Comment by dheera | Hacker News Reader

Comment by dheera | Hacker News Reader

Parent Full thread

Strategies like this can easily be positive EV until enough people discover it and that actually is the driving force behind efficient pricing.

Here's how the mechanism works: I find that something is statistically worth $0.70 but I am able to buy it for $0.60 and statistically sell it for $0.70 (in the average). I make $0.10 each trade on average. Until you come along, copy my strategy and change $0.60 to $0.61 to frontrun my trades. Then someone else does it for $0.62. Until the market finally reprices to $0.70 where it should be. The guy who tries $0.71 loses money and stops, and then it goes back to $0.70. It's a stable feedback loop.

There are lots of positive EV strategies lying around in these inefficient markets that Citadel hasn't (yet) descended upon. The best advice I can give is if you find one, trade the hell out of it and don't open source it or tell anyone about it, because as soon as more people run it, it will cease to be positive EV and then after that it becomes an infrastructure game.

If it's popular on Github it probably doesn't work.

If you found something that works and is paying your rent, don't put it on Github. My 2 cents.

This all really depends on the efficiency of the market. I think these prediction markets would claim that is their goal, but even Wall Street isn't perfectly efficient. I would also guess that sports betting sites like DraftKings or FanDuel would be even less efficient and less likely to be swayed by a popular GitHub repo. Once again, it goes back to the share of the market that is participating for entertainment purposes. That's a lot more common for sports betting than it is for will the US bomb a certain country.

> but even Wall Street isn't perfectly efficient.

Yes, you can find positive EV trades on Wall Street as well. I've been diving into this a lot lately, all I can say is it's 10X harder to find strategies that work on Wall Street than prediction markets.

With one exception.

The one easy long-lasting anomaly to exploit on Wall Street which actually does NOT exist on prediction markets: "American stocks go up most of the time". This is actually a massively exploitable structural anomaly (you just buy and hold forever and effectively reap the rewards of a biased coin) and the source of the anomaly is mostly US monetary policy, US foreign policy, and US tech expertise put together. However, it still is an anomaly. The thesis that SPY will continue going up forever is also predicated on these things continuing to work the way they have in the past.

How is it an anomaly? Global stocks go up (or at least have positive total return) over time on average because companies produce value. Ultimately it's true that that money has to come from somewhere, so you can say printing it is monetary policy, but the reason it can be done without runaway inflation is the tangible value produced by the firms.

> I would also guess that sports betting sites like DraftKings or FanDuel would be even less efficient

Your strategy doesn't work on sportbooks to begin with, because bookmakers don't move the odds with the action.

That is, there is no such phenomenon as "the over is exciting therefore overpriced". Bookmakers price purely based on facts and statistics. Their pricing isn't affected by excitement nor by how many people are betting a certain way.

>Bookmakers price purely based on facts and statistics. Their pricing isn't affected by excitement nor by how many people are betting a certain way.

If this were true, lines would never move unless there was breaking news, but we see lines move all the time without there being any material change to those "facts and statistics".

I think the distinction is that sports betting companies are basically casinos, need to guard their edge, and although they will tolerate some moving of lines, they will kick out players who consistently eat their edge, and will rig the lines at a place where they can still profit.

Different from a prediction market like Polymarket or Kalshi whose income probably comes mostly from transaction fees rather than house edge. Otherwise these platforms wouldn't welcome bots so much. Bots => efficient pricing + transaction volume => profit for them

These are all reasons supporting my point as they would make sports betting platforms less efficient meaning it would be easier to find arbitrage in their prices (at least temporarily, until you're booted for being too successful).

The sustainability of prediction markets depends heavily on continuous liquidity provision - without bots and market makers, spreads would widen and user experience would degrade quickly

> without there being any material change

Without there being any material change you can see. If you had access to all the tips and data and insider information that sportbooks operate with, you could be a bookmaker too.

>If you had access to all the tips and data and insider information that sportbooks operate with

Can you give an example of what you're talking about here? Because it sounds like you're accusing these large publicly trade companies of participating in organized crime. There is regulation when it comes to sports betting that doesn't exist with general prediction markets. An athlete can't just feed a sportsbook "insider information" in the way you're suggesting. The only private info that they are supposed to have is better behavior details that you claim doesn't factor into their decisions.

Books like DraftKings and FanDuel get their lines from market makers like Circa. Market makers use a variety of information for setting initial lines (you'll have to go ask them), but one of the main ways they move lines afterwards is professional action. That is, if some person or entity who Circa knows to be a profitable professional better puts down a large bet in a certain direction, Circa will move their line in that direction.

Where did that entity get that information, and how are they right so often? Your guess is as good as mine. I'm not accusing anyone of anything.

>That is, if some person or entity who Circa knows to be a profitable professional better puts down a large bet in a certain direction, Circa will move their line in that direction.

Well that's the source of our confusion then. I agree with this, but it conflicts with what you said a few comments up:

>because bookmakers don't move the odds with the action.

> Bookmakers price purely based on facts and statistics. Their pricing isn't affected by excitement nor by how many people are betting a certain way.

A bookmaker is a market maker, and they ideally want to end up with no net interest in a position. They then take guaranteed profit in the bid-ask spread, which in sportsbooks is the 'vig'. Bookmakers who adjust their odds in real-time don't have to be particularly clever about the fundamentals, just responsive to the competing demands on either side.

A bookmaker who intentionally takes a position on a game is the equivalent of a proprietary trader or hedge fund. It's potentially more profitable, but it's also adversarial against 'sharp' traders.

Bookmakers who set odds at the beginning and don't move with the action must set larger bid-ask spreads to compensate.

Reply on news.ycombinator.com