UPDATED post re. small sample size
by Cannonballden on Dec 17, 2009 2:07 PM PST
183 comments
UPDATED post re. small sample size
Sabean seems to have little reasoning other than "AH LIKEZ DIS GUY HE GUD!"
Given he recently non-tendered Garko, his motivation in pursuing the same player for more money is mysterious.
I’m pro Johnson like the next guy (um, Nick Johnson that is), but I can appreciate a GM's decision to cut one guy like Garko given, among other factors, an undesirable performance in a small sample size, and pay (a reasonable amount) more to another with similar career stats who he thinks may perform better in the future.
Not revolutionary obviously, but I think that results shown in small sample sizes are too often dismissed as insignificant and irrelevant.
Whether or not they're significant is one interesting question. Was his 127 PAs insignificant?
Marcello had a very helpful link that may go unnoticed b/w all the inflammatory remarks (mostly mine - sorry).
and I don't want to quote the whole thing, but here's the breakdown of how many PAs it takes for generally accepted significance/correlation to kick in:
The question asked most often with regards to small sample sizes is essentially – When are the samples not small anymore? As in, at what juncture does the data become meaningful? Martinez at 4-10 is meaningless. Martinez at 66-165, like he is right now, tells us much, much more, but still is not enough playing time.
. . .
Without further delay, here are the results of [Cutter's] article as far as when certain statistics stabilize for individual hitters:
50 PA: Swing % 100 PA: Contact Rate 150 PA: Strikeout Rate, Line Drive Rate, Pitches/PA 200 PA: Walk Rate, Groundball Rate, GB/FB 250 PA: Flyball Rate 300 PA: Home Run Rate, HR/FB 500 PA: OBP, SLG, OPS, 1B Rate, Popup Rate 550 PA: ISO
http://www.fangraphs.com/blogs/index.php/when-samples-become-reliable/
I THINK the article basically said that a .70 correlation is a generally accepted significance level in baseball studies (unlike .95 or .975 which is generally accepted probability significance in science).
Apples and oranges though I think. .70 is a correlation, not a probability.
Really, we (I guess I if I boned up on p-values and other statistical methodology long forgotten) should be able to figure out if the difference in Garko's stats in 127 PAs was significant - in that it showed a true decline in his performance.
Theoretically, the larger the effect seen, the smaller the sample size has to be to be significant.
2.
The second question I wanted to raise (but did a bad job of raising), was whether or not despite the small sample size being statistically unreliable as a rule. Whether or not, as an exception it should be factored in to the decision of whether or not to tender (using Garko as only one example).
The best we can do is assume he’ll be at the average of the distribution, because that will maximize the amount of time we’re right.
And I'm arguing here that I want a GM that is trying to predict future performance in a way that is better than just what the player's total averages indicate. Obviously even Sabean (i hope) is looking at various splits which are smaller sample sizes, but hopefully large enough to be more informative than a player's total averages.
I think a good GM would use small sample sizes, and perhaps psychology and other reasoning in connection with those statistics, and maybe even a look in the guy's eye, to determine whether or not he will out-perform or under-perform his past performance.
I'll leave these home/away splits here also so as to not erase all the stats talked about below.
Point here is that while some of these stats may be just based on park differences - some of the difference in home/road has got to be due to comfort level in my opinion. For example, why would a RT handed (top-spin?) hitter like Rowan do better on the road and worse at home where Ishikawa, a LT handed hitter does so much better at home, a place known to be harder for LT handed hitters.
In an almost even # of plate appearances, Ishikawa had an OPS of .935 at home, .471 on the road.
Rowan had an OPS of .718 at home, .754 on the road.
Pablo had an OPS of 1.012 at home, .877 on the road. (tOPS+ 114 vs. tOPS+ 86 if you prefer)
Schierholtz had an OPS of .677 at home, .720 on the road.
Bengi had an ops of .856 at home, .610 away.
Winn:OPS .585 Home, .749 away and tOPS+ 75 home and 123 away.
Flew: OPS .802 Home, .674 away and tOPS+ 118 home and 82 away.
Renteria had an OPS of .615 at home, .653 on the road.
Velez’s numbers are almost identical .711 vs. .704.
Like I said, I don’t know why Garko stunk during his short time as a Giant, but I don't think it's insignificant, but may be wrong, and certainly don't think its meaningless.
This FanPost is reader-generated, and it does not necessarily reflect the views of McCovey Chronicles. If the author uses filler to achieve the minimum word requirement, a moderator may edit the FanPost for his or her own amusement.
0 recs |
183 comments
Email
Print
Comments
Sense
You should make some.
WHY IS BOCOCK?!
by Lars The Wanderer on Dec 17, 2009 2:21 PM PST reply actions 0 recs
oh
Matt Cain: He'll save children, but not the Dodger children.
