June 2, 2013
Stats vs. The Will To Win on the Family Feud
In 1980, the Royals and the Phillies faced off in the World Series, on Family Feud. In the World Series, of baseball, also, but on the World Series of Family Feud especially. The video of it, which a minor-league scout tipped me off to this weekend, exists. It's right here:
But 1980 was before sabermetrics, before analytics, when we had to go by what our eyes, and what the scouts, told us. Now sabermetricians tell us we can know more than that. These guys wield their calculators and their spreadsheets and promise to definitively answer the question: Who was really the best at Family Feud, in 1980, among 10 players chosen from the Royals and the Phillies?
There is, of course, the most basic statistic that we've always started with: Most Points. There are folks out there who will tell you Points is the most important stat in Family Feud -- the whole point of the game is to score points, obviously. So if we just start with points, here's how our 10 players rank:
That's not unreasonable, but the stats guys say it doesn't take into account where each player guessed in the guessing order -- a factor that greatly influences how many chances each player had. If we take into account that context, via Points Per Guess, the order changes a bit:
Unser's talent, and Splittorff's situational advantages, become clearer. But that's far too blunt a measure for the stats crowd. Not all Turns offer the same opportunity -- a guess with just one answer left on the board is unlikely to yield as many points as, say, the guess each player gets to make early in the round. So now we're getting into a more statistically complicated measure: Points Percentage, or the percentage of available points the player converted, where the first person to guess at the puzzle has 100 possible points to guess at.
Splittorff drops further, as he had the second most available points to start with, though -- as you can see by his converted points percentage -- he's also a very good hitter. Just, some argue, overrated. Wathan's lack of talent becomes very clear, as he had the most available points -- more than 300 -- and converted just five of them. By counting stats or by rate stats, he's very poor.
Similarly, Wathan is very poor if we measure by a more subjective measure, Errors. Errors, of course, are those answers that either draw a laugh from the audience, an audible "bad answer," or were non-answers at all, in which the player merely stays silent until the buzzer rings.
Again, it's Wathan and Schmidt leading the league in being awful; the consistency between these measures is encouraging, suggesting all these advanced metrics are honing in on something true.
A traditionalist would look at Hits -- guesses that showed up on the board, regardless of point value -- or Big Hits -- guesses that were the top remaining answer on the board -- the stats guys say these measures provide messy jumbles of names that reflect usage and random fluctuation more than true skill.
Whatever the measure, the stats are clear: Wathan is terrible. But then the Bonus Round starts, and the Royals put Wathan up first, and he wins the doggone thing all by himself; they didn't even need to bring Dennis Leonard out for the second half, because Wathan cleared 200 points like a true winner.
Look, I'm pro-stats. The thing about it, though, is that stats don't tell you about the human element. I have nothing against nerds, believe me. I’m just, I’m so tired – not just in Family Fued, in every game show – of the new infatuation with stats and numbers and using this stuff to compare players. The best thing about Family Feud is people can answer the same in the first round and the last round, but you’re trying to tell me that you’re gonna take someone other than John Wathan in the Bonus Round? If John Wathan's guessing a 2 percent conversaion rate, I'm taking John Wathan, compared to a guy converting 40 percent. That's something you can't put in the numbers. You know, the new way to compare people, it’s been a hit, or whatever people want to call it. And some of it’s true; I’m not saying all of its wrong. But the great thing about Family Feud is how people handle pressure and how people handle adversity, and there’s just no way that numbers can do that justice.