« Well alrighty then | Main | You are on the way to destruction »

Do you want to play a game?

Bill:
Well, just when we thought Rob Bradford was the MSM's bloggah MVP, he goes and chews on the hemlock that is "live blogging."

 

Mike:
I don't get it. I don't know a single person who gives a rat's ass about reading a live blog, yet, they persist.

 

Bill:
Live blogging is like Paris Hilton. I don't give a shit about what she does or what she says, and yet she's always on the friggin' TV.

 

Mike:
Yeah, I heard Amnesty International is all up in ahms because down in Gitmo they force the enemy combatants to read live blogs as a form of torture.

 

Bill:
Go ahead and flush my Koran, mofo infidel, just don't subject me to that!

 

Mike:
So Diamond Mind's 2007 projection where they simulated the season 200 times doesn't do any favahs for the Red Sox. The computer predicts 3rd place and only 86 wins.

 

Bill:
Screw that. This is baseball with all its infinite variations and not a game of tic-tac-toe.

 

Mike:
Of course, if the Diamond Mind's algorithm predicted, say, 98 wins for the Sox …

 

Bill:
Yeah, yeah, in that case I'd be all "Woo-Hoo, I'm Alan Turing's bee-atch!"

 

Bill:
Consistency is the hobgoblin of small minds and all that, you know …

 

Comments

FOOLISH consistency and LITTLE minds, H.B. (Sorry, but it's one of my fav lines.)

Let's just hope Josh's foolish consistency of sticking with fastball after fastball in the American League ends tonight.

Yup, we got your statisfuckintistics right here. Let's play the games, ok?

Is it just be or is the water in the clip art rising in the harbor?

lc

PECOTA has the Sox at 92 wins and the WC.

A "WHOPPR" of a post today, hb. Bradford's live blog did little more than Curt's in rehashing pitches and counts other than to include radar results for pitch speed. Don't get much out of it, frankly. I was disappointed with Bradford's game blog. I prefer he return to the insights that made his ST efforts worth a read.

Bob,

Nothing to apologize for. But be mindful that the character speech is meant to mimic real speech. So like when most people talk, they are going to shorten, rephrase, paraphrase, etc. and very well may get things twisted/incorrect.

This is also the case with grammatical constructions like "So don't I" or "How why ya?" etc that you may see the characters saying.

He EGG, didn't the Ghostbusters nail you at that big hotel?

On March 20th, SG at Yankee blog replacementlevel.com used Diamond Mind to run a whole host of different projection systems as part of his pre-season bonanza.

http://www.replacementlevel.com/index.php/P7/

He ran PECOTA, ZIPS, MARCEL, CHONE 1000 times each and posted the results. The SABR types among you might enjoy it.

From what I can tell, most of the projection systems are a lot harder than the scouts are on Schilling, Beckett and Matsuzaka. Most of the statheads I read seem to lean toward the scouting and believe that the Sox troika will outperform most of their projections.

H.B., yeah, the moment I commented, I realized it wasn't you speaking, but Bill.

'Amnesty International is all up in ahms'

And why would they not be:

http://www.smh.com.au/news/world/hicks-unlikely-to-break-gag-order-mcleod/2007/04/04/1175366320724.html

Well that about wraps it up for the first amendment. Huh? Sorry, what I meant to say was:

I predict Josh to be a total jock today. Sox to be like playing 500 ball by the time I wake up.

たくさんの為

I don't know, projections just don't excite me anymore. What was once something for the smahties to show off with, something you could really hang your hat on, now seems like a gag any average joe with a calculator can spin doctor to suit their own homer-tendencies. The other day I saw a Bill James projection of Dice-K going 19-2 w some miniscule ERA... now do I believe it because it's Bill James, grandfather of stat projection, or do I discredit it because the man's on the payroll?

Anyone read Sam Walker's Fantasyland?

//projections just don't excite me anymore.//

Obviously, you didn't go see Ron Jeremy debate last night. :))

Well, good projections are pretty useful if you plan to be an effective talker-of-smack.

Look guys, it's real simple. As Tippett explains in the article, the projections for each player are based heavily on the last couple of seasons, which include Schilling's year of recovery, and Beckett's disaster last season, and Wakefield's injury-shortened effort last season.

