Most days between April and September start with me waking up the kids, letting the dogs out, watering the plants, drinking some coffee, and then …

… and then spending about an hour going over PlateIQ at Rotogrinders, mostly looking at pitcher splits and batter ISO, though groundball and flyball percentage as well as barrel percentage get plenty of play, as well as K%, both for pitchers and batters. Walk percentage slips in there also, as well as a few other data points.

From there, I’m checking out Kevin Roth’s OVERCast app for the weather, then it’s off to Handigraphs, where I can more easily break down current form for the pitchers and batters. And all the while, I’m — pardon my Boomer status here (I’m fully Gen X, but after these next few words I’ll lose my credibility) — writing things down in a notebook to help organize my thoughts.

The above doesn’t include the bi-minute detours to Fangraphs and Baseball Reference, as well as some X and regular ol’ Google search work, and whatever else I stumble upon.

All told, the above take upwards of an hour. I’ve got some music playing in the background, it’s usually pretty fun, and by the end of it, I have a very good idea of how I want to allocate my DFS dollars for the day.

This is a far cry from the way I used to do my daily research, be it for baseball or NBA or NFL action — my two other DFS loves.

The way I used to do it was a little bit quicker: I’d be apprised of the news of the day by reading Rotoworld blurbs and then build my lineups. Took about five minutes.

And while I don’t keep day-by-day, sport-by-sport detailed records of my DFS play like I used to, I also know this: I’m not winning nearly as much, or as often, as I did 10 years ago. Sure, the games have gotten harder and the competition fiercer, but there’s been something nagging at me for a while now: Am I hurting myself by crunching every number? Is it possible that too much information is a bad thing? It feels counterintuitive, but …

Finding the edge

Adam Levitan, one of the co-founders of Establish The Run, which is the E.F. Hutton of fantasy sites (“When E.F. Hutton talks, people listen”) has a term for my former behavior: Phone shitter bros.

As in, people who make their DFS lineups on their phone whilst toileting.

It’s not a compliment.

And while Levitan knows that’s far from the best way to go about building quality lineups, he does recognize the challenge presented by the glut of data that has taken over the DFS industry.

“I think in the age of everything being so data-driven and model-based, what actually wins the most money is understanding and digesting the data,” he told me. “But also having contrarian takes based on soft-skill stuff.”

In short: Know the data, know what’s important in the data, and then use that data against your opponents.

This idea crosses over into regular ol’ sports betting, as Capt. Jack Andrews explains.

“Successful betting is about finding something the market hasn’t factored in, or at least hasn’t factored in enough,” Andrews said. “Conversely, if you and the market disagree, you should ask yourself ‘What does the market know that I don’t?’ That can be harder to answer, but making that exploration is a great way to avoid losing money.”

In today’s sports betting and DFS marketplace, it’s all too easy to find, understand, and use the data. The harder part, as Levitan and Andrews point out, is properly using the data against your DFS opponents or the house.

It’s all reminiscent of the Battle of Wits scene in The Princess Bride. Watch the clip. Replace “iocane powder” with “advanced metrics” and you’ll get the gist.

In short: Sometimes overthinking can be deadly.

Of course, I’m not the first person to come to this idea. Cue the academic studies.

Jam, NBA, and the horses

Back in 1973, a psychologist named Paul Slovic rounded up eight professional horse handicappers. He let each one ask for any five pieces of information they wanted on each horse, then had them predict the outcomes of 40 races. They hit at about 17%, comfortably above the 10% or so you’d get from blind guessing.

Then he gave them 10 pieces of information. Then 20. Then 40.

Their accuracy? Stuck at 17%. Didn’t move. A few of them actually got worse with more data. 

But as the information piled up, the handicappers got more and more confident. Same hit rate, but way more swagger. They weren’t any better at picking horses. They just felt like they were. 

So maybe I’m the modern version of those handicappers, drunk on barrel rate and ISO, dead certain I’ve solved the slate, and hitting at the exact same clip I would’ve when hitting Rotoworld blurbs for five minutes on the toilet.

Then there’s the jam.

In 2000, researchers Sheena Iyengar and Mark Lepper set up a tasting booth at a fancy grocery store. Some days they put out six jams, some days 24. The big spread drew the bigger crowd. But when it came time to actually buy something, the shoppers who saw six jams bought at roughly 10 times the rate of the ones who saw 24. Too many options didn’t light people up. It froze them. They walked off empty-handed, and the few who did choose felt worse about it afterward.

More choice, more paralysis, less satisfaction.

Then there’s this one, out of the University of Bath. Two researchers built NBA betting models and tested two ways of choosing which one to actually bet with. The first way picked whichever model called the most winners correctly. The second picked whichever model was the most honest about its own odds, meaning when it said a team had a 60% chance, that team really did win 60% of the time.

Both models were basically dead even at picking winners. But the honest one made about 35% profit over a season, and the cocky one lost about 35%. Same data, same games. The loser just trusted its own numbers more than it had any business doing, so it kept “finding” edges that weren’t really there and betting into them.

Sound like anybody you know? Some 7 a.m. guy with a notebook, maybe?

Carry on

So do I quit the morning routine? Burn the notebook, cancel the subscriptions, go back to building lineups in five minutes between Rotoworld blurbs?

Nope. I like the new way too much. The coffee, the music, the hour of feeling like a quant. 

What I’m going to try to quit is the thing all three studies are actually warning about. Not the data, but the certainty. That little voice that swears the hour of work bought me an edge, when most of the time it just bought me the (sometimes) false confidence to bet bigger on the same coin flip.

So: Know the data. Know what’s worth knowing in the data. And know when you don’t actually know anything.

To quote Levitan again: Easy game.