Gambling experts, pt. 1

Song: Charlie Musselwhite, "Cristo Redentor"

Are betting experts any good at what they do?

These days, nearly all talk about gambling I see on TV and the internet is sponsored by one of the sportsbooks. How good is all this sponsored advice?

There are quite a few shows that are just about gambling, but more common are ad reads from Youtubers or sportscasters who are sponsored by sportsbooks, but aren't really focused on gambling. These appear on-air in the middle of a game, or an ad break in a Youtube video.

Picks from sports announcers do terribly, as the Youtube channel Foolish Baseball has documented in their wonderful video, Baseball is Ruining Gambling.

As a numbers guy, it's baffling that anybody would follow these obviously sponsored picks at obviously juiced lines, given by obviously casual gamblers, but some people are taking them, because the sportsbooks keep paying for the ads.

Gambling is a social and parasocial activity now, another thing you do on your phone when you're bored that sort of feels like interacting with other humans, but isn't.

Some gamblers want to be on the same side of the bet as their favorite YouTuber, who give their favorite picks as a part of an ad read. The apps also allow you to follow people, and take the same bets they take. It's yet another one-way online relationship.

Other gamblers take bets to feel more connected to their team. Announcer parlays are invitations to take a financial interest in the game that you're already watching, not necessarily because you think the Brewers play-by-play guy is secretly a betting wizard. The baseball announcers don't seem to have much of an interest in gambling, or being touts. They're not there for our wholesome national pastime, gambling on sports, they're true sickos who are only interested the disreputable game of baseball. Putting together some half-ass parlay for the promo is part of their job. It's just another ad read. It may as well be a local roofing company or a personal injury lawyer.

Sportsbooks advertise because it makes them money in the long run. These companies seem pretty well-run, if nothing else. They want to sponsor people who are good at bringing in customers with a high Customer Lifetime Value -- people who will lose over and over again for years, making back the cost to acquire them as a customer many times over. That's it. That's the game. Why would they sponsor people who give good advice about gambling, or good picks?

Do people care whether gambling experts are actually good or not?

Some guys talk about gambling for a living. They discuss sports from the perspective of people who are gamblers first, and sports fans second. Everything's an angle, or a trend, or a bad beat. At the extreme, athletic competitions are interesting because betting on them is interesting, not because sports themselves are. These guys are both living and selling the gambling lifestyle, which I talk much more about in the book:

A parasocial relationship with a guy selling picks or talking about gambling on a podcast causes guys to want to form social relationships around gambling. They're Gambling Guys now. Which leads to an endless parade of dudes complaining about their parlays online, and, I would wager, annoying the heck out of their significant others. "It's a whole lifestyle, Sherri! Of course I had to get my tips frosted! I'm a Gambling Guy now!"

It's all imaginary. An imaginary relationship with a betting guru in the form of a "hot tip". An imaginary relationship with the sporting event or player in the form of a bet. An imaginary relationship with reality itself in the form of the rationalization about why the "hot tip" didn't win. An imaginary relationship between winning and skill.

Being a sports fan is already ridiculous enough.

Poking the bear, a bit

The Ringer is a website about sports and pop culture that has evolved into a podcasting empire. I like a lot of what they do, and I especially appreciate that they publish great writing that surely isn't profitable for the company. For the most part, I can just enjoy their non-gambling content and ignore that it's subsidized by gambling.

But they've done as much as anyone to normalize sports betting as a lifestyle, and deserve an examination of that. The Ringer wasn't worth a bajillion dollars before gambling legalization, back when they were doing MeUndies ad reads.

The Ringer has an incredible amount of content that's just Gambling Guys Talk Gambling with Other Gambling Guys, around 10 hours a week of podcasts, by my count. The Ringer's flagship show is the Bill Simmons Podcast, which usually devotes at least a couple hours a week to discussing which bets Bill and his pals think are good. (Previously satirized in Cool Parlay, Bro)

The site also has an hour long daily podcast about gambling, The Ringer Gambling Show, and several other podcasts that regularly discuss betting. As far as I know, all of their sports podcasts feature gambling ad reads, even the ones with hosts that clearly find gambling distasteful.

Several members of the Ringer's staff are full time Gambling Guys now. They talk about their addictions for a living, which must be nice. I've been talking about my crippling data science addiction on here for months without a single job offer.

Are these guys good at their job, though? (Are they good at their addictions?) The Ringer is currently having a contest betting on the NFL between five different NFL podcasts, four of which primarily cover sports gambling, which they're calling The Ringer 107. Here are their results through Week 9:

Every single one of the five teams is losing money against the spread. They don't make it clear on the website, but some bets are at -120 vig, so the records are even worse than they look.

It might be bad luck, right? I've written at length about how having a true skill level of 55% doesn't guarantee actually winning 55% of the bets.

Based on this data, it's extremely unlikely the gambling pros at the Ringer are just unlucky. I simulated 5 teams taking the same number of bets as the Ringer's contest so far. If every pick had a 55% chance of winning (representing picks by advantage players), then 99.5% of the time, at least one team does better than all five of the Ringer's did.

Even at a 50% winning rate, same as flipping a coin, the simulation does better than the Ringer 91% of the time. Based on this data, you'd be better off flipping a coin than listening to these gambling experts.

