The Sabermetrics of Beer Pong
From the fundamental and elementary to the resplendently sabermetric, baseball drowns in numbers. RBIs, OPS, HR, WAR, wRC+. Baseball’s nature, with its 1-on-1 hitter/pitcher matchups, indepenent and discrete events, and easily quantifiable data, makes it a veritable treasure trove of ones and zeroes (and fours and twos and nines). In that way, it lends itself to number-loving statheads (like me) and we eat it for breakfast, lunch and dinner. But we don’t only love baseball. We love football and basketball and politics and cooking and basket-weaving.
In many of these fields of interest, statistical data is lacking, to say the least. Not only does that leave the pocket-protector crowd wanting, it creates unverifiable truisms and axioms (think “Defense wins championships”) that haven’t stood up to the test of science. Beer pong (or Beirut) is certainly no exception.
It lacks the storied history of football or politics, but Beer Pong is rife with unverified axioms, the most infuriating of which is “No seriously – I get better when I’m drunk!” There’s so much yet to be learned about the numbers in beer pong – an objective truth needs to be sought after. People are trying, but we aren’t there yet. What I really need is data: someone to collect loads and loads of information about which shots people are taking, whether they’re going in, what starts to happen when the booze sinks in, that kind of stuff. That…isn’t going to happen. Yet. So this is all purely theoretical. For now.
But a boy can dream. If I had all the data in the world, here’s what I’d analyze:
The Basics: These’ll be the building blocks – what we work with before we get down to business.
Hit% – simple enough. The number of cups you hit divided the number of shots you take. Period.
2nd Hit% – how often do you hit the shot immediately following a make? A measure of streakiness, which plays heavily into beer pong lore. Also relevant are 3rd Hit%, 4th, etc.
Balls Back Hit% – how does your Hit% change when you’re shooting on a balls-back turn?
Partner Stats: How you interact with your partner and how her/his shots affect yours.
Partner 2nd Hit% – how often do you make any cup after your partner’s made a shot? Could be called balls-back percentage.
Hit% after Partner Miss – same as above, only after they miss.
Same cup Hit% – depends on the rules you play with. If you aren’t playing pull-cup, how often do you and your partner hit the same cup?
1st shooter Hit% after previous turn Hit/Miss – put simply, does the previous turn’s partner success have any bearing on the first shooter’s success? I’m guessing no.
BAC Stats: Imagine, for this set, that we had a non-intrusive machine hooked up to the player that determined how drunk they were at every instant.
Hit% by BAC – critical statistic. Because of the binary nature of Hit% and the linear nature of BAC, I’d do it like a histogram, setting up various “bins” of BAC levels, e.g. .005-.01, .01-.015, all the way on up. What percentage of shots do you make when you’re buzzed? Around the legal limit? Really f***ed up?
Come to think of it, you could do this for every statistic on here. Just pop the number into the BAC bucket/bin. What would be interesting here is how different stats change differently by BAC. Do you get streakier when you get drunker? Do your partner’s hits make your shots better when you’re four or five games deep? The world will never know.
Arrangement Data: All about which cups are hit, not so much whether they’re hit. For our purposes, we’d number the cups 1-10 just like in bowling. Also like in bowling, the 7-10 split sucks.
Cup N% (Ordinal) – Which cups are most often hit in what order? For instance, the middle cup of a full rack (number 5, aka the bitch cup) is probably hit first more than any other cup (or maybe it isn’t. That’s why we’re doing this) – call it 60% of the time, cup 5 is hit before any other. I’m betting 7 (back right) is hit first with the least frequency. Maybe 5% of the time. Similarly, we might guess that cup number 1 (front) is hit second 45% of the time. This gets convoluted when reracks come into play.
I don’t know how I’d format the statistic, but another interesting phenomenon to observe might be how the previous cups made determine future cups being made. Like, if someone just made cup number two, I bet the cups nearest it are most likely to be hit next.
Contextual: How a player does with regards to the “score” of the game
1st cup Hit% – very simple: how often does player A make his first shot?
Last cup Hit% – how long does it take for a player to hit the last cup, on average?
Hit% v. Cup differential – for instance, maybe player A is at his best (highest Hit%) when his team is up by 4 cups, but player B thrives under pressure and performs best when down by 2
So that’s what I have for now. It’s not complete, but neither were Bill James’ Baseball Abstracts or the Bible. If all this does is get your head thinking in the direction of next-level beer-pong analysis, then I’ve done my job.
— Tim Badmington