BtB Breaks Down NL West
Long portion about the Giants
Email
PrintComments
worst team in baseball imo
Matt Cain: He'll save children, but not the Dodger children.
"AT LAST I AM A PARENTS." - Buster
by jponry on Oct 20, 2009 10:21 PM PDT reply actions 0 recs
That seems kind of silly given the fact the padres, pirates, and the royals still exist.
Rafael Rodriguez: #8 on our list, tearing up Scottsdale, and has been alive long enough to see the Warriors make the playoffs once.
by BrianBokake on Oct 21, 2009 11:00 AM PDT up reply actions 0 recs
nuh uh
royals are better because AL
A’s are 9th in mlb because BILLY BEANE
/stats
"he walked 18; new league record! Struck out 18, another new league record! He also hit the sportswriter, the PA announcer, the bull mascot twice..."
by i did my job on Oct 22, 2009 4:04 AM PDT up reply actions 0 recs
The system must be flawed
the Giants have the second-highest clutch rating at FanGraphs (+29 runs).
YOU EAT YOUR DAMN EGGROLL
by heimy25 on Oct 20, 2009 10:44 PM PDT reply actions 0 recs
Given all of this, if we’re going to be consistent in how we read the numbers, I think the defense simply got lucky by ~45 runs or so.
Is it really that easy to say a team got lucky with 45 runs? That seems like an awfully high number to stamp “lucky” on.
by deuce deuce on Oct 20, 2009 11:27 PM PDT reply actions 0 recs
Merkin Valdez? Manuel Mateo? A rose by any other name...
by rotorueter on Oct 21, 2009 6:23 AM PDT up reply actions 0 recs
45 out of almost 700. Less than 10%. I don’t know… doesn’t seem all that far fetched.
Context, people. More context is good. Less context is bad. If you're willing to be reductive, then you're willing to be wrong.
by howtheyscored on Oct 21, 2009 9:31 AM PDT up reply actions 0 recs
23rd?
That must be some kind of typo. They actually meant 85th.
The very bad man traded my son...So now I'd like you all to meet my new son, Ryan "Aaron" Garko...Dammit it's just not the same!
by boonitez on Oct 20, 2009 11:36 PM PDT reply actions 0 recs
I don’t pretend to follow the math, but I’d agree that overall the Giants were much worse than their final NLWest standing. Only 23rd best seems a bit low, but not by a lot. And yet I agree with deuce deuce: that seems anti-statistical to simply assign “luck” to a large unexplained variable like that.
"The part of the roster where most of the money is spent, though, is on free agents and guys acquired through trade — guys Sabean did play a big role in acquiring. And they are not good. When you get 2/5 of a pitching rotation for free, you would think you could do better with $76 million than to field the league’s worst offense."
-Taliesin September, 2009
by Lyle on Oct 21, 2009 6:32 AM PDT reply actions 0 recs
Nonsense! Eugenio Velez was worth at least +50 runs by himself!
by xanthan on Oct 21, 2009 6:41 AM PDT reply actions 0 recs
he did? Well hot damn, I’m gonna go buy me some season tickets now!
by Fresburg on Oct 21, 2009 7:08 AM PDT up reply actions 0 recs
VROOM!
Who’s brain did you bring me?
Brain SabeanOranother.
by daveinexile on Oct 21, 2009 8:33 AM PDT up reply actions 0 recs
OT: but in the same scheme of things
Apparently the best pitcher in the post season isn’t Cliff Lee (3GS, 24.1IP,14H, 3BB, 2ER 20K) or CC (3GS, 24.2IP, 17H, 3BB, 3ER, 20K) but Vincent Padilla (2GS..almost 3…14.1 IP, 8H, 1ER, 2BB, 10 K)
I R 5
by say hey nation on Oct 21, 2009 7:34 AM PDT reply actions 0 recs
Fail. But in all fairness, he’s been the best postseason pitcher for the Bums.
So far. >-)
by esseffgeez on Oct 21, 2009 8:34 AM PDT up reply actions 0 recs
"Well,
everyone knows Custer died at Little Bighorn. What this book presupposes is… maybe he didn’t."
But seriously, what if BtB is… right?
I know we have all been flabbergasted by their ranking, but what if they are right? I have thought this team was going to be worse (record-wise) next year, but what if they are REALLY bad? Actually, I’m kinda hoping for it for personal reasons. Even without BtB’s analysis we know that this year was sort of fluky, but the Brass doesn’t know that, and since I’m sick of ‘em, I’d like to shove a giant, smelly pile of crap in their faces.
