Friday, September 6, 2013

How Not To Handicap A Prop Bet (2013 Edition, V2)

Barnwell, Vegas, props, #notthattricky, etc etc etc. Go:

The NFL season might have started yesterday, but today is the day for my annual NFL over/under and assorted (likely losing) bets column.

Life is so great. What did I ever do to deserve this?

This is the third year I've done this column for Grantland...

Absolutely mind blowing that this guy has been allowed to write about wagering on sports for more than a paragraph, let alone three years.

As a result, a line that opened with the Chiefs over 7.5 wins at a price of -115 rose to the point where it ranged from -175 to -195 at the books I visited during my week in town.

This sentence may not seem notable now, but please remember it for later.

Last year, the bets I chose to make would have lost 5 percent of their value had I not been robbed of the tickets.

I had honestly forgotten this ever even happened. What a pro. I'm the single least responsible person you will ever meet and I've never lost a ticket in my life.

Let's get started with one that looks great right now!

Okay Bill, I'm ready for some math. Let's do this.

Denver Broncos MISS the playoffs: +675
$20 to win $135

I think the Broncos are going to make the playoffs

Great start.

So why bet on them not to make the playoffs? Because the odds are nice!

Okay. Certainly, there are lots of scenarios in which you think an outcome is likely to occur, but would still want to bet against it happening at +675. Surely, our hero Billy, a Very Serious Quantitative NFL Analyst for Grantland dot com, will be backing this wager up with some numbers...

A line of +675 implies that the Broncos will make the playoffs 87.1 percent of the time.

There it is, 87.1%. Would prefer to use the team "breakeven" here rather than "implies", since this is not what "implies" means, but 87.1% is correct and I'm trying to keep my expectations reasonable. Now to see what Billy has on the fair value of Denver missing the playoffs:

That seemed a little high to me in that moment...

I am not cutting him off in mid-sentence here to make him look stupid. I am doing so because THOSE TEN WORDS REPRESENT HIS ENTIRE EXPLANATION FOR MAKING THIS BET.

I'm like 15% of the way through the article here and already at a loss for words. There are so many BS ways you could justify this bet by doing very little actual work, but Barnwell doesn't even pretend. This is not even worthy of the #makingshitup hashtag, as nothing is being made up except the bet itself, out of thin air.

Tampa Bay Buccaneers OVER 7.5 wins: -135
$200 to win $148.15

Betting "to win $148.15" is a really clear sign you do not actually have any talent when it comes to betting on sports for an overwhelming number of reasons.

Tampa Bay Buccaneers WIN the Super Bowl: +6000
$20 to win $1,200

And while 60-1 is a pretty big long shot, there would also be some value in hedging this bet and locking in a smaller victory if the Buccaneers were to make it as far as the NFC Championship Game.

Making a bet and principaplly explaining it with "I can lock in profit if they do well" is a really clear sign you do not actually have any talent when it comes...

Also, despite that mess, no mention of odds on the Bucs to win the NFC? Why, exactly?

How will this even work logistically? Is he going to fly out to Vegas before the Super Bowl if he needs to hedge? Doesn't really seem optimal. Does the Grantland columnist who is confounded by conditional probability have $500+ sitting around in an offshore sportsbook account to hedge with? So many questions.

Kansas City Chiefs OVER 7.5 wins: -165
$250 to win $151.52

Scrolling back up...

As a result, a line that opened with the Chiefs over 7.5 wins at a price of -115 rose to the point where it ranged from -175 to -195 at the books I visited during my week in town.

At this point, it's pretty clear that I have read this article more closely than all of Grantland's editors combined, and I'm fairly concerned for everybody involved. Mostly myself.

To me, in a weak division with an excellent core of talent, [the Chiefs are] a 10-win team. With that in mind, knowing the books had been hit with plenty of action on the over 7.5-win line, I was hoping somebody would offer the Chiefs at eight wins so I could take the over on that, probably at somewhere around even money. Instead, I had to settle for the massive amount of juice.

Price doesn't matter because Kansas City is a lock to win nine or more games? Is that what I'm supposed to be taking away from this?

Indianapolis Colts UNDER 8.5 wins: -120
$100 to win $83.33

Consider this a lesson learned from last year's bet on the 49ers.

