Prediction is a lot harder than people think, even with billions of dollars of research and data at your disposal.
Freakonomics: What Went Wrong? A look at the errors of the Freakonomics franchise. I'll say from personal experience that the Freakonomics folks are committed to getting the story right. I recently worked with them for one of their 'Football Freakonomics' spots on NFL Network.
If you're into data visualization, you'll like these. Or they'll make you really depressed. Either way, here are 10 of the most interesting economic graphs of the year from British economists.
Mike Eayrs looks like an interesting fellow. He's the director of research for the Packers.
The worst hockey game ever. At least it was innovative.
I used this article on double-yolk eggs to explain to a reporter from NPR that the Broncos win streak is not simply the product of their individual game win probabilities. Here was the piece from NPR.
This makes me want to go out and buy a #27 jersey.
The NHL has a video rulebook. The NFL should do this. Helmet-knock: Tango
The future is calling. My take is that it's not about the technology and the mountains of data you can produce. It's about making sense of all that data. It doesn't matter if you design a play on an iPad or the back of a napkin. Does the play work?
At the Community Site, Jim Glass revisits the Big Win Index.
Also at the Community Site, Mike Beuoy updates his betting market rankings.
Do the Packers take their foot off the defensive pedal late in games with big leads, and is that why their stats are poor? Here's FO's analysis.
- Home
- Unlabelled
- Roundup 12/17/11
Roundup 12/17/11
By
Brian Burke
published on 12/17/2011
Subscribe to:
Post Comments (Atom)
Make you want to buy a #27 jersey? I find it hard to believe you don't already have one.
This dated 12/12/11 so it is appearing last week in the list of posts, instead of up top as the most recent.
Thanks Jim.
I know. I should have a whole closet full of Ravens jerseys!
The hockey part is funny. As much as i like the Red Wings, the game itself is soooo wrong designed! That is why i liked it, that the Flyers and Lightning eposed the game at least a little bit.
I talked to many hockey fans and people involved (some low class players): Nobody could convince me that my tactic would not work; Just keep all the players as close as possible to your own goal, may even let some players just simply lay down in front of your own goalie. Never ever will a puck go through, even if the opponent shots 100+ times then.
So once you lead or are secured of a play off spot, just stop playing. The goals are just too small. Even a goalie alone almost covers 100% of the net. I mean you can see it in penalty shoot outs were it is not unlikely that only one goal is scored in 10 tries in one-on-one situations. It´s almost 100% luck to score.
Again, what a wrong designed game! Thanks to Walter Camp that football in comparison is almost perfect (the biggest downside is the needed violence). There is no way a true defensive tactic like in hockey or soccer would work (besides kneel downs inside the TMW).
Karl, Germany
Karl, that is totally and completely wrong. I have seen rec league teams try it. It has never, ever worked.
"Screening" your goalie in such a fashion 1) Allows the offense to pick their shots, 2) Prevents the goaltender from seeing an incoming puck, and 3) Prevents the goaltender from reaching and freezing a loose puck.
Also, shootouts have a success rate of around 25%.
You appear to have next to no understanding of the game you're critiquing.
My mistake: Shootout percentage is actually almost exactly 33% this season.
Anon, i admit i am not into hockey, but you are wrong too.
If i put two players on the left and two on the right side of the Goalie i accomplish
1.) My goalie can see the puck
2.) My goalie can reach out his glove (and stick) for defending shots
3.) No puck ever will go trough into the net, because i also have one man laying in front of the goalie. After all, the net is 100% covered and thus impenetrably.
Yes the Offense can shoot 200+ times a game. But for what? It´s like shooting into a wall without holes. Or playing pool billards without pockets, or football without an endzone etc. etc. ...
P.S.: "Shootout percentage is actually almost exactly 33% this season"
Imagine one on one situations, and only 33% succees. Now imagine one shooter vs. 6 "Goalies"...
After all, Hockey is designed wrong and played with false tactics. Sorry.
