It's Better to be Lucky than Good

Thoughts on my Best Ball Mania finals team, and the extent to which we can make our own luck

Last night was a Merry Christmas. First and foremost it was a pleasure to be back home in Winnipeg, MB, see a few dear friends and my family over the course of the 23rd until now. But Santa left a little extra treat under my tree this year: a golden ticket to the Best Ball Mania Finals.

For someone like me, this tournament is both the fantasy football event I spend the most time thinking about, writing about and drafting, and also the tournament that most torments me in terms of how to judge anyone’s performance - especially my own.

The fundamental paradox of this tournament is that the easiest way to measure success is advancement in the first 14 weeks of the tournament. Two of twelve teams in each draft group advance to the “playoffs,” so the average number is 16.7%. If advancement optimization was the sole goal of the tournament it would be quite easy to approximate how well you played in a given year: check you advance rate, compare it to the average, and maybe factor in some context such as injuries or lack thereof to your most drafted players, and come to a conclusion on the success of your process.

But as I’ve written about at length - while in a vacuum you would prefer to advance as many teams as possible, I earnestly believe that optimizing for advance rate is somewhere between sub-optimal and insufficient, and that advance rate makes a very small difference in terms of the expected value of your portfolio as a whole.

Contingent on advancing into the playoff stage, your odds of *making* a Best Ball Mania final are 1 in 256: (you have to advance out of two, independent 16 person pods). Therefore, if you advance exactly the average rate of 25 teams out of a max entry of 150, you would advance *one* team to the final just 9.8% of the time. Even by posting a 50% higher advance rate than the field, your odds of placing a team in the finals improve by just under 5%, and the effect on your chance of actually winning the tournaments are 0.01% higher.

The odds of winning this tournament are so faint in any event that if there were no opportunity cost to optimizing around advance rate, this net increase would be beneficial even if unlikely to be realized in each given year. My argument has been that like with any DFS tournament, the best lineup come playoff time is one that blends a high degree of correlation, with low-owned players relative to their weekly ceilings, and structural leverage. Because there are 16 weeks of football in between your eventual 441-person final and your draft, it’s quite difficult to actually build those into each team in a predictable manner.

However, I wanted to try, and wrote about how I was trying to do it this summer. Luckily, and I truly do mean luckily, the team I advanced has captured some of the concepts I wrote about in that piece and addressed in other content. Also fortunately, I have quite a clear recollection of my draft in this case and why the team fell the way it did.

But with all that said, this is *NOT* a piece in which I tell you that I solved best ball, found the edge and magically captured it. The reality is that if you agree with me that Best Ball play should be optimized around advancing to the finals, and ultimately seeking 1st place, then we can basically never capture the strength of one’s process in a statistically significant way because the relevant sample is too sparse. Your expected rate of finals teams every year is 0.098. In other words, if you advance two teams to the BBM finals each decade, you’re performing well. That means one holding penalty, fumble, interception or bounce of the ball can determine whether you performed above or below your expectation. It’s an uncomfortable realization that is difficult for me to mentally process both as a player, and especially, as someone trying to sell my advice on a game for money. But I think it’s an uncomfortable realization we have to live with.

When I first started this substack, I said the most important trait in a good fantasy player was to make choices which mitigated the impact of bad luck, and maximized the impact of good luck. At the heart of that notion is a thesis that we cannot solve luck out of the game, but instead need to marry our acceptance of luck’s place in our game with the strategic levers we can pull to will it in our direction.

With all that said, let’s go through my finals team in BBM and I’ll walk through some of the decisions I made, some of the luck that I created, and some of the luck that was gifted to me. (I will note for completeness that I also advanced one team into the Pomeranian 2 Finals)

Best Ball Mania IV Finals Team Breakdown

Here’s the squad! As you can see below, the core of the team is a DAL-DET Week 17 stack featuring Jared Goff, Jahmyr Gibbs, Sam LaPorta, Ceedee Lamb and Brandin Cooks. Around that are several different 1:1 secondary correlations, but no full second game stack (especially with Jimmy Garoppolo no longer starting).

