Prediction markets, explained for NFT investors

The one idea

The edge is the gap between our price and the market's.

▶ Try it: drag the two prices and find the gap

Every game has a price, and the price is a probability. When LazyEdge says a team wins 62% of the time and the market sells that team at 55¢, that 7-point gap is the edge. We don't need to pick winners. We need the market to be wrong by more than the fee.

👆 Drag either bar left or right

Market: 55¢ (55%)

LazyEdge: 62%

Gap: +7 points → that's a bet. No gap → no bet, no matter how much I like the team.

🦁 Own 5 Lazy Lions → get the API

You just did by hand what my bot does every game: compare the two numbers and bet only when the gap is big enough. Hold 5+ Lazy Lions and you get the LazyEdge API: a key that lets your own bot, spreadsheet or AI agent pull LazyEdge's number for every game automatically. Then you set up your own automations with your own rules: your minimum gap, your bet sizes, your sports, alerts to your phone, or bets placed for you. That's exactly how I built DDKBetBot.

▶ How to set it up: Kalshi key, LazyEdge key, Claude

Bigger gap, bigger bet. That's the whole sizing philosophy, and it's why the units on my posts change from game to game.

The scoreboard we keep

Is LazyEdge actually sharper than the sharpest books?

Brier score · 2026 NFL · 106 games through week 8 · lower is better

LazyEdge

0.219

Pinnacle

0.232

Kalshi

0.232

Bars start at 0.18 so the gap is visible. Pinnacle and Kalshi prices taken 10 minutes before kickoff.

Brier score, in plain English: how far off your percentages were, on average. Say 70% and the team loses, and you take a big hit. Say 55% and they lose, and it's a small one. Lower means your numbers are closer to reality.

Pinnacle is the book the pros respect most. It takes the biggest bets and keeps the sharpest lines in the world. Beating its numbers is the bar.

Right now LazyEdge is ahead of both this season. I'll be straight with you about the rest, because that's the point of testing it in public:

106 games is a real sample, but not a final verdict. There's still about an 8% chance the lead is luck.
Last season the market edged LazyEdge's live model (0.180 vs 0.183). The model has improved since, and this season it's out in front.
I'll keep posting this number every week, whether we're leading or lagging.

Sportsbook → prediction market

You already know how to bet. Here's the translation.

▶ Try it: drag a price and see the sportsbook odds

Kalshi is an exchange, like a stock market for game outcomes. You buy and sell with other people instead of betting against a sportsbook. What you buy is a contract: a ticket that pays $1 if you're right and $0 if you're wrong. So the price is the market's guess at the chance: a 60¢ ticket means "60% likely."

What you know (sportsbook)	What it is on Kalshi
Odds of -150	A contract at about `60¢`. Pay 60¢, get $1 back if it hits (40¢ profit). 60¢ = the market thinks it's 60% likely.
Odds of +200	A contract at about `33¢`. Pay 33¢, get $1 back if it hits (67¢ profit).
Moneyline on the Chiefs	Buy `YES` on "Will Kansas City win?"
Taking the other side	Buy `NO` on the same market, or YES on the other team.
Spread (and alt spreads)	"Kansas City wins by over 6.5." Kalshi lists every line from small wins to blowouts, so you pick your own. Taking fewer points costs more but hits more often. That's called "buying points."
Parlay	A combo ticket across several games. You ask for a price and traders on Kalshi offer you one.
The house's cut (bettors call it "juice" or "vig")	A sportsbook prices both sides of a coin flip at -110: risk $110 to win $100. Two people betting opposite sides hand it $220, it pays back $210, and it keeps $10 no matter who wins. On Kalshi there's no house in the middle. You pay a small fee per contract instead, plus a few cents between the buy and sell price.
Cash out	Sell your contracts any time before the game ends, at whatever the market pays.
Getting limited for winning	Sportsbooks often cap or ban bettors who win too much. On an exchange you're trading with other people, not the house, so winning isn't a problem for anyone but the person on the other side.
Units	Your standard bet size. If 1 unit is $50, a 2-unit bet is $100. I size every bet in units so the bet size tells you how much I like it, whatever my bankroll is.

What you're actually buying

A contract is a $1 ticket on one outcome.

▶ Try it: HODL or WETH your ticket at halftime

1 · The ticket

Worth $1 or $0

"Will Kansas City win?" A YES contract pays exactly $1 if they win and $0 if they don't. Its price today, say 60¢, is what the crowd thinks that $1 is worth right now.

2 · Two sides make $1

YES + NO = $1

Every contract is born when a YES buyer and a NO buyer meet: 60¢ from one, 40¢ from the other. That $1 is locked up front, like escrow in a smart contract. The winner collects it.

