
NEW BRITAIN, Conn. -- The date was December 21, 2007. The place: Manhattan College's Draddy Gymnasium. A small crowd, its numbers diminished by school break and holiday shopping, watched the homestanding Jaspers take on the Pepperdine Waves in just another pre-conference tuneup. Those who did show up took in a good old-fashioned back-and-forth free-for-all, as Manhattan won 80-79.
But when the boxscore rolled out of the printer, it made no sense. The total minutes didn't add up to 200 for either team. According to the record, the participants hadn't amassed 159 total points, the cumulative number shown in lights up on the scoreboard. Basketball, like the universe, requires balance and order: just as the earth takes 365.242199 days to orbit the sun, so must missed shots equal the number of rebounds. But this sheet was pure chaos theory.
There was a negative chain reaction. SportsTicker and Stats Inc. called the press row phone repeatedly, asking where in the world the box score was. In the postgame press conference, the coaches couldn't properly review the performances as they shared their thoughts on the game. Reporters scrambled to fill holes in their stories; how many rebounds did Tyrone Shelley have? And in the eyes of the NCAA, with no true and accurate statistical record of the action, this game didn't happen.
It was Manhattan sports information director Michael Antonaccio's last game before taking a similar position with Red Bull New York of the MLS, and this was exactly what he didn't need. I caught up with him after the presser; his brow was covered with sweat and his eyes were wild. He held an open laptop in one hand and a stack of looseleaf papers in the other. "We're going to be here until 2 in the morning," he said, out of breath. "We have to have to reconstruct this game from the tape."
"Why, what happened?" I asked.
As he ran down a hallway, he called out behind him. "Stat Crew screwed up."
---
At every Division I college basketball game, there's an official scorer who sits with a straight view of the timeline, and who wears the zebra-striped shirt of ultimate authority. He or she keeps track of points and fouls in a coil-bound book, and every player who checks into the game must do so with the scorer's permission. The official scorer is the fourth most important person in the arena, as far as upholding the rulebook is concerned.
But the statistical authority, the creators and guardians of the game's official record, sit elsewhere at the scorer's table, huddled around a laptop. Every point, rebound, block, foul, steal, turnover and assist are recorded by Stat Crew, a standard software program used from coast to coast. It is on this record that boxscores are built, and no matter if the stats are individual or team-based, standard or tempo-free, here is their genesis and basis. Player accomplishments and admonishments are entered, one at a time, by an uncredited and anonymous army of work-study students, graduate assistants and SID's. And for one evening, I joined that group.
On an unseasonably warm New England night in a dying year, I arrived at Detrick Gymnasium on the campus of Central Connecticut State University. The Blue Devils were scheduled to play former Northeast Conference foe Maryland-Baltimore County, which left for America East back in 2003. These were all knowledge-items that SID Tom Pincince reminded me of before he introduced me to Mike Bono, a recent CCSU graduate who'd been helping with Stat Crew for years. Mike, in turn, introduced me to the series of signs and symbols that record the fastest-moving, most statistically intensive game on earth.

Stat Crew teams tend to work in pairs, with one typing away at the keyboard while the other -- the "spotter" -- calls out the action in code, which the typist enters into the system.
"If No. 22 hits a 10-foot jumper, that's
J-22 good," Mike explained. "If there's an assist, it's
A, but we usually just say
from. If there's a miss and a defensive rebound, that might be
J-22, RD-14." If it's an offensive rebound, we'd leave out the
D, so if there's a putback, it'd be something like
J-22, R-22 L-22 good. There's a
P for tip-in, which automatically adds an offensive rebound, but we don't really use that here. Multiple actions can be the hardest to keep straight, but the toughest parts for someone new are the fouls and subs."
I was quickly realizing, once again, how little I know about basketball.
"I know this is information overload, but don't worry about the codes too much," Mike said. "It took me a while to get a handle on them, and 'jumper,' 'layup' and 'rebound' work just fine. Especially because you'll be working with Vinnie tonight, who could do this by himself. He taught me everything I know about Stat Crew."