The computer believes those immediate past performances are very relevant. That's how it comes up with 86 wins. If humans don't believe their relevance is high, then 86 wins is not very likely to be accurate.

Or let me put it another way: there are some who think Pedroia is a .200 hitter, because that's what he hit in his late season trial. Not being stupid, you guys know this is not a true measure of his skill level, but caused by a BABIP of .188, way below normal.

The computer accounts for this somewhat by using his AAA numbers, but cannot adjust for Beckett's underachievement, or the recovery from injuries, because it is not just impartial, but limited in what data it's allowed to use.

Hence, "86 wins". We'll see.

3rd place????

Seems about right.

Evan, you're making assumptions about:

- Beckett's "underachievement"
- Schilling's "recovery"
- Wakefield's "injuries"

If all go the same way (that's the most recent evidence), then 86 wins is about right.

At best, all those assumptions go the other way and they win 98 games.

How each of those turns out will determine the season.

3rd place will be the end of Tito, and possibly Theo as well.

Tito got a pass last year, because of the injuries, but even the injury depleted club should have been able to do much better than 14-14 against the D-Rays and Royals, two teams that lost a hundred games each. 21-7 would have put us within one game of the wild card.

And we just lost to the Royals again...

"3rd place will be the end of Tito, and possibly Theo as well"

I do not see this. While 2004 does not necessarily guarantee a lifetime pass for Tito and Theo, they did generate a bit more than a couple of year's worth of goodwill.

Not with what they spent this off-season. That was the Flutie. Still, another September like last and it's hard to argue that won't be the end. After all, if 90% of the 04 team can be considered expendable, why not the GM and manager?

Sorry to be off-topic, but does anyone have experience with using Slingbox? I thought I saw a post about it here last week, but I couldn't find it just now... If so, please email me at home with your opinions.
Now back to our regularly scheduled programming.....

I bank on statistics for my job. That having been said, most statistics are lies.

Let me elucidate that thought. Most statistics are great at showing you where you've been. I do an experiment with thousands of genes firing at once and I can look back and get a picture of what was going on within the system. I can use statistics about average on- and off-times, etc. Baseball 2006 can be evaluated for statistics. We can figure out where we lost and what needs to be fixed and who might be best suited to fix it because of how they did in another system.

What you can't do is extrapolate well from statistics. You can't take what happened in 2006 as a reasonable predictor for 2007. It's not a reliable system that's being evaluated. If I look at that gene experiment, it happened in a well controlled environment and if I reproduce all those conditions with a new batch of the exact same cells, it'll happen again exactly the same 99 times out of 100. That really doesn't sound like baseball now does it? In fact if my cell cultures get too old, I have to throw them out and start over again with younger cells from the same starter cells that have been frozen and waiting. If we had clones of our team from 2004 and thawed them out in Spring Training, then we might be able to use 2004 stats to best predict their 2007 results. Then again, we're not placing them in the same environment (unless we can convince all the other teams to thaw out their 2004 guys to have the greatest chance at winning it all again).

It isn't until a good portion of this season is in the can that we'll have enough current info to predict ahead and only a minor point in the future (like the Wild Card predictions from BP each year). Everything else is just "adjustment" on past statistics based on age, trend, etc. And there are tons of examples of guys bucking the "trends" (i.e. unknown rookies, injury, resurging veterans, steroids, etc). Each one throws a monkey wrench into the statistical extrapolation...meaning no good way to predict October based on April. No research statistician is gonna make bank on a prediction now from the system of an MLB season. I can't find any statistical predictions based on 2005 that put the Sox in 3rd place for 2006 and so it goes.

Statistics are fun and give us a great topic of conversation, but reality gets in the way of using them as some sort of crystal ball to see the future.

Yaz, they've had their couple year pass. This'll be year three. If the Sox miss the playoffs again with their expanding payroll, that goodwill just might evaporate.

FWIW, PECOTA was pretty comfortably the best projection system last year, and it still sees the Sox with the WC despite fairly conservative projections for the rotation.

PECOTA last year would have killed you at the betting tables for the over/under on wins in the AL East.

BP look back on 2006 PECOTA

In that article they claim to have been good at predicting 2003 and 2005, while failing at 2004 and 2006 (AL East anyways, and since this is a Red Sox blog, that's what I'm talking about). So 50-50? I'd rather flip a coin.