Here is the draft board from when I selected this team, in which I’m picking from the 11th slot.

This draft was *if I recall correctly* from late July or early August, and was indicative of some strategies I was implementing at the time from that pick range, right from the start.

As I alluded to in my leverage article I linked earlier, there is some risk with certain game stack pairings that if the players involved in your stack perform well (as you’d expect them to in order for you to advance to the finals) you’re taking on a ton of combinatorial ownership risk with other drafters not only having the same successful players but paired in the same fashion. This was a particular focus of mine with respect to the Dallas - Detroit game because of the Ceedee Lamb - Amon-Ra St. Brown pairing available to drafters at the R1-2 turn.

Pairing Amon-Ra St. Brown and Ceedee Lamb would bring on a *TON* of overlapping ownership when you consider that (A) they likely have high advance rates conditional on you having advanced a team to the finals with then as your first two picks, (B) they are already drafted at the same turn leading to natural pairing of the two, and (C) thoughtful drafters will break ties toward St. Brown conditional on having selected Lamb for the Week 17 game stack equity.

To be clear, I drafted some St. Brown - Lamb teams, but I always envisioned a scenario where if these two offenses were keys to both the fantasy season as a whole (as is fairly necessary from your first two picks), and of the Week 17 landscape, there would likely be multiple hits from them and no certainty with respect to who the highest-scoring or most impactful pair would be. For that reason, I went out of my way to draft several Lamb teams without St. Brown and vice-versa, especially given the alternative pairings available in Lamb-Gibbs and St. Brown-Pollard. Further down the line, I was prioritizing any mix of Sam Laporta, Brandin Cooks, David Montgomery, and Jake Ferguson I could with each of these starts indiscriminately.

A common response would be that you could pair them all together on one team. The benefit of a Lamb-St. Brown-Gibbs-Cooks-Laporta team for instance would be the chance of counting different iterations of the same stack in each playoff week, in hopes of being able to rely on a low-owned member of a high-owned game in the final week, that effectively leverages your own player who got you there. I think that’s valid and I did dabble in these teams as well, but the reason I preferred the thinner stacks in this case was that you’re immediately making life hard on yourself if your win condition in a 441-person final is benefitted by a failing score from one of your first two picks. These mega-game stacks were something I did more commonly with stacks involving an array of cheaper pieces in the mid-late rounds more than this Detroit-Dallas game.

So this is a very long explanation of why I chose not to select Amon-Ra St. Brown at the 2.02. Davante Adams was my common pairing for these teams because (A) he was a draft priority of mine in a vacuum, and (B) he had (at the time) his own common game stack pairing in Jonathan Taylor. Picking one from each game allowed me to build two stacks out simultaneously in hopes of hitting on the best pieces in two different games with my top draft capital investments: it’s thinner in terms of hitting on a ceiling but in my opinion slightly more optimized for a true first place ceiling. Another aspect of this pairing I enjoyed was the optionality in QB pairings which allowed me to play more greedily the rest of my draft.

I immediately had Anthony Richardson and Dak Prescott in play at the 8-9 turn (often), and opened up Jared Goff at the 11-12 turn. In this draft I continued building in a fashion that kept opening up more mid-late QB stacks as I drafted.

At the 3-4 turn I picked up Gibbs as I intended to do as soon as I made the Lamb choice, and then paired Gibbs with Breece Hall, who was a player I would take in Round 4 in almost every build type when he was available, but also opened up possible stacking scenarios with Aaron Rodgers and Deshaun Watson.

At the 5-6 turn my preferred route was to open up another stack with QBs generally available in the Rounds 9+ range and pair a Seattle WR with Diontae Johnson, or George Pickens should he make it to the 7/8 turn. Neither Steeler made it to me at those spots so I went with Jaxon Smith-Njigba, knowing I could still add Jaylen Warren or Pat Freiermuth late if needed. I almost certainly would have taken Michael Pittman Jr. in Round 6 had he made it to me, but with him gone I opted to open up *another* Round 9-12 QB option with Jordan Addison and a possible pairing with Kirk Cousins.