I fucking love learning!

I'm confident that the Colts will regress toward the mean in terms of luck and not win 11 games, as they did a year ago, just as I was confident the Niners wouldn't win 13 games in 2012.

We are properly using the word "regress"!! So much progress. I'm so proud.

They didn't, of course, but they also didn't drop by all that much; San Francisco still won 11 games, which beat their posted total of 10 wins, so I lost my under bets there.

Okay, "proud" didn't last long. I think I see where this is going. Extremely worried.

In that case, what I failed to properly account for was that the regression was already priced into the line.

Deep breath.

The 2011 San Francisco 49ers won 13 games.

The over/under on the 2012 San Francisco 49ers' win total was 10.

It took Bill Barnwell, Very Serious Quantitative NFL Analyst for Grantland dot com, AN ENTIRE YEAR to determine that "regression was already priced in to the line".

The posted win total for the 49ers was a full three wins below their record from the previous year.

Which was...not clear to you at the time?

I don't regret writing that the Niners would regress

Maybe not. But betting into that U10 line based on the fact that you thought the team that won 13 games the year prior would regress? Thinking I might regret that part a little bit. Like cutting-myself-off-from-ever-using-the-internet-again regret.

Philadelphia Eagles WIN the NFC East: +600
$20 to win $120

I won't pretend it's likely, but the 5-11 Redskins weren't exactly favorites to win this division last year, either.


Jacksonville Jaguars OVER 4.5 wins: -160
$100 to win $62.50

I don't have super-high hopes for the Jaguars, if I'm being frank. But five wins in the AFC South? That's very feasible. The Jags were 2-14 a year ago, but they underperformed their Pythagorean Expectation by 1.4 wins...

So, 3.4...

and should have better coaching on both sides of the ball this year.


The average team with two wins or worse in a given season has won an average of six games the following year, with 17 of the 24 previous examples winning five or more.

Completely unfounded implication that the Jaguars are as good as the average team that went 2-14 in the previous season. Jacksonville is a 3.5-point dog at home against the Chiefs in Week 1. They are awful.

That seems relatively safe, and in fact, most books around town have the Jaguars posted at five wins, not the 4.5-win figure.

Using the word "safe" when talking about a wager is a really clear sign you do not actually have any talent...

For someone who explained breakeven/"implied" percentages earlier in this article, Bill sure seems to have no concept of the significance of juice.

Oakland Raiders, MISS the playoffs: -950
$800 to win $84.21

Well, this certainly has potential.

OK. Just as Simmons invested heavily in the Raiders playing terribly for his bets, I'm of the opinion that the Raiders will be either the worst team in football or among the worst teams in football, which led me to these two bets.

Sure. The Raiders suck. Totally on board.

The bet on the Raiders missing the playoffs actually opened up at the Wynn around -1650, but there's been enough action on the corresponding "Yes" playoff bet that the "No" price has dropped down to -950. That's also true of every other long-shot team's playoff odds, though; the Cardinals and Jaguars each made similar sorts of moves, which suggests that the public is betting on just about every long shot to make the playoffs.

Always a good idea to end a paragraph with a blatantly wrong assertion, IMO. Maybe ask Chad what "public" means?

My best guess is that the worst team in football would make the playoffs much less frequently than 9.5 percent of the time. A study estimates that a team whose true talent level is around 5.0 wins — which is where the market suggests the Raiders are — would make the playoffs around 3 percent of the time.

Yep, we're all definitely 100% confident that the Raiders' true talent level is exactly 5.0 wins. NFL teams are know for being easy to predict from season to season. And, even if they weren't, it's not like variance from that 5.0 win level is a bad thing when you're laying -950, right?

Oh, damn. Nevermind.

Pittsburgh Steelers OVER 9.5 wins: +130
$80 to win $104

The +130 line on the over-9.5 win bet implies that the Steelers will win 10 games or more 43.5 percent of the time; I think it's probably closer to 50 percent, given their schedule, the coaching, and the talent on their roster.

No comment. Beyond parody.

Go Raiders.

1 comment:

  1. 8-15 (-$106.35) is a clear sign you do not actually have any talent when it comes to betting on sports for an overwhelming number of reasons.