rzalkikar 🦇🔊

Building @crunchtimeodds | Computer Science @nyuniversity | Statistics @uva | Unit tester | NLP NLU NLG | ΞVM | Ξthfinancier

[rzalkikar 🦇🔊 @rzalkikar]

First-author paper "Measuring Social Biases in Masked Language Models by Proxy of Prediction Quality" was accepted to ACL 2025 Main Conference. #ACL2025NLP

[ACL 2027 @aclmeeting]

🚨 Just a reminder! 🚨 The Main Conference Papers for ACL 2025 are now available! 🤩 Dive into the cutting-edge research being presented in Vienna this July/August. Explore the program here: https://t.co/Yz2f3TlWRj #ACL2025NLP #NLProc #research

[rzalkikar 🦇🔊 @rzalkikar]

@ETHbenezer @696_eth “almost” 👀😆

[Betonchain (🚫 Discontinued) @Betonchaingg]

Hey anon, come experience the future of onchain Fantasy contests 📲 https://t.co/2SKBbt96xE

Join our All or Nothing Sunday @NFL 🏈 pool live on the Sepolia testnet

✅ Trustless
✅ Accessible
✅ Mobile-friendly
..and a whole lot more 🔥

[rzalkikar 🦇🔊 @rzalkikar]

Best Overall 📈

🥇 Pittsburgh Steelers +$3.65
🥈 Cleveland Browns +$2.89
🥉 Jacksonville Jaguars +$2.66

Worst Overall 📉

🥇 Carolina Panthers -$6.00
🥈 Denver Broncos -$3.43
🥉 New York Giants -$3.38

[rzalkikar 🦇🔊 @rzalkikar]

For the 2023 @NFL season so far, betting $1 per game:

Best when Underdogs 📈

🥇 Pittsburgh Steelers +$4.65
🥈 Houston Texans +$2.42
🥉 Jacksonville Jaguars +$2.07

Worst when Underdogs 📉

🥇 Carolina Panthers -$6.00
🥈 New York Giants -$3.83
🥉 Green Bay Packers -$3.00

[rzalkikar 🦇🔊 @rzalkikar]

Best when Favored 📈

🥇 Philadelphia Eagles +$1.75
🥈 Miami Dolphins +$1.42
🥉 Detroit Lions +$1.15

Worst when Favored 📉

🥇 Washington Commanders -$1.70
🥈 Buffalo Bills -$1.52
🥉 Denver Broncos -$1.43

[rzalkikar 🦇🔊 @rzalkikar]

This analysis uses aggregated odds and game data from the @Betonchaingg API (and Chainlink node)

[rzalkikar 🦇🔊 @rzalkikar]

More NHL Betting Alpha 😎 (NFA)

When line totals vary across sites going into games, Bookmaker and JustBet UNDERS are the safest moves (based on past seasons' data). They tend (significantly) towards the push instead of the over in comparison to other sites.

[rzalkikar 🦇🔊 @rzalkikar]

When only considering games with line discrepancies:

- Hertiage has the highest average line at ~5.74
- All of the other sites have average lines between 5.6-5.7
- Bookmaker, JustBet, and Heritage have the highest standard deviation in line total of ~0.52

[rzalkikar 🦇🔊 @rzalkikar]

#BUIDLing

[Dray11 @Dray11__]

If you're bullish on strong, diverse, knowledgeable, alpha-packed and Ethereum value aligned communities but you're not part of @EVMavericks then you're doing it wrong.

[Betonchain (🚫 Discontinued) @Betonchaingg]

Happy Fri-YAY, Players! 🎉

Dive into the weekend with NFL and NHL fantasy contests on our mobile dapp – there's a pool for everyone 🏈🏒

Exciting surprises await our early beta testers who help us fine-tune the magic of our decentralized protocol🔮

[rzalkikar 🦇🔊 @rzalkikar]

@PolygonVentures @playtrophy @superlayerio Love to see it!

Gotta mention @Betonchaingg for significant progress in onchain betting and fantasy sports.

Now, users can act as decentralized liquidity providers to collateralize bet positions, indirectly hedge their bets, and get rewarded for it!

[rzalkikar 🦇🔊 @rzalkikar]

Shifting the paradigm and leveraging blockchain's strengths to transform expectations.

Accessibility, Information Symmetry, Trustlessness.

#onchainfantasy #onchainbetting

[Betonchain (🚫 Discontinued) @Betonchaingg]

The Betonchain mobile DApp for onchain Fantasy Sports just leveled up BIG TIME!🏆

Don't miss out on the action 📲 https://t.co/2SKBbt96xE

[rzalkikar 🦇🔊 @rzalkikar]

Results are from binarized outcomes & predictions where:
if total score line difference > 0 ➡️ 1
if total score line difference < 0 ➡️ 0
if total score line difference = 0 ➡️ push (considered accurate in the metrics)

[rzalkikar 🦇🔊 @rzalkikar]

I fit different ARMA(p,q) models on the difference btw lagged total game score and the ou line for #NHL games

Best results on my prediction data (w pushes as correct):
RMSE: 2.195
Accuracy: 0.641
Precision: 0.649
Recall: 0.925

Here's what I did ⏬ (NFA ofc)

[rzalkikar 🦇🔊 @rzalkikar]

Normality of residuals 😎