Ok, this is me in October. Ask me in April, I will probably say something different.
/auto-defenestrates
something something jhiat00 will swindle
Young Studs for Old Bats: The Brian Sabean Story
FREE KEVIN FRANDSEN!!! Member of the Frandsen 5% Club.
by Uribe nee Gonzalez on Oct 21, 2009 9:13 AM PDT reply actions 0 recs
But seriously, what if BtB is… right?
I think BtB is most likely correct. The only rational hope I’ve been able to come up with is the weights changes a significantly amount in a low run environment. Which might be true but it would take many seasons of low run defense and offensive teams in the same era to begin to flesh out that hypotheses and I don’t see that happening soon. And if it did happen I doubt it removes ammount of “luck” reduced would be enough to please mos of us here.
Who’s brain did you bring me?
Brain SabeanOranother.
by daveinexile on Oct 21, 2009 9:33 AM PDT up reply actions 0 recs
I don’t think BtB is right at all. If you take the “luck” component of our situational hitting/pitching out that others methods like Fangraphs have us pegged for, the Giants were still an 82 win team. BtB is putting us at 74 wins. That’s a HUGE difference. I think one of the problems they have is with our dominating individual performances. An example – when Lincecum starts, we win. When our 5th starter was on the mound, we lost. I suspect our run differential/expected run differential combining these performances would be negative, implying we should win at less than a .500 clip. However, we should win at around a .500 clip for those games.
For most teams, factors like this are negligible. We aren’t most teams. We have one good hitter (and a couple others that performed well – Torres/Uribe), and everyone else is legitimately awful. Truly, truly awful. Our pitching staff has some guys that dominate (Lincecum), Cain dominated but, on average, should only be good, and Sanchez/Zito were good while our 5th starter sucked it up hard. Bullpen was the same way – Affeldt/Wilson, who were the go to bullpen guys, were lights out – the rest of our bullpen not as much, but they were much less relevant to winning games.
Basically, I think a lot of their assumptions work for most teams because their stuff is based on league averages. The Giants are out on the very edge of extreme. I think their assumptions break down to a degree and underrate the Giants. I don’t think we were close to an 88 win team, but I think we were at least as far from a 74 win team…
by Missing Barry on Oct 21, 2009 11:44 AM PDT up reply actions 0 recs
also
I think they weight the AL/NL split way too much.
Sharlon Schoop - de favoriete Nederlandse honkbalspeler van McCovey Chronicles.
You always have to be one step ahead of your drunk friends
--Daisy Owl
by Viliphied on Oct 21, 2009 12:53 PM PDT up reply actions 0 recs
I am also guilty of ignoring people that give huge “adjustments” to one League over the other. I tend to view that as conformational basis. (I can remember a time when the NL was thought to be the stronger league and would see a lot of the same assumptions then as well.)
The reality is the DH really messes up the game for this because team have to construct themselves very differently and tends provide more runs in a line (after all that is why a DH was put in) and that effect gets magnified again because we are using a runs scored logarithm as a major part criteria with a data base line gathered from one of the most offensive periods of the game. Unfortunately I am not strong enough in the math to present a counter to their numbers to contribute to that discussion. So by just mentally zero out that part their estimate comes back towards 500 by 2-3 games. That is pretty darn close to what I thought this teams talent leave was anyways.
Who’s brain did you bring me?
Brain SabeanOranother.
by daveinexile on Oct 21, 2009 1:14 PM PDT up reply actions 0 recs
Fair enough. The one thing we both agree on is the ’09 Giants were working a model not used much in the last 15-20 years. Extremely low run environment on both sides of the ball. I suspect it throws their model a lot farther out of whack then has been previously thought. I am not a sure exactly how far out of whack it throws their model off so I haven taken conservative stance on that compared to you.
Who’s brain did you bring me?
Brain SabeanOranother.
by daveinexile on Oct 21, 2009 1:12 PM PDT up reply actions 0 recs
Again, you are ignoring league adjustments. You can disagree with them if you want, but you’re making it sound like BtB just randomly gave us 74 wins instead of 82, which you know isn’t the case.
HA HA HA LOOK AT ME I'M ALL HAPPY AND STUFF NO REALLY CAN WE STOP WITH THE COOKYMAN IS SAD JOKES?