At 7:20 pm, 10 minutes before game time, Vinnie DiCarlo breezed in, took the seat in front of the laptop, cracked his knuckles, and started inputting the starting lineups with the speed of a piano maestro. If there's any such thing as a stat legend, Vinnie qualifies. He's been using Stat Crew for nearly two decades, and has done sports information director stints at places like Southern Connecticut (Division II) and Princeton. When the No. 13 Tigers upset fourth-seeded UCLA in the 1996 NCAA Tournament, Vinnie was sitting courtside doing stats for radio.
L-32 good from 30.I was positioned on the right side of the Stat Crew "booth," with Vinnie in the middle and Mike closest to CCSU's bench. On my right was
"Jasper" Joe Arnone, the greatest and best public address announcer in all of Hoops Nation. In addition to my spotter duties, I was tasked with handling Joe.
"Joe's going to turn to you a lot and ask you who just scored," Vinnie said. "But the most important thing is that if he starts getting on the refs, pull him away from the microphone. Physically pull him away. He's such a Blue Devil, and he gets really into it. He forgets the mic is on."
The game began. Alongside the flurry of action on the floor, there was a cascade of letters and numbers in the Stat Crew booth.
Y-24 good... Y-15 good... Y-23, R-15, L-15, R-15, L-15 good!... L-24 good!... T-3... J-10, RD-0... L-15-Z good!And before long, I realized I had completely lost track of the score. I had no idea that UMBC was shooting the lights out to a 60 percent tune, or that sophomore big man Robbie Jackson was compiling the game of his life. This view of the game was all trees and no forest.
"When I watch a game on TV now, I'm always doing Stat Crew in my head," Vinnie said during an early media time out. "I remember back in 2007, when we won the conference and went to play Ohio State and Greg Oden in the NCAA's. My wife asked me afterwards how the game was, and I couldn't remember. There was this run in the second half..."
"Honestly, I couldn't tell you what offense either team is running," Mike chimed in. "Defense? Zone, man to man, I have no idea."
Stat Crew makes you unlearn everything that was drilled into your head about basketball at an early age.
Don't follow the ball. To watch a basketball game, one treats the ball as an instrument in a larger context, a drop of fuel in the engine. But recording actions at a granular level requires a pinpoint, persistent focus. To properly perform Stat Crew duties, one must develop a specific and sharp tunnel vision; all that matters is the rock and who has it. The other nine players move around the floor in an irrelevant blur.
And this is why I was an initial failure as a Stat Crew spotter, and why Vinnie's expertise and Mike's experience saved this UMBC-CCSU game from becoming a disaster on a Manhattan-Pepperdine scale. I said
DR instead of
RD, signaling a dunk and a rebound instead of a defensive board; that's how I always write it out on paper, and putting the adjective last was all Spanish to me. On breakaways, I forgot to add the
Z that would automatically add a fast break bucket to the points-in-the-paint column. I tuned my focus too narrow and forgot to add assists; usually, I just wait for the PA announcer to tell me who dished the dime.
But Vinnie would patiently correct me, and Mike was keeping track on paper in case the action moved too fast for Vinnie to capture on the keyboard. Finally, with one second to go in the first half, a UMBC player hit a long buzzer-beating jumper. The other two were blocked out by the backside of CCSU head coach Howie Dickenman and the ref on the sideline. With my unobstructed view, I was finally able to be useful.
"Did you catch that, Kyle?"
"J-10 good from 31."---
Fifteen years ago, Stat Crew software (recently purchased by CBS) was the survivor of a death-struggle with a competing program called StatMan. Like VHS over Beta, or Windows over MacOS, the winner of the war may not have been the superior product, but the one that appealed more to the market.
"With StatMan, you had a window in which to enter the shorthand text of the play," Vinnie recalled. "It gave you an incredible level of control, it was like a stick as opposed to an automatic. The problem was that if you entered something wrong, it would throw an error message, and you'd have to enter it all again. Meanwhile, the action on the court wasn't stopping."