The Addison pick was the first ADP “reach” on the team but is reflective of something I’ve written about over the last two years in terms of how I draft my “targets.” Addison - along with Breece Hall - were two players near the top of my exposure list and players I wanted integrated across all different types of teams. Only taking Addison at his ADP (which for most of the season was around the 6/7 turn) meant I could be over the field on both Dallas-Detroit stacks and Addison, but have very few teams with both. I prioritize correlation quite a bit, so I will rarely have 30+ % exposures to players. But with someone like Addison who I drafted in the 20% range, I wanted to make sure he was layered across several different iterations of teams I was constructing to make sure I maximized my chances of combining my target players with builds that wind up carrying the day. Again, in this case I would have likely taken Pittman for the correlation had he lasted another pick, but after he didn’t there was no clear cut choice, and it meant taking advantage of the chance to add Addison to this build.

Round 7 was essentially an auto-Brandin Cooks pick with the stack I’d built up, and D’Andre Swift was a case of another player who I would just take every time they were available in that round (in his case, round 8).

The next turn was probably the most interesting and where the theme of luck really takes hold. For the first pick I was locked into a TE and had two correlated options with David Njoku and Pat Freiermuth: this is one of the few cases where clearly the other option would have worked out better season-long, and wouldn’t have disrupted my playoff run out. I marginally preferred Freiermuth and preferred Geno Smith to Aaron Rodgers in terms of the stack I was setting up.

The 110 pick is fascinating to me because it was in my opinion, (A) a very sharp pick from a process perspective, (B) my worst pick from a results perspective, and (C) absolutely essential to me getting to the finals of this tournament.

I took Rashaad Penny, who has been a totally dead roster spot all year. I was a big fan of taking shots at multiple backs in the Miami, Philadelphia and Chicago backfields given their ambiguity and price points on the same team, with outs to either securing one major hit with 2 roster spots, or finding two hits at different times: possibly even within the playoff weeks. Clearly the Penny bet didn’t work, and this team would be much better off had I used my 4th RB pick on Jaylen Warren, De’Von Achane or another reasonable option I was in on at this stage. But ultimately with the way I was building this team, the RB I was picking was always going to be Penny here.

The real decision was whether to take Dak Prescott to complete my 3:1 stack of the Dallas-Detroit game, or whether to continue deferring QB. The way I saw it, a Dak pick here was entirely reasonable, but I had a chance to build a more unique roster by deferring. I had already locked up both Lamb and Cooks, and the Pollard drafter took Josh Allen in Round 3. Once the player behind me - one of the only two other teams without a QB yet - passed on the position, I thought my odds of Prescott falling two rounds beyond his ADP were actually pretty reasonable, opening up a chance at a highly valuable Prescott team. In the event he didn’t last, I was now set up to chase late stacks with Geno Smith, Kirk Cousins, Jared Goff, Aaron Rodgers, Kyler Murray (along with the Eagles RBs), Kenny Pickett, and Jimmy Garoppolo. I figured it was worth the risk to see what the room would leave me.

As it happened, almost every one of my QB targets - including Prescott - went between my picks, with the Pollard/Allen team taking an early 2nd QB in Dak Prescott. Nonetheless, Jared Goff was available and I was perfectly happy playing the other side of the game in this case based on what the room dictated. For my next pick I continued to build out my contingency plans for my secondary stack with Rondale Moore in hopes of building a Kyler Murray, Rondale, Swift, Penny game stack.

Naturally, prior to my next turn Kyler Murray was gone. I considered pulling the trigger on Kenny Pickett here but felt Sam Laporta was especially necessary on this team to keep building out my primary stack, and having been locked into a 3QB build, I wanted to add an Indianapolis piece opposite Davante Adams and (presumably) Jimmy Garoppolo with Alec Pierce.

