BP Projected Standings are up
For those without a subscription, Wicked Good Sports Red Sox Blog has a screenshot of BP's projected standings up.
Some surprises:
The A's take the AL West with 82 wins. Their OBP also is 6th in the AL, a big improvement over the last place they sported last year.
The White Sox are projected to go from first to last in the Central, and the World Series Teams are both in 3rd place in their divisions. Computers obviously have no glue about how many wins Championships add the following season.
The Mets allow the fewest runs in the NL? Tim Redding makes that much of a difference from last year's rotation?
over 3 years ago
iamawesomer
83 comments
0 recs |
Email
PrintComments
maybe this is the year they get within 15 wins for the white sox.
i mean, statistically, it’s gotta happen sometime.
our schtick is old
they project the White Sox pitching to get over 100 runs worse.
I'm spent, just can't rosterbate anymore.
I’m just saying that’s a lot of decline, I kinda think they’re going to miss again.
I'm spent, just can't rosterbate anymore.
Well they certainly won't be 15 wins over.
So yeah, I think under is the only possibility.
by Brendan Scolari on Feb 10, 2009 9:22 PM EST up reply actions
They pretty project the Mets pitching to duplicate last year’s RS, but they project a league wide uptick in scoring by 1032 runs.
I'm spent, just can't rosterbate anymore.
They project the Rangers to win 73 games
and (as a Ranger fan) I can’t argue with them.
With just a little bit of bad luck the 2009 Rangers could lose 95 games.
I’m a little shocked at seeing the Angels below .500. They were 68-40 before trading for Tex and also didn’t have Lackey the first 6 weeks of 2008.
"Dying ain't hard. It's living that's hard."
Am I the only one
who has a problem w/ a website taking subscription material and posting it on their website? This is the reason BP charges for their service and that blog should respect BP’s copyright.
"free" does not mean "free to grab and post on my website".
at least they screen grabbed the copyright at the bottom, too.
our schtick is old
Uh, they linked to it and it wasn’t that much data. Not a problem at all. I’m also pretty sure BP likes the hits and publicity they get out of it.
ah, yes. they linked to it. and it wasn't that much.
i can see you passed copyright law with flying colors.
our schtick is old
So your saying it was wrong that BtS did this then, right?
http://www.beyondtheboxscore.com/2009/2/5/751161/however-other-observers-se#comments
my fault
I have a subscription and just assumed, based on iamawesomer’s first statement — “for those w/o a subscription” — that one had to have a subscription to see this and therefore, posting it would violate copyright. Since it’s free content, problem solved. My mistake.
I’m disappointed with Pecota’s projections this year. They don’t really seem to be done very carefully. They have Citizens Bank Park in Philly playing as a pitcher’s park (EqAVG,EqOBP,EqSLG) >> (AVG,OBP,SLG) for Phillies hitters. They also flat out didn’t bother to do one for Jamie Moyer yet.
On top of that, their insanely bullish on the Mets and Braves. Admittedly, I’m coming with a Phillies fan slant, but I just mean compared to the other projection systems. Some examples:
Carlos Beltran: BABIP of .314 despite ranging from .263-.300 over the last 5 years for BABIP and being 32 years old.
Jose Reyes: BABIP of .330 despite never being above .320 and every other projection system sitting him closer to .310 for BABIP.
Carlos Delgado: BABIP of .303 despite being between .276-.290 for his three years with the Mets and being 37 years old this June.
I thought PECOTA was all about regression to the mean for above average players but it doesn’t seem to do so for lots of their hitters.
For the Braves: Chipper Jones: .361 BABIP? Last year was the only year he’s ever done that and he’s 37. They’re also far more bullish on a lot of Braves hitters than other systems. There seems to be almost a 10-win cumulative discrepancy between what PECOTA says and what ZIPS says about the Braves.
I haven’t looked into their numbers for the Angels but those really do seem way off. Just looking at the NL East, it seems like nearly every projection system has the Phillies about 90, the Mets about 88, the Braves about 84, and PECOTA really seems to have some weird numbers. I’m biased, but I’m usually pretty grounded and tend to be more bearish about Phillies players than a lot of the projection systems this year. I don’t even think PECOTA is way off there, but I just have a hard time seeing the Mets or Braves being anywhere near as good as they predict.