That having been said, I used the 2007 BP book and PECOTA to help draft my fantasy team. Why, if I just got done denying statistics have any predictability in baseball? Because in the absence of any actual predictability, I'll take the next best thing, potential predictability. Plus I find that they are more accurate about how a player is likely to do than a team. If a team is a sum of potential predictions...then each variance from prediction can easily sway the total too far from the total team prediction to make the prediction valid. If 20 players each drop 0.5 win shares over the season that they were predicted to contribute, your 100 win prediction just dropped to 90 wins. But if I'm trying to figure out which guy to draft for a fantasy team, being within 0.5 win shares of true is a pretty good prediction.

PECOTA was pretty comfortably the best projection system last year, and it still sees the Sox with the WC

If that's truly the case, then I disagree with lou. Let's just give the SOX the trophy and schedule the N.E. tour!

Joking of course. Nothing beats a cold beer on a sunny day watching the boys of summer. Hey, let's play two!

"Statistics are fun and give us a great topic of conversation, but reality gets in the way of using them as some sort of crystal ball to see the future."

I agree, Kaz. It's a semi-intelligent way of bragging about or defending your team. It keeps us busy between games.

What...no AI talk? Did Sanjaya win?

Or should I say...not lose...

BWF,
I have a slingbox hooked up. Haven't used it all that much yet. Just for a few STR games and Bruins games. It was kind of a pain to hool up but seems to work well now.

QUESTION:

Would any of you pay $1,750 at auction for this?

Same price would get you two seats in box 36 for opening day at Fenway off StubHub.

Jay- thanks for the input. My biggest fear is that I will be blocked by the IT police from using it, and I don't know if I want to be bothered trying.

Kaz- "I do an experiment with thousands of genes firing at once...." how WAS the Ron Jeremy debate anyway?

in re AI: Sanjaya's longevity proves that Howard Stern has the capability to submarine the entire concept of AI. I don't know whether to be sad or impressed.

FNL tonight, y'all. Tune in, 8pm, NBC. You won't regret it! Best show on TV...

Rob -
That $1750 would be well worth it if Circle, Lisa and Tara were involved. No doubt they would make the squad.

I don't think this is all due to the power of Howard Stern.

Sanjaya is a huge sensation in the Indian/Pakistani community here in the US and even is being followed as a page one story in India. While I don't think you can vote from a foreign country, I mention that to show how big a sensation he is.

You combine that with the tween crowd who thinks he's cute and votes for him, with the Howard Stern/vote-for-the-worst crowd, and you can see why he's where he is.

You're right, hb. Never underestimate the truly horrid taste of the tween. :)

This Sanjaya individual has even been making the news downunder - how fucked up is that!!

Let's cut the crap right now.

33 million people on average vote for AI.

Howard (notK) Stern is not even the most famous Howard Stern these days. Given the limited subscriber base of Sirius, published reports demonstrate that, on his best day, Howard (Not K) Stern has approximately 400,000 people CUMULATIVELY listening to him in ANY given week. How, then does he even move the needle?

By the way, he is on vacation this week, so regardless of outcome, he won't affect things.

The Media reporting on the relevancy of HS is like PHB saying that The Curse hangs over us.

HS mattered in 1989. Not today. Let it go.


lc

lc- point well taken, but I think the number is 33 million votes cast, not individual voters. People can vote more than once. So if those 400K people were each to vote for Sanjaya 10 times, that's 4M more in his bucket, which (as hb pointed out) combined with the tweens certainly tips his needle.

I actually like the idea that some kind of counterculture viral campaign can derail the world's most popular karaoke contest.

Forget Sanajaya, where's his delicious sister at? Get me some more o' that! (insert sexual harrassment growl noise.)

I don't watch AI, so maybe she's in the audience.

da kine-
yeah, I'm with you. Was watching AI a couple of weeks ago (the girlfriend wanted to- I swear :-), and his lovely sister was looking mighty, er, perky whilst cheering him on. Highlight of the season definitely

Let's try that again...

Extra Innings package to be on cable.

The comments to this entry are closed.

The Soxaholix eBook Spinoff

The captivating and long awaited Soxaholix eBook spinoff is finally available!

There's No Crying in Pocket Pool

cover

Purchase at Amazon.

T-shirt

Logo t-shirts now available, several colors, even pink.

'Soxaholix logo t-shirt