)
Among the top 20 builds, teams had an average of 2.65 LB, 1.95 DL, and 1.4 DB. All of the top 20 builds featured at least one LB, but there were two builds with zero DL, and three builds with zero DB.
Successfully drafting 2-3 LBs is the clearest path to weekly season-long D-UP points with top DL as the preferred supplement and DB also often viable.
While the above highlights that some configurations tend to outperform others, it raises a secondary question of whether purposefully building a configuration is better than drafting the best available players.
Once again using the 95 “crystal ball” teams of the top 100 2024 IDP performers and applying D-UP scoring, there is a correlation of .937 between average season-long individual scoring and team scoring (using the top 4 weekly scorers). This strong relationship between the best players and the best D-UP team (displayed below with team 1-6 on the left and 95-100 on the right) is despite the wide variety of risk profiles in terms of positional configurations described above. Note: the ‘dip’ in the middle of the chart is due to some unlucky teams that only had 3 or fewer players score any points in multiple weeks due to overlapping byes and/or injuries.
Despite some positions having greater variability week-to-week, drafting the best players tends to yield the best results for season-long scoring.
Both of the previous sections make a significant assumption that we’re able to draft the best possible players when we are on the clock. While we know this is not the case, we can use data from 2024 IDP best ball drafts to determine what positions we were better at drafting last year by comparing ADP to finish.
For this analysis, I used player overall finish by Big 3 IDP scoring since this was the scoring system from which the ADP were generated. I focused on the top 68 plus DB4 and DB5 ADP since there are currently 68 players being drafted in at least 10% of D-UP drafts, and there are 5 DBs in that group. Also, when I mention LB and DL below, I am using D-UP designations since many are dual-designated on Sleeper.
Below is summary data on the success of various positions drafted in the top 70, noting the percentage of players that finished in the top 30, 60, 100, and 100+ of all 2024 IDP. I would not draw strong conclusions from the DB group since it only represents 5 players, but there is a substantial enough number of DL and LB to draw some inferences.
First, we were clearly better at drafting LB than DL in 2024, as all three hit rates are higher and the miss rate (100+) is lower. The “Top 30” column is particularly enlightening as it shows that only 7 of the 40 DL drafted finished in the overall top 30 in 2024 scoring. Conversely, 8 of the 25 drafted LBs finished in the top 30 in 2024. Still, there is optimism for those who prefer DL early, as only 3 of the top 9 drafted linemen finished outside of the top 60 in scoring, with Micah Parsons and Aidan Hutchinson missing that cut due to injuries.
We are better at drafting LB than DL, but we tend to do well at identifying the top end of the DL pool (i.e. top 10).
How to use boosts?
The final wrinkle of the D-UP contest is that you can assign scoring boosters to four of your six selections. Proper booster allocation can likely turn a mediocre roster into a contest-winning team, while poor use of boosters can derail a team that looks fantastic on paper. While the following is speculative since there is no past contest data to pull from, below I lay out the case for boosting based on your best players and boosting to embrace inconsistent players with more week-to-week variance.
My first thought about boosters is probably similar to many of yours and goes something like: “Remember that 62.7 point game by Trey Hendrickson in Week 9, how awesome would it be if that was 125.4 points!” However, some of you may also think along the lines of: “Kaden Elliss was dependable last season and had upside, so imagine if those points were doubled!”
This led me to another data experiment, which looked at these two players who would have finished 7 and 8 overall using D-UP scoring in 2024, and how much would have been gained by using a max booster on either player. To do so, I first calculated the median score among the top 100 for each week to get a sense of the floor that needs to be exceeded to enter a lineup. I then multiplied each of Hendrickson and Elliss’ scores by 2 (the maximum boost) to determine how many points over the median would have been gained by using the booster on both players. The result:
Boosting Elliss would have netted you 310.25 points against the median
Boosting Hendrickson would have gained you 304.5 points against the median
The reason why both answers were right is twofold. First, both players had phenomenal seasons, and if you remember nothing else from this, remember that boosting your best players is never a bad idea. Second, they were both high or medium-high variance players among the top 100 last season. Elliss was a “steady” performer with upside, while Hendrickson was the second most “up and down” of the group. These are the kinds of players who have the opportunity to raise your ceiling in a given week, and both are likely viable strategies.
To provide one additional example using a wider pool of players, I compare the effects of a 2x booster on Jamien Sherwood (IDP17), Jonathan Greenard (IDP19), Germaine Pratt (IDP24), and Travon Walker (IDP53). Despite comparable finishes, Sherwood ranked 56th in variance, Greenard was 92nd, Pratt was 8th, and Walker lagged in scoring but had many boom games and had the 95th most variance of the sample.