You probably know better than I do
but PECOTA also regresses to “player type” based on size and handedness. Whether that’s the best way to do it or not, I don’t know, but that’s one thing that’s different about it…
FOr another shock, look at PECOTA’s projections for Cristian Guzman and Russell The Muscle Branyan.
Bringing you more-or-less replacement level analysis and commentary since sometime in 2008.
by Matt Klaassen on Feb 10, 2009 5:22 PM EST up reply actions
It's February.
Stuff will change. And I’m not sure what would make PECOTA less “carefully done” one year than in other years.
I thought PECOTA was all about regression to the mean for above average players but it doesn’t seem to do so for lots of their hitters.
Not really.
I think PECOTA is giving the Rangers
much more credit than they’ve earned.
"Dying ain't hard. It's living that's hard."
Maybe the Mets are gonna clone Santana.
Because, that would, you know, help them make the playoffs.
by Brendan Scolari on Feb 10, 2009 9:13 PM EST up reply actions
Is defense incorporated into these projections?
Apparently they released initial depth charts last year, and then updated the projections about a week or two later to include defensive efficiency.
by CapgrasDelusion on Feb 10, 2009 6:17 PM EST reply actions
Any projection system that doesn't include a third of the game is inherently flawed.
by Aaron Campeau on Feb 10, 2009 7:19 PM EST up reply actions
PECOTA's methodology has a lot of problems.
It’s not without its uses, but I don’t think anyone should take it all that seriously.
by Aaron Campeau on Feb 10, 2009 9:33 PM EST up reply actions
Point me towards a more reliable projection system.
I’ve seen maybe MGL and that’s about it.
Also, please detail the problems in the methodology and who arrives at better conclusions with different approaches.
I don't know of any formal league-wide sims that are any better than PECOTA
but I think CHONE is better on an individual basis.
In reality though, being better than other systems doesn’t amount to a hill of beans if the system is flawed. PECOTA doesn’t value defense properly, it doesn’t value walks properly, it completely whiffs on certain types of players and it inflates runs scored numbers. I’m not saying no one should pay any attention to it, people should just be aware of its limitations.
by Aaron Campeau on Feb 10, 2009 11:32 PM EST up reply actions
Could you give more details?
Some explanations of why and how?
I’d also be interested in evidence as to why CHONE is better on an individual basis. Something I can easily believe but I need something ressembling proof before I take one absolutely over the other.
And if something is right more often than anything else, that amounts to a lot more than a hill of beans.
My point is that being right more often than things that aren't right very often isn't that impressive.
I’m not saying no one should pay any attention to PECOTA. I’m saying people should take the projected standings with an enormous grain of salt and recognize its limitations.
by Aaron Campeau on Feb 10, 2009 11:47 PM EST up reply actions
But I would like to find what forces me to take what it says with the smallest grain of salt possible.
Since PECOTA is right more than me, or you, or anyone I know (except maybe this newfangled MGL thing), I will believe it has a better chance of being right than any of you, them or me.
Therefor, worth a hill of beans.
I would wager that a DMB simulation using CHONE and regressed UZR projections would prove to be more accurate than PECOTA.
by Aaron Campeau on Feb 11, 2009 12:03 AM EST up reply actions
Cool.
You should test that. I would think you could even go backwards with something like that to be more thorough.
I’d be surprised if this hasn’t already been done and published, though.
As someone who’s in a DMB league, though, and gets frustrated by their simplistic defensive “ratings,” how would UZR work?
It would be a ton of work to be accurate, which is probably why no one has done it.
by Aaron Campeau on Feb 11, 2009 12:15 AM EST up reply actions
I would argue that you shouldn’t be disappointed with the defensive ratings (unless you’re talking about separating out a double play rating or something, which is different). Having 5 categories, from EX to PR, is actually a fair mathematical representation of the errors inherent in measuring defensive performance. We simply don’t have the tools that give us the granularity that one needs to do much else than put players into tiers – the probability that an offensive player projected to be 5 runs better on offense will be actually be better is far higher than what you would get with a similar defensive evaluation.