Karl, Germany
1. If you station two men to either side, the shooters are going to go low every time. The only reasons goalies can make low saves with regularity are because they were enormous, outsized leg pads, and because they endlessly practice butterfly style, where they drop to their knees and turn them inwards to stack the pads, creating a wall. A skater can't do that, and it is incredibly difficult to stop a low, fast shot. You've either got to stop it with your stick (and anyone who's actually watched or played a hockey game can tell you how difficult it is to receive a hard pass, let alone a shot) or you've got to stop it with your body, which is dangerous. And once again, the goalie can't see it coming in past his defender's skates.
2. Goalies, again, wear much, much more protection than skaters. They can stand to take a slapshot to the chest, or even the face, though it still hurts like hell. A skater taking the same shot probably won't finish his shift, and maybe not the game. You will see skaters go out of their way to block shots with their bodies, even to the point of lying down, but this is an injury-intensive desperation tactic. Asking them to do it even more than your average team will deplete your roster very quickly.
3. Even if a goalie makes a save but can't catch the puck, that many bodies in the crease prevents him from freezing the puck. A loose puck in the crease is a dangerous puck.
4. Oh, and the many bodies will also greatly increase the chances of a skater accidentally freezing the puck or dislodging the net, which are delay of game penalties.
5. If the save is made and a skater gets to it first, what next? There's no one high in the zone to lead a breakout. You can't dump the puck down the ice, because the opposing team will either recover and begin the attack again, or it results in an icing call, which means the faceoff is right outside your net, and you're not permitted to substitute your tired players.
6. Oh, and even if the goalie can freeze the puck - the faceoff is still right outside his doorstep.
7. Pucks deflecting off defending players into the net is a regrettable occurrence that happens several times per season. It will happen several more times per season under this scheme.
This scheme makes it difficult for the goalie to either see or freeze the puck, increases the risk of injury to skaters, increases the likelihood of taking a penalty, keeps the puck in the defensive zone, and prevents you from substituting tired players.
And on top of its defensive deficiencies, good luck scoring a goal with your team clustered around your own net and no one to lead a breakout. You are essentially dependent on the other team gifting you with a power play, and teams penned in their own defensive zone are far, far more likely to take a penalty than the attackers.
The scheme you describe is superficially similar to penalty kill strategies, except no one who actually plays or coaches hockey is stupid enough to clutter up the crease. The box and diamond penalty kills only succeed because a shorthanded team can't be called for icing, and can thus clear the puck with ease. Even then, power plays have a 20% success rate, or a goal scored on average every 5 minutes. A goal every 5 minutes is 12 per game. NHL average today is a little over 2.5 per game.
This is, without a doubt, a
Craigo
thanks for the nice summary, since i am not into hockey that much (only take notice when "my" Red Wings lose another playoff series to lower seeded teams after finishing their usual No. 1 in the regular season :-)).
The injury points are good ones. Let me think about it for a while...
Karl, Germany
The simplest way to say it is that you're asking your team to play as if they're on the penalty kill when they're even or ahead.
And as far as the 1-3-1 goes
1. Every team uses it to some extent, just like every post-1995 team used the earlier forms of the trap.
2. The 1-3-1 isn't as impenetrable as it sounds. The simplest way to beat it is to send a forward over the red line to deflect a hard pass into the corner - not ideal, but usually effective. And I've heard someone suggest a very innovative tactic - deliberately pass the puck to the "forechecker" on the blue line so that another forward can take a clean run at him. I'm not sure what kinda result it would have on the immediate play, but it would pay dividends over the rest of the game.
Wait, how does the egg article explain why the Broncos' win streak is not simply the product of their individual game win probabilities?
Also, great article critiquing Freakonomics. I love Levitt, but like a lot of economists, particularly those specializing in micro, he tends to be dogmatic and dismissive of criticism. In that regard, he reminds me a little of Dave Berri, whose writing I also enjoy but who has gotten into trouble for sometimes dismissing those with valid critiques of his methodology.
Because the serial probability of finding a double-yolk egg is higher than the independent probability. (They tend to come in bunches.)
If the model is truly wrong in assessing Tebow's Broncos, then it will be wrong week after week.
Never mind, I think I get it. You're speculating Tebow's the reason their wins are currently coming in bunches, correct? And this factor isn't fully reflected in the GWP because he hasn't played the full season?