3 ms·
Depending on the market you are betting on, how quickly you need updates and what you can afford, you have a choice of subscribing to a live data feed (e.g. fro
by seanos 16y ago
Depending on the market you are betting on, how quickly you need updates and what you can afford, you have a choice of subscribing to a live data feed (e.g. from http://www.enetpulse.com/ http://www.enetpulse.com/ or http://www.pressassociation.com/sport/data.html http://www.pressassociation.com/sport/data.html), scraping a live score website or updating manually from the venue or live TV (although this does have a delay!). You do not always need ultra fast updates since:
Some markets are suspended and all unmatched bets cancelled when a significant event happens (such as when a goal is scored in football).
Any in-play market places a 5 second delay between placing a bet and having it matched, giving the market makers a chance to get out.
- greyman 16y agoDid you try to write some bot yourself? If so, for which sport? You can also PM me if you want, I do have interest in this area.
- seanos 16y agoI have written a few bots before but only in certain soccer markets such as over/under goals.
- greyman 16y agoThat's cool, was it long-term profitable? I traded under 3.5 and 4.5 (manually) in a few matches with quite a good results.
- seanos 16y agoThey have been profitable, although I don't use a fully automated strategy yet, the bots show me where to bet based on statistical analysis (Monte Carlo analysis) based on past games, the current situation in each game and the available odds - I place the bets myself combining information from the bots and my own subjective analysis. The fact that profits are tax free helps! I'm not sure how feasable full automation is, maybe in certain sports like cricket that can be analysed well through stats alone, but for sports like soccer it is difficult to build an accurate statistical model that can outperform a market that can make subjective judgements based on the precise circumstances of a game. Reading the article, it seems like this hedge fund is using experienced traders to provide judgement backed by statistical modelling too.
- greyman 16y agoI've been chatting with a few people and they confirmed to me that outperform football market long-term (by statistical model or any other way) is indeed very difficult, and most professionals currently do tennis and horses. The reason is that predict a goal in football is difficult, while the tennis match can be partially predicted by watching the game.