Artificial Intelligence is the ability of a computer or computer-controlled robot to perform tasks that are commonly associated with the intellectual processes characteristic of humans, such as the ability to reason. Why can’t we do this to increase our volume in Best Ball? Move over ChatGPT, Dall E and Copilot, there’s a new AI in town; introducing Autodraft Intelligence.
Best Ball is a weighted lottery that is inherently risky, not all tickets are created the same and only a select few will come away with life changing money. If I play the Powerball, I have a 1 in 292.2 million odds of hitting the jackpot and the only way I can improve those odds is through buying more tickets. In Best Ball Mania 5, I have a 1 in 672,672 odds of winning the jackpot. Unlike the lottery, we do not have unlimited entries to improve our odds as we can only buy 150 tickets. In Best Ball, the most powerful thing we can do is create a volume of tickets that have an advantage over our opponents and there are almost always more contests to be entered. A well constructed ticket will be worth more than the entrance fee post draft as multiple drafters will construct teams without tournament winning upside. The key is to balance the need to get as many tickets as possible and building intelligently constructed teams.
Time is not infinite and most people do not have infinite resources ($) to play Best Ball. If you are able to draft the volume of teams you are planning for this season manually, you should do that! Especially with a tool like the Draft Hacker. I sometimes fall into a category of people who want to draft more teams, have the funds available to do so, however don’t have the time available. When autodrafting, we are likely to bulk draft multiple teams at the same time based on the same exact strategy. Our objective with Autodraft Intelligence is to create directionally accurate teams with tournament winning upside. By defining what you are attempting to accomplish, it will help dictate your inputs. Let’s jump in! We’ll talk about positional limits, creating rankings and site specifics, then a few potential pitfalls.
Positional Limits
The first step is determining your positional limits; these should be determined based upon what specific or similar roster build(s) you want and your preferred team structures. Moderate structures will struggle to be autodrafted well; be specific and live towards the extremes. What do I mean by this? Dual RB is a moderate strategy and one that will be hard to create; Zero/Hero RB and Hyper Fragile RB are more extreme and easier to create.
I’ll commonly set limits for a 2-7-7-2 or 1-7-8-2 roster build on Underdog. On Drafters and DraftKings, limits of 2-8-8-2 or 2-7-8-3. These are intentional as I want to draft Zero/Hero RB teams with a high likelihood of an Elite QB and/or Elite TE. If you wanted to do a Hyper Fragile RB team, you may choose something like 2-4-10-2.
Quarterback – Do you want an elite QB or not? If you do, set your limit for 1-2 depending on your personal risk tolerance. If you do not want an elite QB, 2-3. I personally want an Elite.
Running Back – If you are building Zero/Hero RB teams, set the limit at 7-8, if you are building Hyper Fragile RB teams set the limit at 4 RB.
Wide Receiver – If you are building Zero/Hero RB teams you’ll want 7-8 WRs, if Hyper Fragile RB teams set the limit at 10-12 WRs. Underdog allows a maximum of 10 WRs on a team.
Tight End – Do you want an elite TE or not? If you do, set your limit for 2. If you do not want an elite TE, 2-3. I personally want an Elite.
Creating Rankings
This will be the hardest part and it’s going to take a bit of effort, especially your first time going through this process. Using a rankings set from a content site won’t yield the result you are looking for and using adp definitely will not. Your ranks must be custom and designed specifically for autodrafting. Based on the Team Structures you chose above, we need to group positions by tier, clustering positions more than normal manual draft rankings. The further down the draft board the more you want to move players up (and down) as drafters in general become less tethered to adp as the effective draft capital lessens.
Running Back/Wide Receiver – For Zero/Hero RB, move CMC through Gibbs down 3-4 picks each, then everyone after this tier down 24 picks all the way to the players going around adp 105 (Post Bowers).
- – What if you want Devon Achane? He is inviting you to get drafted a Dual RB start if you do not tank CMC’s rank.
- – You can’t have it both ways. Achane can’t be in your top 25 if CMC is your 1.01 and you do not want to start Dual RB.
- – If you want Achane, CMC should be in the 1.04 range.
Hyper Fragile RB drafters, move the running backs in the first four rounds up, hoping to draft three. Then move every running back after this range down by 24-36 picks, hoping to land your fourth and final as a mega value. This will get your running backs and wide receivers in a good range.
Quarterback – Now focus on Elite QB, moving targets up by 2-3 picks. Have 3 targets at most. This will bring these players into range but not too far ahead that you would typically get multiple. Attempt to avoid multiple Elite QBs on the same team. If you set up for Solo QB, you can be a little more aggressive with your ranks. If you are anti-Elite QB, drop all the quarterbacks until the range you want to start drafting them by 48 picks.