StatCrew destroyed StatMan and its command-line accuracy by introducing on-screen modes, rotating screens, and real-time input. By the time Y2K rolled around, StatCrew was the industry standard, and its system of codes became the foundation for all college basketball statistics.
But StatCrew never really and truly entered the 21st Century. The program has remained fundamentally unchanged. For starters, it's still written for
DOS, the Disk Operating System introduced to consumers in 1981 by IBM and finally put to bed in 1995, when Windows graduated from a pretty graphic interface for DOS to its own independently coded operating system. There is no Windows 7 version of Stat Crew, and some newer laptops have to contain a DOS emulator in order to run it.
But like the air traffic control system, Stat Crew is too firmly entrenched to be replaced by a superior product. It's too standardized for a version 2.0, which would likely shock current users into mass protest.
"They tweak it every year," Vinnie said. "Hooking it into the Internet was a big step, so that ESPN and SI.com and Yahoo can have live stats. But it's DOS. You can only take it so far."
I asked Vinnie how he'd improve on Stat Crew.
"The core of it, the inner workings, those will probably never be changed now," he replied. "But I still think there are a lot of possibilities in terms of input. Maybe a touch screen interface? Some kind of graphics, anything."
But Stat Crew is still waiting for a
Windows 3.0 moment to revolutionize it. In the interim, the arcane system of codes and letters remain, passed down from one generation to the next. And there are plenty of people who don't pick up the torch.
"As you might imagine, there's a lot of spotter turnover," said Vinnie. "It's not like there are a lot of out-of-work SID's out there to draw from."
The second half moved quickly. As with many games with no television coverage, regulation time was complete in just over an hour and a half. But after UMBC tied the score at 66 with 38 seconds to go (
L-23 good from 10), CCSU dribbled its way into an empty possession. There was nothing for us to record except for a last-second formality:
T-3, S-21.When overtime began, "Jasper" Joe clicked on the microphone and intoned with his inimitable class and grace, "Let's have a big hand for both teams and the fine game they're both playing." The CCSU crowd, well-trained to be utterly dignified through years of having the best public address announcer in Hoops Nation, complied.
The first overtime period, with the game tight and on the line, passed with the same level of emotion as the very first minutes.
L-10 good... Y-3 good... L-32 good. Y-32, RD-15... And that was it. Seventeen seconds of clock ticked away without a number or a letter, and the game was tied again at 76 after five extra minutes.
With a
Y-3 good, CCSU shifted into the odd digits, making a third overtime unlikely. And then, an
L-32 good broke the game, and the Blue Devils prevailed by an 89-86 score. With a few keystrokes from Vinnie, the final box score came rolling off the printer. The official scorer came over to the booth, quickly crosschecked the sheet against his handwritten book, and the boxscore was ready to be faxed to local news outlets, Stats Inc. and ESPN. As the CCSU staff collapsed the bleachers and the UMBC players gathered in the lobby to catch their bus, the numerical record of this completed game zinged around the wires and fell into order in giant databases, including the NCAA's data mine in Indianapolis. An hour after the final buzzer, the raw stats would trickle down to Basketball State and be automatically sorted and processed into a
tempo-free boxscore.
And as for Vinnie, Mike and me, we'd long since passed in our unsigned handiwork:
the book.I'm not sure how this new perspective will affect the way I watch basketball, if I'll need some time to adjust from watching a game from courtside through a telescope. But later that night, stopping in a McDonald's parking lot on the way back to Rhode Island, I cracked open my laptop to check scores. I saw an e-mail alert from BBState's statistics provider that the boxscore from the
Alcorn State-Central Michigan game would not be available for an indeterminate amount of time. For once, I understood.
And then I noticed that there was a late game on the online video streaming. Northeastern was playing Kent State in the Cable Car Classic out on Santa Clara's home floor. As I watched the tiny figures run and leap in 480-pixelvision, I found myself whispering under my breath.
Y-3, RD-2... home, 22 in for 3... T-3, S-32, L-23 good!This is the third in the ongoing Kyle Doin' Work series. Previously: Assistant Coach, Assistant SID.