User Comments, Suggestions, or Complaints | Privacy Policy | Terms of Service | Advertising
Buy MLB playoff tickets, plus 2011 World Series, 2011 ALCS tickets and NLCS game tickets. We also have Texas Rangers playoff schedule, tickets to Red Sox games and Yankees game tickets. Plus, buy Phillies baseball tickets, Tigers playoff tickets and the biggies like ALDS baseball tickets and 2011 NLDS tickets. |
Demarini, Easton and TPX Baseball Bats
|
AllianceTickets.com has cheap MLB Tickets. Get all your Colorado Rockies Tickets, Seattle Mariners Tickets, San Francisco Giants Tickets and all your favorite baseball tickets here. We also carry cheap Denver Broncos Tickets, Seattle Seahawks Tickets and Denver Nuggets Tickets. |
Page rendered in 0.3920 seconds
54 querie(s) executed
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
1. Esoteric Posted: November 14, 2008 at 08:23 PM (#3009398)Can't say I'm disappointed or that I wouldn't do the same thing. If money and fame are there, and you can still do something you enjoy, why not take it? Besides, predicting politics may be a bit more important than predicting what Mike Jacobs hits next season.
I'd buy both.
Gladwell would like you to pay him $0.10 for using that phrase without blinking.
They had Tom Bradley's campaign manager on NPR a couple weeks before the election, and he said the Bradley Effect was b.s., that mostly they underestimated turnout in GOP areas. He had every reason to want to blame racism and said it wasn't. There, I saved Silver a chapter.
Besides, predicting politics may be a bit more important than predicting what Mike Jacobs hits next season.
Predicting politics is what gives the political season its awfulness. It's not a goddam' horse race. I despise all such polling and punditification.
Gladwell would like you to pay him $0.10 for using that phrase without blinking.
Does Gladwell have a copyright on "intolerably insufferable"? Maybe he should. 10,000 hours my ass.
That makes two of us. Nate Silver is on my very short list of "If he/she says it, I'm listening closely."
Blasphemy!
I think you're right. And I know I've been banging this drum for a while now, but that really cuts BP close to the bone - I think the only full-time hard-core numbers guy they'd have left is... would they have one? I don't know what Davenport's up to these days. It's possible he's working behind the scenes, but his website presence is exceedingly small. The Woolner/Click/Fox set has all been hired by teams.
Predicting is very important (mostly to campaigns, to the public it's pretty useless). Pundification is what makes it insufferable.
The first is a Freakonomics-style guide to politics that answers questions like “Is there really a Bradley Effect?”
The Bradley Effect could exist, but pretty much had no shot this election. The media really really wanted it to happen, because nothing says November sweeps like racial tension! Talk about a fabricated story.
It matters who owns the algorithm for PECOTA. Since it seems that and VORP are the only stats they ever use anymore.
I think the real problem is that pundits tend to view all issues through the prism of the horse race. That's not what Silver does. He just tries to give people a clearer picture of where things stand in the horse race.
What are the odds of there ever being a MLB season where every team, after 162 games played, all have .500 records, leading to a cataclysmic playoff of DOOM!?
The exact year, month, day, hour, minute and second that mankind lands a man on Mars?
Where is the Beale Treasure?
How many licks does it take to get to the center of a tootsie pop?
Where is Osama?
Was JFK's death a conspiracy?
When will the World end?
Maybe. Met him five years ago and he was a really nice guy, though. Very down to earth, no chips on shoulders or the like.
Not that I'm all that privvy to the inner workings of BP, but my impression based upon what I've read is that PECOTA is an absolute rat's nest of Excel programming (this is why Silver says BP can't do in-season PECOTA projections). I have no doubt that Silver can't take PECOTA with him, but I really don't know if they can run PECOTA without him. Maybe some of the behind-the-scenes greasemonkeys know all the wickets, I dunno.
As far as I can tell, development of VORP pretty much ceased when Woolner left. SFR and the baserunning figures seem to have disappeared when Fox left - if I'm wrong on this I'd love to be corrected.
About the only thing of theirs that I check regularly are the support-neutral pitching stats.
That's awesome! Such an influential metric, and it's based on Excel?!
