ESPN analyzes Donaghy debacle: Which games did he allegedly affect?

More than a decade after Tim Donaghy was sentenced for conspiracy and wire fraud, the disgraced NBA official is back in the news.

Last week, ESPN published an in-depth report detailing allegations that Donaghy not only tipped off gamblers with insider information, but also most likely made in-game calls with the intention of fixing the point spreads of games he officiated.

From their article:

Donaghy favored the side that attracted more betting dollars in 23 of those 30 competitive games, or 77 percent of the time. In four games, he called the game neutrally, 50-50. The number of games in which Tim Donaghy favored the team that attracted fewer betting dollars? Three.

In other words, Donaghy’s track record of making calls that favored his bet was 23-3-4.

If one assumes there should be no correlation between wagers and the calls made by a referee, the odds of that disparity* might seem unlikely. And they are. When presented with that data, ESPN statisticians crunched the numbers and revealed: The odds that Tim Donaghy would have randomly made calls that produced that imbalance are 6,155-to-1.

Wait...what? Didn’t we already know Donaghy fixed games!?

The ESPN report created a wave of publicity but, to me, it didn’t feel like the earth-shattering event the World Wide Leader likely hoped for. The crucial takeaway from the article — that ESPN’s analysis can identify games Donaghy likely fixed — felt like something we already knew.

I’ll admit that before reading the ESPN article I assumed it was accepted by all parties that Donaghy altered outcomes of some games for gambling purposes (that is not accepted by everyone, in actuality). Matt Moore from the Action Network, a guy writing for a freaking “gambling-adjacent” website, felt the same way:

I can’t decide whether it’s a function of my being older, my working in a gambling-adjacent media enterprise or the general direction of Western society over the past three years that caused me to ask myself as I read the article: “Wait, didn’t we already think he fixed games?” If you had asked me 10 minutes before the article was published, would I have parroted the same line about his not fixing games, just betting on them? Did I always believe he had fixed games? The report didn’t produce any “Wait, what?!” moments. Instead I was left saying “Well, yeah, that make sense.”

What do ESPN and the NBA disagree on?

Shortly after ESPN published the article the NBA reminded the public that their analysis agreed with an FBI investigation that said there’s no evidence Donaghy made in-game calls to aid his betting choices:

The ESPN Article attempts to revive this old story. Unfortunately, it is replete with errors, beginning with its statement that the Pedowitz Report “concluded that Donaghy, in fact, did not fix games.” The Pedowitz Report made no such conclusion. Rather, the investigation found no basis to disagree with the finding of the FBI and the U.S. Attorney’s Office that “[t]here is no evidence that Donaghy ever intentionally made a particular ruling during a game in order to increase the likelihood that his gambling pick would be correct.” ESPN ignores this important distinction.

What can ESPN do to strengthen their argument?

ESPN hasn’t done themselves any favors in this dispute with the NBA, I’d argue, but there is one easy way to improve their argument. They mention 40 specific games that they identified as possible fixes and then discuss the detailed analysis of those games, including re-watches by officiating experts.

Next, we pulled game videos for all 40 games and employed a researcher with an extensive background in officiating to watch them closely, logging all of Donaghy’s and his fellow referees’ foul calls. (Of those calls, 2.6 percent could not conclusively be attributed to a referee and were excluded from the study.)

But the authors never actually tell the public what games they’re speaking about — the 40 games remain a mystery to anyone outside of Bristol.

Strangely, they have an italicized vignette(?) earlier in the article narrating Donaghy’s travel and activities for a few days and identify several games their informants claim were fixed (complete with box score!):

Donaghy on New Year’s Day in Charlotte calling 14 fouls against the Bobcats, five against the Timberwolves; the Wolves covering

If they’re willing to identify those games via anecdote, why won’t they give the public a list of the data they used for their statistical analysis?

My main takeaway from this is that ESPN could strengthen its argument by releasing more details of their analyses so that other statisticians and officiating experts could also review the data.

Additional notes:

Henry Abbott’s coverage of the Donaghy scandal for True Hoop provides some great retrospective reading. A few highlights:

• In 2010 Haralabos Voulgaris, possibly the most well-known professional NBA gambler ever, broke down the influence a referee can have over a game and also identified a game that he believes Donaghy attempted to fix and failed. Insightful stuff.
• Henry Abbott’s 2009 review of Donaghy’s book remains a must-read.
• An e-mail between Abbott and a gambling expert from 2007 with some interesting circumstantial evidence.
• Should anybody listen to Donaghy?