Tabitha Garwe, PhD

This is a large, single-center, cohort study demonstrating improved shock index after administration of blood prior to emergency department arrival. Further study will be helpful in determining which patients stand to benefit the most from transfusion in transit. #pediatrictrauma @NatalieADrucker @ksutyak @UTH_Pedisurg journals.lww.com/jtrauma/abstra…

@ryanmouquegolf Thanks Ryan, this is super useful. I feel like I should pay you:)

Have you ever watched pros on TV hit those wedge shots into the green that bounce once or twice & then STOP immediately? Well, below is everything you need to hit this shot & control the distance it goes. We are looking for a low launching, high spinning wedge shot that produces predictable flight, spin & distance control every time. Let’s dive in

It's PAR 3 CONTEST day at The MASTERS! If you want to become a better wedge player, here's the ULTIMATE wedge thread for you to "master" shots between 50y - 125y They'll be using this technique in the Par 3 Contest 🧵👇

@AnthonyCrudele 💯

When I really think about success, in trading or in life, it always comes back to one thing: Making the right decision in real time. Not later. Not after the fact. Right there in the moment when it is hardest. Because most of the time, you already know what the right move is. You just do not make it. You go for what feels good right now. You react instead of pausing. You repeat patterns you have already paid for. I have done it more times than I can count, knowing it was the wrong move as I was doing it and dealing with it later. That does not make me unique, it makes me human. The goal is not to be perfect. It is to do that less. To slow things down. To sit still for a second. To look past the immediate and think about where that decision is actually taking you. That is where things start to change. This goes way beyond trading too. Business, life, everything. So take a minute this weekend and think about it: How many times did you know the right decision and still went the other way? And how different would things look if you started getting that right, even a little more often? Be patient. Be self aware. Be intentional. You will get where you are trying to go a lot faster than you think. Cheers, DELI

Spring thoughts…running it back to scratch

Weekend projects…

@LacksonMudenda Congrats!

I never imagined I would be an Assistant Professor of Economics let alone be teaching at a University in the USA. I say this because I can surely be lazy sometimes! But today a surreal moment hit me. I am part of a fantastic faculty team reviewing scholarship applications for students. I have spent a good part of the last 3hrs reading students’ aspirations and goals in their essays. I feel honored and proud to be doing this job. It is truly a privilege I do not take for granted. A huge thanks goes to all the people who were patient with us when we just started out when we made all the incredible mistakes possible and they kept correcting us like babies. Those are the heroes of the story. Do not be afraid to make a mistake. We all start from somewhere. PS: There are still some students who can write in a deep & raw tone. Some essays are AI generated and that can be off-putting!

Bacon mode re-activated…

@FelipeGuirao 💯 And we are not even talking advanced statistics

If you don't know your distribution, you're gambling, period. Knowing your distribution is what separates traders from gamblers, and it's one of the most important things you can do to actually grow your account long term. When you know your stats and system distribution, nothing comes as a surprise because you EXPECT things to happen. I know exactly what my system delivers over a large enough sample size of trades in the market: - 65% of trades are winners - 80%+ of months are profitable - Max drawdown around 15% - Expected return 100%+ per year I swing trade stocks, so the market phase will determine 100% of my distribution. That's how the market works. So when I hit a losing streak, there's no panic because I know it's part of the distribution. When I'm in a drawdown, I know these come in choppy 5-15% pullback phases, just like the one we had since Oct '25 until now, since the market decided to ride momentum on the bear side... That's CONTROL, and it's what keeps you in the game. Most traders are flying blind. They take a few losses and spiral into panic mode. They don't know if their drawdown is normal or catastrophic, they don't know their win rate, their max DD, or if they even have an edge. They're just hoping things work out, which is the fastest way to blow up your account. When you trade with data and know your distribution, you know WHEN your system works best according to the market phase. You know what to expect in bull markets, bear markets, and choppy 2-sided action. You stay calm when everyone else panics because you've seen this before in your data. You execute without emotion. All I do is: > show up 30 minutes before the close, > run my scans, > take any signals that meet my criteria, > manage the risk, > and let the market do the rest. Repeat for days, weeks, months, years. Same thing, again and again. After a large enough sample size, the distribution plays out exactly as expected. That's how you trade professionally. Boring, but it works. I've averaged 100%+ returns with a max DD of 15% over 6 years, swing trading stocks, trading EOD (<30 min/day), with a systematic routine and pure rule-based when to enter/exit trades. Apply statistical thinking to a trading style like this, and you get the best between risk-adjusted returns, max DD, and time required to pull it off, not to mention the lack of discretion so you have peace of mind and don't have to babysit trades during the day. Base your decisions on data, know your distribution, and stay in control 📈

@TheVixhal All of it is overrated anyway, been there done that

Most people with “AI/ML” in their bios don’t even know a real symmetric matrix always has real eigenvalues.