--
Dan Szymborski
dan@baseballprimer.com
by D.Szymborski on Feb 11, 2009 9:02 AM EST up reply actions
I don't think PECOTA really tries to do anything with fielding.
And if its data is converted to EqA or something like that, that’s an EqA problem, not a PECOTA problem.
If you only require PECOTA to do things that something like CHONE is doing, it’s basically just as effective.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
Projected Standings are tough not because of the projections, but because of the unknowns
I think I remember seeing a post somewhere where all the major projection systems all where with in a few percentage points of each other in terms of accuracy for individuals- including Marcel.
That may have just been for offense though, I’ll see if I can find it.
On the team level, the hard part is predicting the injuries and the fact that the “misses” (either high or low) may not be evenly spread between teams.
by Dan Turkenkopf on Feb 11, 2009 8:14 AM EST up reply actions
Nothing out there is significantly better...
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
by Sky Kalkman on Feb 11, 2009 12:55 PM EST up reply actions
Except defense isn't a third of the game.
Its maybe 10-15 percent.
by Brendan Scolari on Feb 10, 2009 9:14 PM EST up reply actions
Was hyperbole, but I would disagree with your statement nonetheless.
by Aaron Campeau on Feb 10, 2009 9:33 PM EST up reply actions
Really?
How much of the game do you think defense makes up? Lets say that scoring runs is 50% of the game and preventing runs is 50% of the game. That 50% run prevention can basically be divided into pitching and defense. I don’t think you can say they are equal value, pitching is like 35-40% and defense is 10-15% sounds about right to me. Otherwise pitching is not nearly as important as its made out to be if its less the 1/3 of the game.
by Brendan Scolari on Feb 10, 2009 10:44 PM EST up reply actions
It really depends on the player and the construction if the team.
A team with an average pitching staff and a lineup full of offense first players is going to derive more than 50% of its value from offense, and a team with an average pitching staff and a lineup full of defense first players is going to derive more than 50% of its value from run prevention.
by Aaron Campeau on Feb 10, 2009 10:55 PM EST up reply actions
I don't understand what the problem with that statement could be.
by Aaron Campeau on Feb 10, 2009 11:34 PM EST up reply actions
If you're asking me to admit I didn't run a thorough study on this, okay.
I didn’t run a thorough study on this. If you would like to do so, or point me towards one done by someone else, feel free and if it turns out I am wrong I will cheerfully admit it.
by Aaron Campeau on Feb 10, 2009 11:49 PM EST up reply actions
Tango's done the breakdown
Although I’m still not sure I agree with his approach / results (and lucky for me I’m not the only one).
Take a look at this thread, starting around post 16 (although the earlier stuff is good too).
by Dan Turkenkopf on Feb 11, 2009 8:21 AM EST up reply actions
That doesn't change the fact that defense makes up only 10-15% of the game
Just because you have a team that is good at preventing runs, that doesn’t mean the offensive side of the game goes away. You’re still going to have trouble scoring.
by Brendan Scolari on Feb 10, 2009 11:39 PM EST up reply actions
I'll put it another way.
Two teams have exactly the same true talent level. They have pitching staffs with exactly the same skill sets. Team A derives its value by scoring more runs while the other Team B derives its value by preventing more runs. Would you argue that defense is not more valuable to Team B than it is to Team A?
by Aaron Campeau on Feb 10, 2009 11:43 PM EST up reply actions
Well, no. The induvidual players are more valuable.
I’m a Dodgers fan, so I’ll use the their team last season as an example. Our pitching was really good, (fewest runs allowed in the NL I think), but thats because Billinsgley, Lowe, Broxton, etc. were really good. The reason our team wasn’t great is because scoring runs is 50% of the game too, and our offense was mediocre. So just because our pitchers were very valuable, doesn’t mean preventing runs is more important than scoring runs.
Anyway, the point is that the projections take into account a large part of what makes up teams, that being hitting and pitching. Defense is not nearly as important as those two, as its only going to move teams a few wins either way. So I think a projection that doesn’t include defense is fine, if not completely finished.
by Brendan Scolari on Feb 10, 2009 11:57 PM EST up reply actions
If BP's defensive metrics have improved, this would be hugely important.