Funny; I heard a Demo version of you say this about Dick Cheney yesterday.
You can do a lot with Excel. I would hope he's using the VBA features. The Chone projections are done in excel, but there are too many calculations for it to handle all the players at once. So it grabs the input data from one sheet, one player at a time, performs the calculations, outputs to another sheet, and goes on to the next player. It handles 1500+ hitters in about 6 hours. I usually let it run overnight.
I'm set up to do in-season updated projections, but I'll only do it for one player at a time, if something requires timely analysis, like a trade or whether Cliff Lee/Chipper Jones can keep it up. In a case like that I'll manually add the data.
To process all the players, I get major and minor league stats from different sources, and without uniform player IDs it takes forever to match them up correctly.
Which is strange because people like Tango have been pointing out flaws which don't get fixed.
My guess is they have some sort of license agreement to continue using VORP but changing it's components would be outside the use license.
Which is basically unchanged from what Michael Wolverton had at least as far back as 1995.
I doubt it. Among other things, last time I checked Keith was using Dave Tate's improvements to runs created (Marginal Lineup Value) plus his own calculation for replacement level. You could take MLV (or any decent metric in the public domain) and calculate your own replacement level (and there are at minimum three viable approaches here)
My money's on "good enough". MLV's got a standard error in the 7-8 run range for a full time player. Best metrics I'm aware of can cut this to around 5. I know Keith was always planning on changing to a better metric but just never got around to it (that I'm aware of) -- and it's no big deal. For my money things like VORP and OPD are where the conversation starts. If you're terribly concerned about precise rankings (as opposed to a general sense of where a player ranks -- call it a letter grade) it's best to use as many good metrics as you can. (and as much other info too)
Maybe a dumb question but the "they" in the second sentence above- who is that? Who actually owns BP?
(Which is probably why quality writers like Zumsteg et al left, since they didn't have something tangible like that to offer, and so, were shut out of equity. This part is mostly guessing. And of course, a darn shame if true.)
Also note that BPro owns a stake in 538.com (based on their press release). So again, my (pure) guess is that there's been another transfer of IP from Nate to BPro in return for equity.
***
They do have full access to VORP, but it becomes a programming issue. Trying to alter someone else's code is never fun, especially if it's a pet project coding that you did, and never considered to document the work.
In any case, BPro-ers believe that their bread-and-butter is the talking of baseball, rather than trying to get accurate numbers out there.
***
If I had the time, I would offer to convert Rally's Excel program into something database-centric. I'd do the same for Nate. Stuff like that should take just a few to several minutes to run. Unfortunately, I've got my other sh!t that I've let pile up.
Well I know who started it, but it seems to me that over the years the people who have a stake has changed.
I'd be surprised if Christina Kahrl hasn't maintained her stake. Last I heard, Dave Pease was still there, and I'd be surprised if he didn't have a share. Same with Clay Davenport.
Wouldn't the transfer of IP be the other way around? i.e.
Nate: I want to use this PECOTA feature to predict elections.
BP: Sorry, you gave us the rights to that, and you can't have it back for free.
Nate: Here, have [X%] of 538.com
BP: Deal.
BP is owned by the Chi Coms.
I plan on doing a full open-source release of my Marcels reimplementation here in the next few weeks, once I have my tutorial series finished up.
Brian: never thought of that, but perhaps. The "idea" of PECOTA is not protected, and if that's what Nate used, then it would work my way. However, if he actually used part of the actual code ("trade secret"), then you'd be right. Come to think of it, that seems to make the most sense. It didn't make sense that BP got involved in 538.
This sounds like the relationship that Lorne Michaels and Vince McMahon has with their employees. We'll never see another Wayne's World or "The Rock" because of that. (Which is not necessarily a bad thing.)
But Kuhn is a Hall of Famer!
Hey, I know that one. Thre answer is three.
Maybe. Met him five years ago and he was a really nice guy, though. Very down to earth, no chips on shoulders or the like.
Exactly. Let's not confuse "he's insufferable" with "we resent his success." Nate is a class act, and deserves every bit of this.