"Empty your mind. Be formless, shapeless like water. You put water into a cup, it becomes the cup. You put water into a bottle, it becomes the bottle. You put it in a teapot, it becomes the teapot. Water can flow, or it can crash. Be water, my friend."

@RealSimpleAriel @markminervini But I simply learned from observing the market and coding the observations over time into a statistical model 🤷🏽‍♀️

If anyone tells you that in order to be successful in trading you need to make some unknown discovery on your own... This is a lie!! - I learned about volatility contraction and progressive exposure from Mark. @markminervini. - I learned about momentum bursts and Episodic pivots from Pradeep @PradeepBonde - I learned about the importance of creating a model book for myself with 100s of past examples. As well as the idea of focusing on higher ADR stocks from Kristjan @Qullamaggie - I learned about stage analysis from Stan @StanWeinstein13 - I learned how to read and interpret COT data from Jason @Crowded_Mkt_Rpt - I learned the right side of the V concept and proper bet sizing on A+ opportunities from Lance @TheOneLanceB - I learned to think a little more contrarian from watching @TheShortBear - I learned about the flat base breakout and value in leading groups from @PatrickWalker56 - I learned about the HVC/HVE edge from @AmeetRai - I learned about trading more aggressively during high momentum periods from @DanZanger - I learned about the Undercut and rally or Double top short sale setups from Gil Morales @gilmoandco - I learned how to think about creating and implementing systems from @Peoplewish - I learned the failed follow through setup on an intraday timeframe from @InvestorsLive - I learned about the importance of prior day channels from @danshep55 - I learned about support and resistance gaps for entries from @NickDrendel - I learned about using the 50sma as a guide to measure extensions from @jfsrev - I learned about creating a daily trading plan from Marcel Link. - I learned about the CANSLIM methodology from Bill O'neal I'm sure there are many I missed but the point is: Trading knowledge is passed down from one generation to the next. But it is up to us, the trader, to implement what we have learned in a safe manner while we put together all the pieces for ourself. We live in a time where you DO NOT need to make up some magical elixir for trading in order to be successful. All the people mentioned above have found an exploitable edge in the market, and like myself relentlessly execute that edge over and over. I am personally grateful for all of the educators I've had along my journey; which is part of the reason I so willingly share any bit of knowledge I acquire along the way. The same way they all graciously imparted knowledge on me.

@hjluks Super useful, thanks

Pain on the outside of your hip is one of the most common problems I see in my practice. Walkers get it. Runners get it. People who've never been to a gym get it. For decades, we called it bursitis and injected it. We were treating the wrong thing.

@MoonDevOnYT Also AUC has many limitations, net benefit and decision curve analysis better assesses market utility

@MoonDevOnYT Try informed network analysis, works much better

Beyond Accuracy: The One Metric Every Quant Developer Needs To Scale Their Trading Bots most people think applying ai to trading is a magic money button but the actual results will punch you in the throat if you are not ready for the truth. i spent years losing money and getting liquidated before i realized that a bot that is 90 percent accurate can still leave you completely broke. if you want to know the one metric that actually matters when the machines take over keep reading because most gurus are hiding the real math from you i used to be that guy staring at charts until four in the morning just hoping the price would move my way. i would get that liquidation email as soon as i fell asleep and feel like the market was personally out to get me. it took spending hundreds of thousands on developers for apps that didn't work to realize that code is the only equalizer in this game the first loop you need to close is why your "perfect" strategy fails in the live market even when the backtest looks like a vertical line. most people find a strategy then dump all their capital into it because they have fomo and fear. they skip the rbi system which is research then backtest then implement with tiny size before ever scaling up when i first tried to apply machine learning like random forest models to my liquidation data i thought i hit the jackpot with 75 percent accuracy. the harsh truth is that my model was just a professional pessimist that learned to always predict the market would go down. because the data was mostly bearish the model just cheated to get a high score while missing every single profitable move to the upside this is what happens when you have a class imbalance disaster in your data set. if 74 percent of your data is down moves then the ai realizes it can look smart by never betting on a bull run. you end up with a model that is a "down only" predictor which is essentially a broken clock that is right most of the time but useless for making money most people think they can just throw 10,000 trees at a random forest model and it will magically find the alpha. i learned the hard way that after about 200 or 300 trees you hit diminishing returns and just start wasting your computer's life. the secret isn't more trees but better feature engineering and narrowing the noise down to the signals that actually move the needle i spent ten years in technology scared to code because i thought it was for geniuses who went to stanford or knew advanced math. the reality is that i just needed to lock in for a few hours a day to realize that i could build systems that trade for me while i sleep. once you learn to code you stop being a gambler and start being a data dog who actually understands the math of the market if you want to actually win you have to stop looking at accuracy and start looking at the area under the curve or auc metric. a model with high accuracy but low auc is just guessing based on the majority class and will leave you holding the bag. you need a minimum auc of point 65 or ideally point 7 before you ever even think about letting a bot touch your real capital the reason i show everything live is because the trading industry is the most secretive business in the world and nobody wants to share the real sauce. most people are out there selling "set and forget" bots that make 20 percent a day but those are all scams meant to take your money. if you don't build your own edge and iterate your systems every single day you will eventually get caught by someone who is jim simons ran up a net worth of over 30 billion dollars because he understood that you just have to make your systems better and better. he wasn't smarter than everyone else because of some magic crystal ball but because he treated trading like a science experiment. he used an rbi framework of researching and testing until the math was undeniable and the emotions were completely gone ambitious people often fail at trading because they think they can just work harder to get better results. in almost every other industry working more hours equals more money but in trading it just equals more overtrading and more mistakes. the only way to work harder in this game is to put more hours into your code and your research so the bot can do the heavy lifting for you we found that focusing on specific signals like btc liquidation cascades and feature reduction actually improved the model more than complex math ever could. by dropping the noise from 1,600 features down to just 6 key liquidation metrics the model started to actually catch the up moves it was missing before. it's about finding the signal in the static rather than trying to listen to everything at once you have to realize that you are chasing the same prize as firms like jane street and renaissance who have more computing power than you can imagine. the only way to compete is to use the same tools they use and stay persistent enough to outlast the gamblers. code is the great equalizer because it allows a regular person in a bedroom with a whiteboard to compete with wall street giants if you're still hand trading with 40x leverage and praying to the chart gods you are just waiting for the day your account hits zero. the real results of ai in trading show that it is a brutal game of iteration where you must be willing to fail until you find the parameters that stick. stay hungry and stay foolish because once you automate your trading your life changes forever and you never look back always remember that the 777 is near to the brokenhearted and that this journey is about persistence more than anything else. if you can't fly then run and if you can't run then walk but by all means keep moving toward your automated future. the math doesn't lie and once you have a system that works you can finally stop staring at the screen and start living your life on your own terms