Once again, Kenny Pickett goes before my next pick and all of a sudden my once plentiful options for a secondary stack at dried up. Ironically, by this point in the draft I was exasperated! (it’s why I remember this team so vividly) I had felt totally in control of the draft for the first 10 rounds, setting up several avenues to land the plane while still making aggressive decisions to set myself up for a high-upside Week 17 build. And yet, multiple irrational drafters taking 3-4 QBs in the first 12 rounds had spooked the room into taking QBs I had earmarked. I had already pegged Jimmy Garoppolo as my QB3 at my last turn, but at this point needed to backdoor a third stack from scratch. My typical favoruite options for this were C.J. Stroud (with Tank Dell and/or John Metchies) and Matthew Stafford (with Puka Nacua and/or Tutu Atwell). Thankfully I chose the second option on this team and wound up making this a Puka team rather than an Atwell team. There’s no particular reason why I did this, I just did, and those two players were completely essential to my advancing this team into the playoffs and through Week 16.

I do wish this was a mid-late August team, as my Round 17 pick would have almost certainly been Kyren Williams if it had been. I drafted him aggressively to close my draft season, but mostly stayed away prior to that unsure if he was the clearcut backup over Zach Evans. Instead, my last RB is Jeff Wilson Jr. who saw six touches last week so may not be a truly dead roster spot, but is as close as it gets. Such is life.

All in all, it’s an interesting draft to look back on: so much of the stuff that led to this team was very intentional and thoughtful and it’s great to see it work out. At the same time, taking Rashaad freakin’ Penny - an entirely dead roster spot - instead of Dak Prescott is the sole reason I have Mathew Stafford - and by proxy Puka Nacua on this team. Without the Rashaad Penny pick, I’m never in the finals.

Similarly, my team to this point would have been unquestionably better had I simply taken Amon-Ra St. Brown over Davante Adams at the 2.02. It would also project better this week! But at the same time, writing right now I am *extremely glad* Adams is on this team instead of St. Brown. Every member of the Cowboys-Lions stack will be heavily owned given the success of these teams this year, especially Ceedee Lamb and the four primary Lions skill players. There will be such a high number of Lions-Cowboys stacks in the final that even if the game goes nuclear and produces the BBM winner, the win condition for whoever takes it home likely depends more on who you have around that stack.

I would rather hope that the Lions offense goes through Gibbs and Laporta this week, while Adams crushes as a low-owned option outside that game than be content to capture the entirety of the Lions offense but have little else outside of it.

In terms of my chances to take this home, I see my win condition something like this:

1. The Dallas-Detroit game stack is the must-have element of the week as the game meets or exceeds its slate-high over-under and several high-owned pieces it a ceiling performance

2. That game stack’s primary benefactors are Jared Goff, Sam LaPorta, Jahmyr Gibbs and Ceedee Lamb, with Dak Prescott, Amon-Ra St. Brown and David Montgomery in particular falling slightly behind.

3. At least of two of Davante Adams, Puka Nacua, and D’Andre Swift produce a high-end ceiling game with lower ownership relative to slightly better projecting options due to YTD results or in Puka’s case, a lower overall draft percentage

4. One of the beta-pairings in my game-correlations (Alec Pierce, Rondale Moore, Pat Freiermuth, Brandin Cooks) has a spike week, ideally as part of a game-wide eruption

The odds of taking home even a top-50 prize remain quite low, but regardless I thought this would be a fun and helpful exercise to walk through a draft that wound up having a degree of success. And yet, even after doing so to the tune of 3300 words, I remain unable to truly determine just how much credit I should take for each element of this team.

I was extremely lucky that my team stayed almost entirely healthy, with just 2-3 dead roster spots (depending on Jordan Addison’s status) entering Week 17. I was lucky to have had teams take away picks I wanted to make, forcing me into a Round 18 Puka Nacua pick that was nowhere on my radar until the 15th round. And I was lucky that Christian McCaffrey was stuffed at the 1 yard line, a week after Jaxon Smith-Njigba caught a game-winning TD, to advance my teams in the last minute of the last game for two consecutive weeks.

Maybe my luck will run out this week. But here’s hoping it doesn’t. And if I’ll give myself any credit with respect to this team, I think this team is in a pretty good spot to take advantage of any luck that comes my way this weekend. I look forward to reporting back after this week’s results!