:-) :-) :-)
by Cookyman on Oct 21, 2009 1:21 PM PDT up reply actions 0 recs
I don’t mean to make their value sound random that is not how I view their work anyways. At no point do I view the model as perfect but then nothing really is in single season baseball. Though I joked about it most of the season but I truly figured this as a 79-80 win team in talent level. So once I ignore the AL adjustment that is pretty close to what I thought we were at.
Who’s brain did you bring me?
Brain SabeanOranother.
by daveinexile on Oct 21, 2009 1:59 PM PDT up reply actions 0 recs
So perhaps I am guilty of my form of confirmation bias here.
Who’s brain did you bring me?
Brain SabeanOranother.
by daveinexile on Oct 21, 2009 2:00 PM PDT up reply actions 0 recs
I replied to Missing Barry. I think.
HA HA HA LOOK AT ME I'M ALL HAPPY AND STUFF NO REALLY CAN WE STOP WITH THE COOKYMAN IS SAD JOKES?
:-) :-) :-)
by Cookyman on Oct 21, 2009 2:14 PM PDT up reply actions 0 recs
My bad then.
Who’s brain did you bring me?
Brain SabeanOranother.
by daveinexile on Oct 21, 2009 2:41 PM PDT up reply actions 0 recs
I just figured out I either made a typo or math error, it should be 76 wins. That 76 wins is before the league adjustment, though. It’s 71 wins in their hypothetical, league adjusted setting. At least I think that’s what their calculations are saying. I’m not saying they “randomly” gave anything, it’s not like they aren’t fairly transparent about their process or anything and they show the inputs at various stages, I just don’t think their model adequately explains the Giants talent level. For instance, this is included in their explanation of the Giants:
A big part of that is fielding, though—the Giants rate here as the second-best fielding team in the National League at ~34 to 47 runs above average (bUZR = +53 runs, THT = +40 runs, Catching = -6 runs). The other 39 to 52 run difference, though, remains unexplained.
Given all of this, if we’re going to be consistent in how we read the numbers, I think the defense simply got lucky by ~45 runs or so.
We’re not talking about trivial numbers here. We’re talking about the Giants getting lucky – which we all concede when we look deeper at the numbers, but even after making our most reasonable guesses as to how lucky they are based on things we know, we’re left with a huge disparity to try to explain. The larger the disparity is, the less likely it’s simple random chance. We have a similarly sized disparity on the offensive side.
Now, I don’t know that it’s not luck. I have my questions regarding their methodology, but I don’t have definitive answers. I just see a system that was very, very wrong with the Giants. Maybe they really were just that lucky, but that’s an immense deal of luck that suggests to me there’s probably a better explanation. Furthermore, if you look at other ways of evaluating the Giants, they have us pegged as closer to a .500 team (for instance, I just did the math using Fangraphs WAR, if I did it right they have the Giants as an expected 80.32 win team based on their context neutral WAR). That suggests to me BtB’s model doesn’t adequately evaluate the Giants. I think we broke their system, even if their system works really well for most teams.
by Missing Barry on Oct 21, 2009 2:05 PM PDT up reply actions 0 recs
Well, you see, you did it again:
Furthermore, if you look at other ways of evaluating the Giants, they have us pegged as closer to a .500 team (for instance, I just did the math using Fangraphs WAR, if I did it right they have the Giants as an expected 80.32 win team based on their context neutral WAR)
Once again, WAR is not adjusted for AL/NL. What do you think is the difference between the AL and NL? 3 wins? 5? Adjust for that, and you get something like 75-77 wins, which is still a few wins higher than what BtB gave the Giants, but nothing crazy. It definitely doesn’t suggest that the “model doesn’t adequately evaluate the Giants.”
HA HA HA LOOK AT ME I'M ALL HAPPY AND STUFF NO REALLY CAN WE STOP WITH THE COOKYMAN IS SAD JOKES?
:-) :-) :-)
by Cookyman on Oct 21, 2009 2:31 PM PDT up reply actions 0 recs
It doesn’t matter that WAR isn’t league adjusted, because I’m using BtB’s numbers before they included their league adjustment as the comparison. Apples to apples. It seems to me you don’t think a 4 win difference in estimated talent level is a lot? It’s also a 12 win difference in actual vs. predicted, and that’s….extremely large to say the least, some of which we can easily explain, some of which….well, not so much.
by Missing Barry on Oct 21, 2009 2:41 PM PDT up reply actions 0 recs
I don’t know how you define “a lot,” but no, 4 wins for an entire team over 162 games isn’t as crazy as you’re making is sound.