Tight End – Elite TE is very similar to Elite QB, with the exception that I have more target players and once draft capital lessens (around Round 5), I am okay with a Bully TE build in a 2 tight end. If I draft Andrews plus Kittle, I am not going to cry. The price is steep enough on Laporta that I am not likely to move him up at all (plus Kelce on DK), I am actually more likely to drop Laporta behind adp to focus on WR early.
Once you have your positional tiers reorganized, it’s time to move on to stacking.
Team Stacks – Select around five teams you want to target to team stack, move players from those teams to the top of their positional tier. Pick around five teams you want to avoid, plummet those players to the end of their tier, ideally avoiding plummeting Week 17 matchups with your target teams. An example of this may be the New York Giants, bringing up Malik Nabers in your early WR zone, then Devin Singletary as a Zero RB target.
QB Stacks – Consider moving quarterbacks up one QB tier if the QB is on one of the five teams you are trying to stack, especially if you are not creating rankings that you will likely get an Elite QB. Maybe you love Anthony Richardson and decide to move him up 12 spots. You’d then also consider moving Pittman, Mitchell and Downs up in ranks.
Week 16 / Week 17 Game Stacks – If a team is playing against a team you are targeting, move them up in the tier to just behind the targets in the positional tier. Wait, how about combining the NYG and IND examples from the Team Stacks and QB Stacks sections above. Another example of this, if you are targeting the Miami Dolphins stack, D’Onta Foreman may be at the front of his RB tier.
Site Specifics
Underdog – This may be the easiest site to use AI on as you can modify positional limits mid-draft. Let’s say I want to draft 2-7-7-2’s with an Elite QB. I will set limits of 1-7-8-2 and enter 3-4 drafts, then do something else for 25 minutes. After the 25 minutes, I’ll do a mid-draft spot check to ensure they’ve all passed the Elite QB tier of the draft. If so, update the limits to 2-7-7-2 and go back to life. This mid-draft spot check is a huge advantage.
For the Marathon contest, Week 16 and Week 17 Stacks don’t matter anymore more than Week 1, allowing that step to be skipped.
DraftKings – No special notes.
Drafters – This is the worst site to end up with Bully QB on, be more selective on the quantity of early QB’s you are targeting. Also, Week 16 and Week 17 Stacks don’t matter anymore more than Week 1, allowing that step to be skipped.
What are the pitfalls?
Risk Tolerance – Risk tolerance is a massive factor in my decision to draft Best Ball teams and use Autodraft Intelligence. You should never play more than you can afford to lose and always gamble responsibly. I personally have the mindset that I am going to lose every cent I play, this allows me to free my mind, detether from ADP and draft the best team I can. I understand the risk I am assuming. I believe by using Autodraft Intelligence, I can build teams that are better than a substantial amount of my competition, thus still having a lottery ticket weighted in my favor even if it isn’t as good as if I drafted manually. If you aren’t willing to sacrifice a little bit of quality for the added quantity of teams in your portfolio, AI is not for you.
Spot Checks – I personally do spot checks to ensure what I am targeting is occurring. I am not simply entering all the drafts at once and walking away. If I was to do a batch of 15 drafts, I would probably review them after drafts 3 and 10. If I get a player in an area that does not fit my structure, I will analyze what happened and attempt ranking modifications to address moving forward. As mentioned above, I don’t mind mid-draft spot checks to toss a player or position in the queue to get you back on track with a structure or stack.
Bye Week Duplication – It is very possible you end up with both players at a onesie position with the same byes (ex. Mark Andrews and Dalton Schultz), this is the cost of doing business.
Draft Slot – You can’t control your draft slot; if you enter 3 drafts at once, you may get the 1.01, 1.06 and 1.12, your ranks need to be structured in a manner where all three draft slots can find success.
Condensed Ownership – With the heavy modification of rankings away from adp your ownership will condense at both an individual player level and the combinatorial ownership amongst your largest stands. Let’s say you were going to auto draft a batch of 75 teams, you may consider tweaking the ranks every 25 drafts to combat the condensation.
Pomeranian Autodrafts
The primary reason I was inspired to write this article was the recent Pomeranian contest on Underdog. I used Autodraft Intelligence to draft 10 teams that I wouldn’t otherwise been able to draft. I landed on 11 QB Stacks, which I find impressive considering seven of teams were Solo QB. I also was able to hit many of my targeted game stacks.
QB Stacks
Week 16 / Week 17 Game Stacks
Favorite Team
