Tuesday, February 03, 2009

Super Bowl Prop Redux

TPS regular Rob "Prop Isn't My Middle Name, But Probably Should Be" Holub hit all 5 of his prop bets after we posted sportsbook.com's 16 pages of wagers. Myself? 2-2-1.

In discussing the finer points of prop bets with Rob, however, the following emerged. I'll put up the numbers:

1st player to score TD
(Player, Team, Open, Close)

Willie Parker (Steelers) 6-1, 6-1
Hines Ward (Steelers) 7-1, 7-1
Santonio Holmes (Steelers) 8-1, 8-1
Heath Miller (Steelers) 10-1, 7-1
Nate Washington (Steelers) 15-1, 15-1
Gary Russell (Steelers) 18-1, 18-1
Mewelde Moore (Steelers) 35-1, 20-1
Carey Davis (Steelers) 35-1, 35-1
Ben Roethlisberger (Steelers) 25-1, 15-1
Larry Fitzgerald (Cardinals) 6-1, 9-2
Anquan Boldin (Cardinals) 12-1, 5-1
Steve Breaston (Cardinals) 15-1, 10-1
Tim Hightower (Cardinals) 15-1, 10-1
Edgerrin James (Cardinals) 20-1, 8-1
Leonard Pope (Cardinals) 20-1, 20-1
JJ Arrington (Cardinals) 20-1, 22-1
Jerheme Urban (Cardinals) 20-1, 22-1
Kurt Warner (Cardinals) 20-1, 22-1
Field (any other player) 5-1, 6-1
No TD in game 40-1, 40-1

Notice any trend? The vast majority of the discrepancies between the opening lines and the closing lines are adjusted downwards, i.e., you as the bettor get more return for the same bet placed initially after the odds come out as compared to right before the game starts. And the bets with large swings in odds are all downwards-- the only upwards movements were two Cardinals players from 20-1 to 22-1, and then the field from 5-1 to 6-1.

1) Why the movement in one direction? Perhaps the book didn't have a solid foundation upon which to open bets. If that's the case, however, why not open the odds too low and then adjust upwards? Are they trying to pull extra traffic to the site with appealing odds?

2) Is the structure of betting on just one player from a pool lend itself towards movements in this direction? Parimutuel pools handle this just fine; do we see the same general downward pattern in similar, non-parimutuel pools-- like the Kentucky Derby? I'd suspect not, though the betting volume (Derby vs. 1st touchdown pool) in the Derby would be a lot larger.

3) Don't underestimate the importance of the field moving from 5-1 to 6-1. That could well even out books.

4) Large movements could also be indicative of small amounts of money being bet overall, in comparison to a few relatively large bets. Would sportsbook.com splice the bets that way? Or would they view all 16 pages of props as the pool they want to make a profit on?

Anything else? For the record, the first touchdown in Sunday's game went to Gary Russell, paying a respectable 18-1.

1 comment:

rolub said...

1) I think they're more willing to err on the high side of the odds to draw in bettors on these props. I'm not sure if you could fairly predict the "true" odds for each player, but I'd be willing to bet (not ironically) that over 90% of the offered odds are lower than their true odds. I also believe that less money is risked on the higher payouts, so as little money comes in on a certain amount of longshots (James, Moore, and Boldin come to mind considering their opening/closing odds) with potential high payouts, the books adjust downward to stop the pace of cash flow on those particular longshots. Skipping ahead, I'd like to think that the fact that Russell's odds did not move means the books fared very well on this prop.

2) Contary to popular belief, sportsbooks are no longer just trying to grab the juice and attract even action on both sides; sportsbooks are actively seeking sides to beat the public in some cases, although with this prop it's via different means but to the same end. I know little about the thought process behind horse racing odds, other than that they do react much more quickly and apparently evenly to incoming money.

3) No defensive players were listed, and therefore, all fell into the Field for the first TD at 5:1 or 6:1. It was interesting to me that Troy Polamalu did not have his own odds, somewhere in the range 40:1 to 50:1. I think he could have attracted some action, with his true odds of scoring the first TD much higher than that. Sportsbook offered +145 (or roughly, 3:2) that a defensive or special teams TD would be scored in the game. I thought that was extremely low, but it did pay off on Harrison's TD. As I think this out, I may have answered my own question: Sportsbook.com's low odds on a DEF/ST TD being scored represented the likelihood it would happen, and therefore, preferred to list all defensive players in the field at low odds rather than some individual stars at very high odds.

4) I addressed this briefly in #1. Overall, I think the book is trying to gain an edge on each type of prop, and not necessarily its entire list. If it left itself susceptible to a particular longshot without adjusting the odds and it hit, one large payout to a large amount of money risked could cripple their entire prop list.

My own final tidbits:

Despite a 5-for-5 showing on these posted props, Sunday could not have gone any worse. I took a happy-go-lucky shot on Warner's completion total of 29 at 20:1 odds. The TD to Fitzgerald with under 2:30 to play was his 29th completion. Thanks to Pittsburgh scoring, and scoring with too much time left, Warner would rack up 2 more completions before fumbling.*

I had somewhat of a "free bet" on Nate Washington scoring first at 15:1. He had his man and the safety beat down the middle of the field by over 5 yards, yet Roethlisberger underthrew the ball so horribly that the defender closed in to deflect a sure first TD of the game.