HA HA HA LOOK AT ME I'M ALL HAPPY AND STUFF NO REALLY CAN WE STOP WITH THE COOKYMAN IS SAD JOKES?
:-) :-) :-)
by Cookyman on Oct 22, 2009 2:24 PM PDT up reply actions 0 recs
4 wins for an entire team over 162 games isn’t as crazy as you’re making is sound.
I think this quote isn’t specific enough for this situation. A team being 4 wins off a preseason prediction wouldn’t be a lot at all. A team winning 4 more games than their expected performance wouldn’t be a lot, either – all it takes is a good record in one run games or something lucky like that. What we’re talking about is estimates of expected performance, based on what actually happened. Two systems with the exact same information on what actually happened coming up with estimates 4 wins apart strikes me as a very sizable difference; a difference bigger than I’d expect to get from two different systems that are working properly.
by Missing Barry on Oct 22, 2009 2:55 PM PDT up reply actions 0 recs
But it’s not the exact same information – BtB included a more elaborate baserunning stat, catcher defense, and used tRA instead of FIP. And there are probably some other minor differences I don’t know about. Only if all those things don’t amount to 4 wins, then we have a problem.
HA HA HA LOOK AT ME I'M ALL HAPPY AND STUFF NO REALLY CAN WE STOP WITH THE COOKYMAN IS SAD JOKES?
:-) :-) :-)
by Cookyman on Oct 22, 2009 3:25 PM PDT up reply actions 0 recs
If they did add up to close to 4 wins (or even 2-3 wins), I’d agree with you. The problem is they don’t. The difference, as far as I can tell, in the stuff included in BtB and excluded in Fangraphs is less than 10 runs…
They’re really not that similar in methodology – BtB uses component stats to figure out how many runs they’d score and give up, and from there convert that to a winning percentage. My own wild guess (I really have no idea if it’s true or not) is most of the difference comes from the winning% conversion.
by Missing Barry on Oct 22, 2009 6:17 PM PDT up reply actions 0 recs
Why are they not that similar? wOBA is the hitting stat, UZR is the fielding stat, and some variation of DIPS is the pitching stat (like you mentioned below, the difference between tERA and FIP is small), plus some other minor differences.
Anyway, did you ask them about it? I’m sure they could explain why there’s such a big difference between their ranking and WAR in this case, or if there really is a mistake in their calculations.
HA HA HA LOOK AT ME I'M ALL HAPPY AND STUFF NO REALLY CAN WE STOP WITH THE COOKYMAN IS SAD JOKES?
:-) :-) :-)
by Cookyman on Oct 23, 2009 9:59 AM PDT up reply actions 0 recs
And in any case, WAR is pretty much identical to what BtB used, except it doesn’t include catcher defense, basrunning, and uses a slightly different pitching stat (FIP instead of tRA). If anything, a contradiction implies a problem with WAR, not with BtB’s method.
HA HA HA LOOK AT ME I'M ALL HAPPY AND STUFF NO REALLY CAN WE STOP WITH THE COOKYMAN IS SAD JOKES?
:-) :-) :-)
by Cookyman on Oct 21, 2009 2:34 PM PDT up reply actions 0 recs
a contradiction implies a problem with WAR, not with BtB’s method.
??
I used WAR as one example of an estimate that pegs the Giants production as substantially greater than BtB’s. I’ve had a conversation with owlcroft where he said his estimates had the Giants pegged at more of an 82 win team. I’ve also read a couple of other people that had the Giants pegged at more around 82-84 wins (though ofthand I can’t remember where, so feel free to dismiss this sentence if you want).
I’m also not sure why WAR and BtB are pretty much identical? The included components in BtB and excluded in WAR aren’t a very large factor, and don’t explain most of the disparity we’re dealing with here. +10 BR, -6 C, and a small (but I’m not sure what the exact number is, I think it’s around -10 or so) difference in FIP and tRA (3.93 FIP, 4.00 tERA).
by Missing Barry on Oct 21, 2009 3:00 PM PDT up reply actions 0 recs
getting lucky by 12 wins is getting lucky by 120 runs!
That’s huge! Either we just witnessed one of the modern marvels of the baseball world or there is some omitted variable bias there.
I do have to say that none of the stuff that jinAZ is using is an implausible way to evaluate a team on it’s own, but I do share in the bewilderment at such an extreme result.
