[Carl-Haakon]

...or do they simply follow market movements? I mean from a purely mathematical perspective, many of the different types of odds (1X2, O/U, DNB and others) seem vulnerable to almost all types of arbitrage that I can think of that there aren't online calculators for. Sometimes I see this in opening lines. Sure, some of these types of bets have much lower limits than others, but still. Has anyone else noticed this? Does anyone have any idea how these market failures come to be?

To clarify, I'm not talking about arbitrage between books, but between different types of bets on the same event all placed at the same bookmaker.

(It's also possible that I simply don't understand the rules of the different bet types well enough)

[brettd]

Yeah you're seeing this inefficiencies correctly.

Exploit them though, and you'll be restricted pretty quickly.

[Foxy Dread]

These inefficiences will disappear in the near future as more books employ quants teams to build automated pricing tools that are driven by a couple of indicators, therefore guaranteeing consistency within the book.

[Carl-Haakon]

I was wrong, and I'd like to apologise to anyone who might have believed me. The reason I posted this is that I had programmed a bot that would log on to oddsportal, collect the odds for different bet types and check for arbitrage opportunities. Because of a subtle (but important) mistake in the way that my bot calculated return on risk, it would find arbs almost everywhere. I recently found the time to correct it and it seems to me that pretty much no mistake is made by the oddsmakers when it comes to setting odds with regard to their own odds on other bet types.

brettd may have found something that I haven't, and it's possible that mistakes are made when different bet types are updated at different paces (I could only check closing lines), but the egregious results that I found where simply untrue. There are of course arbitrages between bookmakers, but those are usually small.

For anyone who's interested, my mistake in calculating return on risk was that I would exclude the nonrisked money from the amount I had bet, but include it in the payout side. For instance, if there is a DNB bet in soccer with odds of 3.40 (+240) for the away team, with 1X2-odds of 1.79 (-126) for the home team then I (or my program) would calculate an artificial anything-but-draw bet as having odds of (and I'm using randomly chosen stakes that sum to 100 here):

earnings when 1 or 2 1.79*65.5 + 34.5*3.4
---------------------- = ------------------------ = 1.79 (-126)
losses when x 65.5

when the true odds should have been


1.79*65.5 + 34.5*3.4 - 34.5
--------------------------------- = 1.26 (-384)
65.5

I realise that few people care, but I just didn't want anyone with bad math skills to waste time searching for inefficiencies that do not exist while I establish a reputation for myself as an idiot.

[dalogester]

Very interesting. I would lean towards yes.

[Pot luck]

Hey Carl, off topic but how are you collecting the odds from oddsportal? Would love to talk to you about this. I did this a while back but they changed their site a bit, started fetching their data with AJAX and I never fully got around to updating my scraping module.

[Carl-Haakon]

No worries Pot luck, the topic of this thread was pretty dumb to begin with. I use selenium for oddsportal. It's slow, but reliable and easy. There may be better ways to do it though, and I'm no expert. I've never learned how to use AJAX. We can trade code, if you wish.

[Ced's shadow]

Hey Carl, if you'd like to know what happens as an insider you may want to check this post of mine, there's a lot of factors that affect the outcome of this as well as how it's done. Best regards.

http://www.sportsbookreview.com/forum/players-ta...you-stand.html

[Pot luck]

I sent you a PM

[sorinnn]

Hey Carl, off topic but how are you collecting the odds from oddsportal? Would love to talk to you about this. I did this a while back but they changed their site a bit, started fetching their data with AJAX and I never fully got around to updating my scraping module.

if you still need help with coding you can PM me

[HoldenRg]

Good stuff my friend


PEACE!