3 · Nobody can stiff you

Fully paid for

The $1 is already sitting there before the game starts. Kalshi isn't hoping it can pay you, and it isn't a bookie who might limit you. It just releases the locked money to whoever was right.

4 · HODL vs WETH

HODL or WETH your contract

HODL to the final whistle and collect $1 or $0. Or WETH it: accept a buyer's offer any time before that, just like taking a WETH offer on your NFT. Bought at 40¢ and your team goes up 14? Someone might offer you 75¢ for that ticket at halftime.

Prediction markets, explained for NFT investors.

You buy 162 tickets on a team at 60¢ before kickoff: $99.92 with the fee. From there, it plays out exactly like an NFT you're holding:

HODL · they win

+$62.08

You held to the final whistle. Every ticket pays $1: $162 back.

Your listing sold · team rolling

+$37.42

Up big at the half, you listed your tickets at 85¢ and a buyer bought them. Like someone buying your listed JPEG. Listing is the cheap way out: only $0.36 in fees.

You WETH'd it · team up

+$27.86

Feeling good but not sure it holds? You took a buyer's 80¢ offer on the spot and locked the profit before the game ends.

You WETH'd it · team struggling

−$53.70

Down two scores at the half and you don't think they come back. You took a 30¢ offer and got about half your money back instead of riding it to zero. A partial loss, like selling a JPEG into a weak offer to save what you can.

HODL · they lose

−$99.92

You held to the end and they lost. The tickets are worth $0.

Same 162 tickets and real Kalshi fees in every case. Selling pays the fee again, the same as gas on the way out. At a sportsbook you'd only get the HODL outcomes, plus whatever "cash out" number the book decides to offer you.

Fees: think gas.

You already pay gas to move ETH. Kalshi charges a fee each time you buy or sell contracts, the same way. Here's how it compares:

	ETH gas	Kalshi fee
When you pay	Every transaction	Every buy or sell. Nothing when a winning ticket pays out.
Win or lose?	Paid either way	Paid either way
Spikes when it's busy?	Yes	No. It's a fixed formula, shown before you buy.
What sets it	Network demand	The price. It's biggest on coin flips near 50¢ and tiny on near-sure things.
How to pay less	Wait for a quiet network	Place an offer and wait instead of buying instantly: about ¼ of the fee.

Fee per contract, buying instantly

Ticket price	Fee	On $100
10¢ (+900)	0.6¢	about $5.90
30¢ (+233)	1.5¢	about $4.70
50¢ (even)	1.75¢	about $3.40
70¢ (-233)	1.5¢	about $2.00
90¢ (-900)	0.6¢	about $0.70

The fee is 7% × price × (1 − price) per contract. Cheap long shots mean more contracts per $100, so the fee per $100 is higher on long shots.

Two ways to buy · you already know both

Buy the floor, or make an offer

Buy now = sweeping the floor. You take the cheapest listing right away and pay the full fee.

Place an offer = a collection offer. You name your price, wait for a seller to accept, and pay about a quarter of the fee. My bot does this first, 2 hours before kickoff, and only buys at the listed price if nobody has filled its offer 15 minutes out.

A real receipt · $100 on a 60¢ team

162 contracts × 60¢ … $97.20
Fee (1.68¢ each) ……… $2.72
Total spent ……………… $99.92
They win → 162 × $1 = $162.00 (+$62.08)
They lose → $0 (−$99.92)

The fee is why a "gap" has to be bigger than about 2 points before it's worth betting. My bot only bets when the gap is still 3+ points after the fee.

What I built around it

Five pieces, two loops.

Follow the arrows. It starts with LazyEdge's number and ends with my followers on X, who see every play. Any of them can pick up Lazy Lions and get into LazyEdge themselves.

Loop 1

The automatic loop: moneyline bets

No taps. DDKBetBot bets the gap by itself and tells me what it did.

Loop 2

The parlay loop: my picks, my taps

Nothing gets bet until I tap ✅ and choose the amount.

The brain

LazyEdge Sports

kaneco's models price every NFL, NBA and college game before kickoff and live during play. It answers "what's the real chance?" for every side and every line.

The market

Kalshi

Where the gap gets bet. A regulated exchange, with a live price for every game, every spread rung and every combo.

The trigger

DDKBetBot

My bot makes my moneyline bets for me, automatically. It reads LazyEdge's numbers and Kalshi's prices side by side. When LazyEdge finds a gap of 3+ points after fees, it places the bet as a standard play: 1U for a 3–5 point gap, 2U for 5–10, 3U for 10+.