With 538, I notice the same lack. What's the monkey with regard to Presidential elections? What value does 538's system add over the Princeton guys or RCP or the latest polls or whatever? The primaries suggested that there might be some value added with his method, but it appears that everything is ancedotal at this point.
I really like his political writing though.
One thing's for sure, don't ever trust somebody to evaluate the projections when he's got a dog in the fight. That includes me, even though I'm not selling, and especially anything from BPro, as they are selling. I leave it to Tango Tiger, he's done some projection evaluation for the last 2 years.
$700k - not bad.
I hate to say it, but based on the buzz factor and context, his closest comp may be Ana Marie Cox (i.e. the original Wonkette). Her book tanked.
Hopin' for better Nate!
I don't know -- Wonkette's book was a novel. Nate's books at least seem like they actually build on his area of expertise and are the kinds of things people who like his site would want to read. Then again, maybe 538 The Novel -- upstart political blogger discovers that Zogby is being paid off by sinister forces and he finds his life threatened as he attempts to expose them using his own proprietary algorithms -- wouldn't be half-bad.
No part of PECOTA is used in the election forecasts except (afaik) application of the idea of similarity scores -- which is not unique to either PECOTA or the election projection model. Nate doesn't owe BP anything for using that publicly available technique in some other types of analysis.
The methodology underlying the 538 model is more openly disclosed (see the FAQ) than the methodology underlying PECOTA.
AROM is right when he says you can't trust the guy with the dog in the fight. Not that they will cheat or anything, but simply that they may be subconsciously biased and may only see the evaluation from a narrow view.
In my case, I actually publicize that I want other systems to beat Marcel, and bend over backwards to make sure that Marcel doesn't get any benefits.
Anyway, enough yapping. The threads, which desperately require me to provide summaries for, can be found here for hitters (start at post 29) and pitchers (with post 23 talking about PECOTA and forecasting systems in general).
I try to stay away from talking about projection systems, other than just talking about the actual nuts and bolts, because I have an obvious conflict of interest. Self-promotion isn't exactly my strong point.
Of course, he also has to write the two books... as long as Will Carroll isn't his editor though, he should be ok
I'm old enough to remember how an office worked without PCs at every desk. Unless you had access to a mainframe, some simple number-crunching was incredibly tedious. I'd guess that working as Bill James's assistant back then was even less fun than my hypothetical's intern's job.
Gary Huckabay left, proving that BPro had nothing to offer front offices.
Keith Woolner left, proving that Keith Woolner did have something to offer.
Dan Fox left, proving that Dan Fox was smarter than the rest of the ones that were left.
Joe Sheehan's bulls**t doesn't sell, Will Carroll's bulls**t doesn't make sense, and no one's going to hire the fabulous Ms. Kahrl in today's Prop 8 state.
But Nate Silver and Pecota were all that really were difference makers. If that's gone, congrats to Nate and good riddance to the shell of BPro. It had long since outlived its usefulness.
No need to get all Bruckner on us, Dan. You're better than that.
Also, he was a nutcase.
The baserunning figures are still there, both individual and team, in their statistics section. Some of their writers, off the top of my head, Kahrl and Jaffe, sometimes reference them.
Sounds like Numb3rs for the NPR set. "There was an unauthorized push poll in Nebraska; who could have been behind this??"
I'm not terribly familiar with 538. Did Nate just compile and review the data that pollsters gathered? What kind of independent research did he bring in?
From what I gathered, he took the polls' data and used his own weighting based on the poll's credibility (not a simple average) to assess the "true value" of the candidates' support; he then tied those results into electoral vote math. The archives of fivethirtyeight.com should still be up; I wouldn't read the thread comments if you value your time.
I agree with the others above who said that the true "extra" Nate added was his prose analysis of the results and tying in to the mechanics of polling. IME Sean Quinn's on-the-road reports added a lot to the evaluation of the numbers and served the function of independent research.
And... let it be said that the Suck.com book was good in its day, and ahead of its time.
Source:
http://www.zogby.com/news/ReadNews.cfm?ID=1641#Anchor-37902
http://johnziegler.com/editorials_details.asp?editorial=176
You must be Registered and Logged In to post comments.
<< Back to main