Conducting the same experiment, the 2x booster yielded:
Pratt would have netted 244.45 points
Sherwood gains 259.65 points
Greenard adds 272.15 points
Walker jumps by 225.85 points
This larger example reiterates the two previous points. Very low-variance players (like Pratt) tend to benefit less from boosts than higher-variability players (like Greenard), but this logic has limits. Travon Walker had much higher variance than Germaine Pratt, but given how much Pratt outscored Walker (nearly 30 points on the season), Pratt gains more from the 2x booster.
One final note to consider is your build. All of the above analyses assume that you will have players at median scoring to fall back on. A roster of all high-variance players may very well win, but all ceiling and no floor is a strategy I do not plan on adopting. Boosting riskier players is more likely to be successful in a balanced portfolio where you have reliable players to fall back on for the down weeks.
Players like Pratt, Roquan Smith, and Josey Jewell might not be the most effective to boost, but each scored at or above the median in 10 weeks last season, which allows down weeks by the bigger swings to not hurt as much. The full top 100 scoring players, along with their overall rank, games over median (unboosted), games where a 2x booster would have still resulted in a below median score, and season-long variance are all presented below.
Don’t overthink it, boost your best players!
When you have coin flips, go with the player who will score well but have more up-and-down weeks.
Boosts on “ceiling” players will be more effective when you also have good floor players on your roster.
My Ideal Build
To tie everything together, I thought I would highlight what a build based on the above discussion could look like with current contest ADP picking out of the 8 slot (my favorite position to draw this year):
1.08: Nick Bosa, DL, SF - 2x boost
Bosa (ADP 8.05) is the last of the DL that I have a high level of confidence in heading into 2024. While his talent makes him a 1st round pick for me, his variability makes him an ideal max boost to capitalize on his big weeks.
2.03: Danielle Hunter, DL, HOU - 1.75x boost
Had I not taken Bosa in the first round, Fred Warner (ADP 13.36) would have also been a strong consideration here, but Hunter is at the top of my second tier of DL after he would have finished 29th overall in last year’s contest. Hunter has a very similar variance score to Bosa, so using the 1.75x booster on him feels like the way to go and solidifies my week-to-week ceiling.
3.03: Blake Cashman, LB, MIN - 1.5x boost
I have drafted DL out of this spot in other drafts (this is a great spot for teammate Jonathan Greenard—ADP 20.51—if he is there), but given the build I have started with my first two picks, I wanted to grab one of the best upside LBs out there. Cashman (ADP 27.25) was one of the higher variance LBs in 2024 but had a solid floor, resulting in a 27th overall finish in D-UP scoring. I will not always be so predictable with boosts, but he is a very boostable player, so I use my 1.5x here.
4.08: Foye Oluokun, LB, JAX - no boost
Oluokun (ADP 39.27) represents a nice floor but does not have as much upside as some other LBs still commonly available at this pick. After missing 4 games early in the season, Oluokun averaged 12.667 points Weeks 8-17 with a floor of 7.5 and a ceiling of only 21.5. While an argument could be made to boost him, I will save my final booster for a player with more upside.
5.03: Kyle Hamilton, DB, BAL - 1.25x boost
I take my final medium-high variance swing in the 5th round with Hamilton, who would have finished as IDP31 in 2024 (ADP 44.18). Hamilton’s variance is more modest than you might expect, but the 30+ point game in Week 8 (without a TD) gives me optimism to make a slight projection for more booms in 2025. Lavonte David (ADP 50.48) would have also warranted strong consideration here given he outscored and had greater variance than Hamilton in 2024.
6.08: Bobby Wagner, LB, WAS - no boost
The ageless Bobby Wagner (ADP 57.86) would have finished 45th overall in last year’s contest, so getting him with the 58th pick feels like a strong way to end my build. Teammate Frankie Luvu was the high variance LB in Washington last season, posting a mark of 100.35 compared to Wagner’s 25.39, suggesting that Wagner might be a boring point getter. But that is needed to raise my floor with big swings early in Bosa and Hunter, and adopting medium week-to-week risk with Cashman and Hamilton in this build. Had I selected David in the previous round, I would have still picked Wagner here and not been concerned that I had 4 LBs since they each carry different risk profiles.
Conclusion
The FastDraft D-UP contest is one of the most exciting developments in IDP in recent memory. While we will learn what strategies are successful after the fact, drafters who enter with a plan (based on sound data) will tend to do better than those who do not. I hope that this guide gives you some things to consider as you enter the draft room that may ultimately help you perform well in the contest.
Be sure to subscribe to The IDP Show on Apple, Spotify, or YouTube.
Where are those overall idp rankings for dynasty?