If not, not so much.
by Aaron Campeau on Feb 11, 2009 12:08 AM EST up reply actions
Are you talking about FRAR?
Or DER?
And if their defensive metrics are so terrible at projecting the season, why is no one more reliable?
Because no one else has used more reliable projections as part of a larger system.
by Aaron Campeau on Feb 11, 2009 12:16 AM EST up reply actions
You didn't answer my first question.
And who is it that has the better defensive projections that aren’t used in a larger system?
Why doesn’t anyone? How do you know this?
BP's defensive metrics are all pretty bad.
And at this point I think you’re just being argumentative for the sake of being argumentative.
by Aaron Campeau on Feb 11, 2009 12:21 AM EST up reply actions
I'm aware of the problems with FRAR.
What are the problems with DER?
And I’m being argumentative for the sake of geting something tangible out of you.
Scoring more runs than the opponent is 100% of the game.
If you manage to do that in large part through defense, and projection system that doesn’t take that portion of the game into account is inherently flawed.
by Aaron Campeau on Feb 11, 2009 12:01 AM EST up reply actions
But who does that?
Defense just isn’t important enough to be the sole (or main) reason your wins. The projections don’t take into account baserunning either, but you’re not arguing that. The reason is because its a small part of the game, just like defense, albeit baserunning is even smaller.
by Brendan Scolari on Feb 11, 2009 12:12 AM EST up reply actions
Defense is vastly more important than baserunning.
by Aaron Campeau on Feb 11, 2009 12:15 AM EST up reply actions
Is it?
http://www.baseballprospectus.com/statistics/sortable/index.php?cid=421535
Thats a link from BP’s page of player baserunning values. The leader is Ichiro at +12.6 runs and the last guy is Dioner Navarro at -8 runs.
Defense does have somewhat bigger disparities, but a 20 run difference is nothing to sneeze at.
by Brendan Scolari on Feb 11, 2009 12:26 AM EST up reply actions
The key question isn't the maximum spread but the shape of the curve
Contrived example, but assume you had one player who was +10, one who was -10 and everyone else was 0.
Would it be worth including in the projections? The biggest impact it could have would be just over 1 win for a given team.
If instead you had half the league at +10 and half at -10, and they were unequally dispersed, then it probably makes sense to include.
by Dan Turkenkopf on Feb 11, 2009 8:25 AM EST up reply actions
Thats true
I don’t think baserunning is nearly as important as fielding, but it definitely has an effect. The point is neither of them make that much of a difference, so its not a big deal if preliminary projections don’t include them. I’m not sure how you find the standard deviation but I’d be interested to know.
by Brendan Scolari on Feb 11, 2009 1:16 PM EST up reply actions
A little birdy
A little birdy told me they have the DBacks scoring 60+ runs more than last year all with a lower obp/slg than last years team, and that they have Garland with as many quality starts (yeah I know) as both Webb and Haren combined. Seems odd.
vr, Xei
60 runs more without Hudson or Dunn (most likely)
That Justin Upton breakout must be REALLY big!!
by Brendan Scolari on Feb 10, 2009 9:15 PM EST up reply actions
It probably will be
But not enough to make that projection right.
vivaelbeñsheets
by vivaelpujols on Feb 10, 2009 9:18 PM EST up reply actions
That was a joke
but i’d be surprised if he really becomes a star by next year. And I mean that relatively. He might be capable of a year like his brother has had, but I don’t think he’s gonna go all Pujols on us.
by Brendan Scolari on Feb 10, 2009 9:20 PM EST up reply actions
Comments For This Post Are Closed
Secondary Sidebar
User Tools
Related FanPosts on Beyond the Box Score
FanPosts
Community blog posts and discussion.
Recent FanPosts
Follow us on Twitter!
BTB: @BtBScore
Justin: @justinbopp
Satchel: @SatchelPrice
J-Doug: @RationalPastime
Josh: @garik16
Julian: @giantsnirvana
Bill: @BillPetti
Lucas: @DBITLefty
Jeff: @jeffwzimmerman
Lewis: @LewsOnFirst