The easiest way to know they can't do it

If I talked about gambling for a living and was demonstrably good at it, I'd want everyone to see the proof. These guys almost never post their actual records over a long stretch of time, though. They will crow about their wins, and offer long-winded explanations as to why their losing picks were actually right, and reality was wrong. But it's hard to find actual win-loss records for them.

Last season, The Ringer's Anthony Dabbundo didn't appear to keep track of his record betting on the NFL against the spread at all (example article). This year, he has, so we know he's gone 37-37, and 20-25 on his best bets, despite some of them being at worse than -110 odds. His best bets have had a return of -16.7%, about the same as lazy MLB announcer parlays, and almost 4 times worse than taking bets by flipping a coin.

What odds would you give me that the Ringer goes back to not showing his record next season?

To be clear, I don't think Dabbundo is worse than the average gambling writer at gambling. I've never found a professional Gambling Guy that is statistically better than a coin flip. Dabbundo's job is writing/podcasting the little descriptions that go along with each of his picks. His job isn't being good at gambling, it's being good at talking about gambling. There is no evidence that he can predict the future but plenty of evidence he can crank out an hour of podcast content every weekday that enough people enjoy listening to. Predicting the future isn't really the value-add for these sorts of shows.

The curse of knowledge

I think it's significant that by far the worst team in the contest is the Ringer NFL Show, which is primarily not about gambling. It's a show by extreme football nerds, for extreme football nerds. Once in a while I listen to it while doing the dishes, and I'll go like 20 minutes without recognizing a single player or football term they're talking about. They are all walking football encyclopedias, as far as I'm concerned.

If gambling really were a matter of football-knowing, they should be winning. They aren't, because betting isn't a football-knowing contest.

I can see how being a true expert might make somebody worse at gambling. Every bet is a math problem, not a trivia question. A bet at a -4.5 spread might have a positive expected value, but a -5.5 spread have a negative expected value. People who are good at sports gambling can somehow tell the difference. Like life, sports gambling is a game of inches.

The folks on the Ringer NFL Show know the name of the backup Left Tackle for every team in the league, but I don't think that's really an advantage in knowing whether -4.5 or -5.5 is the right number for a particular game. Most of the cool football stuff they know should already be baked into the line, or is irrelevant. Sports betting is a lot more like The Price Is Right than it is like sports.

A great chef might know practically everything there is to know about food. They might know what type of cheese is the tastiest, how to cook with it, and so on. But if they went to the grocery store, that doesn't mean they would notice that they had gotten charged the wrong price for the cheese, or that they could get it 20% cheaper at another store. That's a totally different skillset and mindset.

People on these gambling shows spend several minutes explaining each of their picks, listing lots of seemingly good reasons. I think that having to give a good reason for each pick forces them into going with what they know, not considering that all the obvious information and most of the non-obvious information they have is already reflected in the line. Needing to be seen as an expert could lead them to pick the side with slightly less value on it, because they want to make a defensible pick. It's much better to have a good sounding reason for making a pick, and losing, than it is to have no reason for making a pick and winning.

In general, I think every bet has a more reasonable side and a crazier side. Anybody prognosticating for a living has a disincentive to pick the crazier side if they're going to have to explain the bet if they happen to lose. It doesn't if there was actually more value than the Browns. Losing on the Browns and then having to live with being the guy who bet on the Browns is two losses at once.

The Mathletix Bajillion, week 10

I figured I should take a crack at it. The season is mostly over, so it will be a small sample size, but let's see what happens. At the very least, it will force me to publish something at least once a week for the next couple of months, and if things go bad, work on developing the shamelessnesss of someone who gets paid to predict the future, even though they can't.

I'm going to make two sets of picks, one based on a proprietary betting model I created, the other purely random, based on rolling dice. At the end, I will reveal which team is which, and see how they do against the Ringer's teams.

I am an extremely casual fan of the NFL, so little to no actual ball-knowing will go into these picks. I probably know less about football than the average person who bets on football. I don't think that's really a disadvantage when the football experts are going 18-27 on the year.

I will take all bets at -110 odds or better. No cheeky -120 bets to goose the win-loss record a bit, like some of the teams in the Ringer competition have done. But I will also shop around between sportsbooks and take the best lines I can find, by checking sites like unabated, reduced juice sportsbooks like lowvig and pinnacle, and betting exchanges like matchbook and prophetx.

For the betting exchanges, I will only take spreads where there is at least $1000 in liquidity -- no sniping weird lines. Part of what I'm trying to show is that bargain shopping, which has nothing to do with football knowledge, can make a big difference. So I will shop pretty aggressively. Even if I lose, I will lose less fake money than I would at -110, which means I don't have to win as often to be profitable. I'm pretty confident I can beat the Ringer pros on that point at least.

Sources will be noted below.

Team names come from the 1995 crime drama Heat, starring Tom Sizemore, in accordance with the Ringer house style.

Picks were made Friday evening.

The Neil McAul-Stars

• ATL +6.5 -108 (fanduel)
• NO +5.5 -109 (prophetx)
• SEA -6.5 -110 (prophetx)
• PHI +2 -109 (Rivers)
• LAC -3 +104 (prophetx)

The Vincent Hand-eggs

• DET -7.5 -105 (prophetx)
• PIT +3 -109 (prophetx)
• TB -2.5 -110 (fanduel)
• CHI -4.5 -105 (fanduel)
• CAR -5 -110 (harp rock casino)