Hey guys! Found this post from Playerrs.com
Does this guy making any sense?
On occasion, when you're searching for patterns in sports data, something will turn up that initially surprises you, but when you have time to think about it, you see how it makes sense. Take this example in the historical betting data: games the home team wins tend to go under and games the road team wins tend to go over the book totals condition. It's unusual to be able to find a consistent bias over such a large sample of games; the betting market should be too efficient. But the reason for it is simple when you think about it; when a home team wins, it'll likely not bat in the ninth, but when it loses, it will. That's the half-inning difference that, over a long stretch of games, accounts for the difference in totals. It's not so easily exploitable -- you're basing a totals bet on your confidence in the outcome of a side bet -- but it's nice to find these patterns and be able to make inferences on what they reveal about the structure of the game.
Oakland Athletics v. Detroit Tigers
The total headed into the game is 9 runs. I don't have a model built yet for forecasting scoring in individual games, but I like the two teams to score over 9 runs. Why? Galarraga, an excellent rookie starter for the Tigers, struggled badly in his last outing against the Twins. The announcers for that game observed that he appeared to have lost his release point during a stretch of consecutive first inning walks. I think that counts as adversity for a rookie, who's already facing it by stretching his arm for this long and this late into the season. So I like the A's to have a better than usual chance of scoring and possibly winning, which would trigger the over rule from above. As for the A's winning the game, with Gallagher as their starter, the team tends to lose. And a losing record means that the pitcher's scoring distribution shifts to the high end. Meaning the Tigers, with a very good hitting lineup, will be facing somebody they can get too as well.
Does this guy making any sense?
On occasion, when you're searching for patterns in sports data, something will turn up that initially surprises you, but when you have time to think about it, you see how it makes sense. Take this example in the historical betting data: games the home team wins tend to go under and games the road team wins tend to go over the book totals condition. It's unusual to be able to find a consistent bias over such a large sample of games; the betting market should be too efficient. But the reason for it is simple when you think about it; when a home team wins, it'll likely not bat in the ninth, but when it loses, it will. That's the half-inning difference that, over a long stretch of games, accounts for the difference in totals. It's not so easily exploitable -- you're basing a totals bet on your confidence in the outcome of a side bet -- but it's nice to find these patterns and be able to make inferences on what they reveal about the structure of the game.
Oakland Athletics v. Detroit Tigers
The total headed into the game is 9 runs. I don't have a model built yet for forecasting scoring in individual games, but I like the two teams to score over 9 runs. Why? Galarraga, an excellent rookie starter for the Tigers, struggled badly in his last outing against the Twins. The announcers for that game observed that he appeared to have lost his release point during a stretch of consecutive first inning walks. I think that counts as adversity for a rookie, who's already facing it by stretching his arm for this long and this late into the season. So I like the A's to have a better than usual chance of scoring and possibly winning, which would trigger the over rule from above. As for the A's winning the game, with Gallagher as their starter, the team tends to lose. And a losing record means that the pitcher's scoring distribution shifts to the high end. Meaning the Tigers, with a very good hitting lineup, will be facing somebody they can get too as well.