The remote

My phone + Telegram

My agent texts me 30 minutes before every slate: what DDKBetBot bet, plus my two favorite alt-line parlays rated 1–5 stars. I tap to bet and pick the amount. Nothing extra gets bet without my tap.

The receipts

X · @DDKing987

Every slate gets posted: the moneylines, the parlays, the units. Then a recap the next morning, wins and losses. No deleted tweets, no "tailed it" after the fact.

A slate, start to finish

How one NFL Sunday actually runs.

A Sunday has three slates: the noon games, the afternoon games, and Sunday Night Football. This whole cycle runs once for each.

2 hrs outDDKBetBot starts watching every game. Where LazyEdge's price beats Kalshi's, it puts in an offer below the current price and waits to see if someone sells to it.

30 min out · Telegram notification of bets placedMy phone buzzes with every moneyline bet DDKBetBot has placed for the slate. Right behind it: my two favorite alt-line parlays, ⭐ to ⭐⭐⭐⭐⭐.

My callI tap Parlay 1, 2, both or skip, then pick the size: from Conservative up to Scorched Earth.

5 min outBetting closes. The bot never chases a line after that.

The postOne post for the slate, "here are my plays," goes up once every bet is in.

Next morningThe recap: what hit, what missed, units won or lost.

The alt-line parlays are where moneyline edges get extra points. If LazyEdge likes a team enough for DDKBetBot to bet the moneyline, a bought-down spread on that same team is often the cleanest second bite.

Your rules, not mine

There are endless ways to use LazyEdge.

LazyEdge publishes a number for every game and every line. It's built to be filtered. My way is just one way, and it's something I keep tinkering with as LazyEdge learns every week. A few others:

The simplest

Any gap above zero, 1 unit, every time. Let the volume do the work.

The patient

Only bet gaps of 5+ points. Fewer bets, bigger edge on each.

The scaler

Size by the gap like I do: 1u, 2u, 3u.

The live trader

LazyEdge's live engine reprices every play. After a turnover, they measured the market lagging the model by about 7 points for two minutes.

The points buyer

Take LazyEdge's favorite sides and buy alt spreads down. Fewer points to cover, built into parlays.

The builder

Connect your own bot to LazyEdge's data feed (the API: a direct line programs use to pull the numbers) and give it your own rules. That's what I did.

Get in

Your Lazy Lion is the key.

LazyEdge is utility for Lazy Lions holders. Hold Lions to unlock the LazyEdge terminal: the website where you see the model's numbers next to the market's for every game. Hold 5+ Lions and you also get API access: a key that lets your own programs, bots or AI tools pull those numbers automatically. It's the same feed my bot runs on.

Step 1

Pick up Lazy Lions

Your wallet is your login.

Step 2

Link your wallet on LazyEdge

See the model's price next to Kalshi and Polymarket for every game.

Step 3

Find your gap

Bet it your way, or follow along with mine on X.

Step 4 · 5+ Lions

Automate it with the API

Hold 5 Lazy Lions and you get an API key. Plug LazyEdge's numbers into your own bot or AI agent and set your own rules: alerts when a gap opens, your own bet sizes, even bets placed for you. Step-by-step setup →

Open LazyEdge → Build your own bot → @LazyLionsNFT @kaneco Follow my bets · @DDKing987

Not financial advice. I'm a live tester sharing my own bets and my own results. Past results don't guarantee future ones. Only bet what you can afford to lose. Prediction market availability varies by state; check Kalshi's eligibility rules where you live. Brier figures: 2026 NFL regular season through week 8 (106 games); market prices taken 10 minutes before kickoff.

Findingthe Gap.

The edge is the gap between our price and the market's.

Is LazyEdge actually sharper than the sharpest books?

You already know how to bet. Here's the translation.

A contract is a $1 ticket on one outcome.

Worth $1 or $0

YES + NO = $1

Fully paid for

HODL or WETH your contract

Prediction markets, explained for NFT investors.

+$62.08

+$37.42

+$27.86

−$53.70

−$99.92

Fees: think gas.

Buy the floor, or make an offer

Five pieces, two loops.

The automatic loop: moneyline bets

The parlay loop: my picks, my taps

LazyEdge Sports

Kalshi

DDKBetBot

My phone + Telegram

X · @DDKing987

How one NFL Sunday actually runs.

Bigger gap, bigger bet.

There are endless ways to use LazyEdge.

The simplest

The patient

The scaler

The live trader

The points buyer

The builder

The rules I hold myself to.

Real money

Every bet posted

Posted before the result

The scoreboard stays up

Your Lazy Lion is the key.

Pick up Lazy Lions

Link your wallet on LazyEdge

Find your gap

Automate it with the API

Finding
the Gap.