Thoughts on Flexies

The Rickies are a storied institution, a particular blend of humor, tech, and chaos that has no equal in the podcasting universe. They have a healthy tradition of change over time, sometimes with little to no warning. In the spirit of this tradition, Federico Viticci has explored new territory with an unprecedented 10 Flexy picks in the 2021 WWDC Keynote Rickies.

I wanted to jot down some of my thoughts on the Flexies so that readers may better understand why I believe that Federico’s recent boldness puts The Flexies at a crossroads, and also offer a compelling argument for the correct direction. The resulting conversation, both on the show itself and in the live chat room, made it clear that there is confusion as to how, exactly, the Flexies are actually graded: by number correct, or by percentage or ratio of correct to incorrect? In a world where each host makes 5 picks, or 6 on a particularly bold day, the results are the same. Federico has entered a brave new world, and the rest of us must move forward to join him. Of course, the decision ultimately rests on the hosts of Connected, who seemed in the heat of the moment to assume the second path to be correct, but I believe that in fact the opposite is true. Let me explain.

First, there’s the letter of the Bill.

“In the case of a tie, the ratio of correct to incorrect Flexies will be taken into account” is the exact phrasing. This phrase is somewhat vague, but it isn’t vague about when it applies – it’s only “in the case of a tie”. It is a classic example of “the exception that proves the rule”, implying that more general rules apply when there is not a tie, and the general rule is that the one with the most points wins.

Second, there’s the historical precedent.

Originally, they were “non-graded picks” and so deciding on a winner or loser was unnecessary. The first stakes to be added to this portion of the game were in 2019, when the “loser" of the non-graded picks had to buy a round of drinks at WWDC. Myke and Federico actually tied in this competition, but it didn’t matter – all that mattered was that Stephen had lost.

After WWDC, the non-graded picks returned to normal – no stakes, no punishments. They were renamed The Flexies for 2020, but were still essentially the same non-graded picks.

It was only when WWDC rolled around again that issues arose. With a round of drinks on the line, it was necessary to determine the loser. Myke and Federico had both gotten 1 Flexy correct, a tie for last place. Listen to the discussion here, and it is clear what the original intent was – getting the most correct predictions matters most. As Myke is working to understand the new ratio rule, he wonders aloud what is preventing anyone from simply choosing 2 Flexies to easily win, but then realizes that Federico’s 7 picks had given him many more chances to earn points and “win easier” as a result. This would not be the case if percentage were all that mattered.

Immediately after that discussion, The Flexies as we know them today were codified and put into practice in the next event, with no substantial changes up to this day. But the details of the rule have become fuzzy over time, and Flexy grading has often been accompanied by discussions about percentages outside of a tie breaker scenario. This has not affected the results so far, as each host has picked either 5 or 6 Flexies, not enough of a difference for the number of correct picks and the ratio to disagree. Looking back to the original conception of the rule, when it was fresh in everyone’s minds, gives clear guidance as to how The Flexies were originally intended.

Third, the most important reason: the “spirit of the game”.

Needless to say, the Flexies are about flexing. Even before the current format, a Flexy pick was meant to encourage braggadocio and boldness, a place where a competitor could explore beyond the normal boundaries of sense and logic without compromising their chances in the main event.

Grading by ratio encourages a strategy of minimizing losses. There is no incentive to make more than the bare minimum of picks if 5/5 is better than 9/10, because each additional pick just represents one more chance to be wrong. It creates a system where taking a risk is punished.

If, instead, we take the rule as it is written, a more nuanced system emerges. Federico has more chances at victory, true – but he also has more at stake. If he does lose, he could be on the hook for up to a $250 donation, twice as much as the highest donation to date. In this version of the game, each host must make a calculated choice – make fewer picks and play it safe, or make more picks to increase both the chance of winning and also the consequences of losing. The system is self-balancing – boldness is rewarded, but caution is necessary.

Conclusions.

Let’s think of it another way. If a sport is being played and Alice scores 7 and misses 3, Bob scores 4 and misses 2, and Carol scores 4 and misses once, who has won? Bob and Carol may argue that their ratio of scores-to-misses are better, and this is true. They can also argue that they missed fewer attempts. The fact remains that Alice scored the most points, and that is what matters.

I modestly propose a tweak to the rules: “In the case of a tie, the first tiebreaker will be whoever has the fewest incorrect Flexy picks.” Changing the rules mid-competition is undesirable, but this is not actually a change – mathematically, this is the same as before, it just eliminates even the hint that percentages are what matters. With a clear understanding of the rules, a new era of bigger and more strategic flexing can begin.

Jason Biatek
June, 2021