@aw_trades_ @_not_a_fish I do this for a living. All this may not mean anything until you do live testing, a Phase III diagnostic study

Me too!! I ran out of sample validation tests to check if there was overfitting Trained it on 2024 data then tested on unseen 2025+ and the edge held and actually improved in some scenarios. If there was overfitting this test would’ve collapsed I’ll be posting the whole testing walkthrough Monday and then my validation tests as well! Still more to go!

Fascinated to see how these results play out In my Numerical Techniques we discussed a topic called “over fitting” essentially when you fit a line so perfectly to every data point in becomes a worse predictor of future results in some cases a 60% win rate in backtesting would be far better than a 90% because the 60 is much more reliable

@systematicls Stochastic parrots 😂

AI Hedgefunds are a conceptual dream - Because who wouldn’t want to make money from a few LLM calls? But most are hype and the class of people that could make it work is smaller than you think. If you get agents to trade trivially - you’ll get stochastic parrots, not alpha.

They do, they make a market diagnosis which in itself is a prediction, probabilistic just like in medicine. Similarly, a golfer reading a putt predicts how the putt is likely to break. Diagnosis of the present moment is a prediction and informs a likely prognosis (future prediction)

Most traders think they need to predict the market to make money. Predict the next crash. Predict the next bull run. Predict which stock will explode. But after years of trading, I’ve learned something surprising… The most profitable traders I know don’t predict the market at all. Think about it. If predicting the market was the key to making money, then economists, analysts, and TV experts should be the best traders in the world. They spend all day forecasting interest rates, GDP, inflation, and where stocks are headed next. But how many of them are running profitable trading accounts? Here’s the uncomfortable truth: You don’t need to predict the market to profit. In fact, trying to predict the market is one of the fastest ways to lose money because predictions create: • Bias (you see what you want to see) • Overconfidence • Emotional decisions. Once you form an opinion about the market, you become attached to it. You think the market should go up, so you hold losing trades longer than you should… or worse, you add more. Professional traders approach the market differently. Instead of asking, “Where do I think the market will go?” they ask, “What does my system tell me to do?” Their decisions are based on rules, not opinions, because opinions are like belly buttons: everyone has one, but most of them are useless. For example, a systems trader might have a rule like this: if a stock drops three days in a row, buy. Then sell when it closes above the 5-day moving average. They don’t need to know why it dropped, what the news says, or what analysts think. They simply follow the rule. Sometimes the trade wins. Sometimes it loses. But over hundreds of trades, the edge plays out. You don’t need to be right about the next trade. You just need a system with a positive edge. So if you’ve been struggling with trading, the problem isn’t your discipline. The problem is that you’re trying to predict the unpredictable (which is like trying to predict what my wife wants for lunch). And that’s a game no trader can win. Because the traders who survive long-term don’t predict the market… They build trading systems for it.

Learning is like peeling an onion— each layer reveals another layer, and occasionally, it makes you cry.