Please hit better, Randy Winn.
by oldjacket on Oct 22, 2009 8:43 AM PDT up reply actions 0 recs
I think the problem we these rankings is that we expect them the be highly accurate when they’re actually got huge errors. That’s nothing against the systems themselves they’re really nice ways of working things out but we should take them as approximations rather than at face value.
Fangraphs did a nice post recently examining how WAR fitted the actual results. They found the standard deviation to be 6.4 wins. That means that our actual record was about 1 standard deviation high and the BtB rankings were around 1 standard deviation low, so nothing particularly shocking. I’d imagine the BtB rankings would have similar errors.
What that suggests is that the rankings are a nice construct and give you a general trend or approximation but they’re not a very precise measurement and so shouldn’t be taken as a “true talent level”.
Proud parent of Waldis Joaquin!
by GiantFan on Oct 22, 2009 10:53 AM PDT up reply actions 0 recs
Well, if BtB has a similar standard deviation with actual results we should be pretty close to being able to reject their system as a true estimate at the 95% confidence level…
As for standard deviation between BtB and Fangraphs WAR, I don’t think that’s an appropriate comparison because we’re looking at the difference in actual vs. predicted on one hand (and actual is influenced by factors we willfully ignore because we know they aren’t useful for predicting the future like “clutch hitting” and such), and the difference in two measures of expected performance on the other…
by Missing Barry on Oct 22, 2009 2:58 PM PDT up reply actions 0 recs
Giants ain't gonna be the same next season.
Unless Bowkermania is unleashed.
by AmorVincitOmnia on Oct 21, 2009 9:44 AM PDT reply actions 0 recs
IT’S OBVIOUS THE GIANTS ARE MUCH WORSE THAN A BASEBALL TEAM, MUCH MORE LIKE A REALLY LUCKY ALASKAN DOG SLED TEAM
Brian Sabean strongly encourages you to disregard the drudgery of your employment responsibilities and join him in the consumption of spirituous libations.
by satyricrash on Oct 21, 2009 10:08 AM PDT reply actions 0 recs
Awesome.
/auto-defenestrates
something something jhiat00 will swindle
Young Studs for Old Bats: The Brian Sabean Story
FREE KEVIN FRANDSEN!!! Member of the Frandsen 5% Club.
by Uribe nee Gonzalez on Oct 21, 2009 10:59 AM PDT up reply actions 0 recs
Worse than the Royals obviously
Chris Dominguez: Bringing dingerz back to The Bay (In a while)
SCIENCE
by CB30 on Oct 21, 2009 10:21 AM PDT reply actions 0 recs
I am thinking that one way to compare teams with simlarities in stats would be look at each game as a sample and calculated that stat (any really) for each game. Plot it on a curve and find the standard deviation. A better team would probably have a smaller std dev and would thus be more likely to sustain that, or that value is closer to their true talent than a team that has a large std dev. Example:
If the Giants and Padres have similar OPS’s, .699 and .701 respectively, can you really say who is better? Probably not. But if you treat each game as a sample and take the OPS of each game and plot it, you could take the Std Dev. It might be useless but if the Giants Std Dev was +/- 100 (meaning that 95% of the scores are within +/-100 points of .699) and the padres had a STD Dev of +/- 50 I would say that the Padres have a better and more consistent offense while the giants offense varies greatly and may have more Potential but has less consistency
I R 5
by say hey nation on Oct 22, 2009 9:22 AM PDT reply actions 0 recs
But that only deals with one side of the ball. How do you find comparible defense? And that is the rub with teams like the Giants. Maybe even toss the M’s a good bit before the Injuries ( say end of July) and Jays tell arround mid July and they had to start making roster choices for budgets and such. All three of those squads were defensiviely strong and don’t remember them ranking very high as either.
Who’s brain did you bring me?
Brain SabeanOranother.
by daveinexile on Oct 22, 2009 11:42 AM PDT up reply actions 0 recs
I don’t know, but I am thinking we need to account for “consistency” somehow
I R 5
by say hey nation on Oct 22, 2009 11:44 AM PDT up reply actions 0 recs
/nods
It is a rough go what they are trying to do. That, and their openness, is why their efforts has kept me interested and curious more agitated by the Giants’ results all season long.
Who’s brain did you bring me?
Brain SabeanOranother.
by daveinexile on Oct 22, 2009 11:55 AM PDT up reply actions 0 recs
Comments For This Post Are Closed
Secondary Sidebar
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2009 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