"AT LAST I AM A PARENTS." - Buster
by jponry on Dec 17, 2009 2:22 PM PST reply actions 0 recs
You’re just blinded by your blind statistics rule and your ubiquitous use of view factors while considering Andrew Luck.
I don't know about that, to the groin.
by howtheyscored on Dec 17, 2009 3:42 PM PST up reply actions 0 recs
Well then
how about that?
Adopted brother of Jason Jarvis. To pass the time during the offseason I decided to try my hand at blogging about photography and music.
by j14 on Dec 17, 2009 2:23 PM PST reply actions 0 recs
I don't get it...
But if you think that 127 AB is incredibly important then look at the difference in these two lines…
A: .283/.355/.450
B: .279/.351/.441
That’s the difference between his career line before coming to the Giants and after. If you think 14 points of OPS is that big of a difference then I guess that’s your choice.
The Basil Fawlty Moderating Strategy:
"We could run a nice blog here if we didn't have all these members getting in the way."
Kevin Frandsen, come back!
by WalrusMan on Dec 17, 2009 2:34 PM PST reply actions 0 recs
I’m not sure where you’re getting your numbers, but this is what I found on baseball-reference…
CLE: .283/.355/.450
SFG: .235/.307/.330
by FPTV on Dec 17, 2009 5:42 PM PST up reply actions 0 recs
(i’m assuming you’re still talking about Ryan Garko)
by FPTV on Dec 17, 2009 5:42 PM PST up reply actions 0 recs
He’s talking about his career line before being traded to SF and his career line after his time in SF.
Matt Cain: He'll save children, but not the Dodger children.
"AT LAST I AM A PARENTS." - Buster
by jponry on Dec 17, 2009 5:44 PM PST up reply actions 0 recs
which is why its misleading
see mass confusion below
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 6:05 PM PST up reply actions 0 recs
It’s not misleading at all. He says very clearly what he means.
I don't know about that, to the groin.
by howtheyscored on Dec 17, 2009 10:33 PM PST up reply actions 0 recs
lolwut?
It's my blarg! Quick Pitch
And I tweet (more often than I blarg).
by can of corn on Dec 17, 2009 2:55 PM PST reply actions 0 recs
Stats aren’t everything unless they’re come in too small of a sample size? What the hell kind of argument is that?
Also, the correct answer is that Ryan Garko was never really that good to begin with.
Please hit better, Randy Winn.
by oldjacket on Dec 17, 2009 2:57 PM PST reply actions 0 recs
If you have to put a caveat in the thread title, you probably shouldn’t post it.
by KCE on Dec 17, 2009 3:07 PM PST reply actions 0 recs
You guys might all think I'm completely nuts - and I mean every last one of you - even giantsrainman and that honkyfella... BUT!
I don't know about that, to the groin.
by howtheyscored on Dec 17, 2009 3:48 PM PST up reply actions 0 recs
Adopted brother of Jason Jarvis. To pass the time during the offseason I decided to try my hand at blogging about photography and music.
by j14 on Dec 17, 2009 3:23 PM PST up reply actions 3 recs
you win the internet
These are my principles. If you don't like them, I have others. -Groucho Marx
by RDreamer on Dec 17, 2009 3:29 PM PST up reply actions 0 recs
YES! You found a use for my creation!
say hey nation is the Ralph Nader of McC.-Xanthan
by say hey nation on Dec 17, 2009 6:06 PM PST up reply actions 0 recs
That's pretty much what I do
I find ways to use work that someone else has done. It is how I got through college.
Adopted brother of Jason Jarvis. To pass the time during the offseason I decided to try my hand at blogging about photography and music.
by j14 on Dec 17, 2009 6:22 PM PST up reply actions 0 recs
But I’m a little surprised at the seemingly ubiquitous view here that blind statistics rule and that there are no other factors to consider (other than age and random luck).
That opinion has never been stated by anyone ever, anywhere, in the history of time and space and the Internet. Not here. Not there. Not even over there. Not even way back then. Not even in the future. Not even waaaay in the future. Not even waaaaay in the future over there. Never. Nobody. No one has ever said it. No one has ever thought it. No one has ever dreamed about it. No one has ever even expressed that accidentally.
It does not exist.
I don't know about that, to the groin.
by howtheyscored on Dec 17, 2009 3:36 PM PST reply actions 0 recs
but will you try green eggs and ham, howie-I-am?
by tyrannoman on Dec 17, 2009 9:15 PM PST up reply actions 0 recs
Absolutely not.
I don't know about that, to the groin.
by howtheyscored on Dec 17, 2009 10:33 PM PST up reply actions 0 recs
Wow - Pretentious and illiterate much?
Why do you all comment if you don’t even read?
Not once did I say that
127 AB is incredibly importantwalrus.
And are you trying to be completely misleading? 14 points is actually a BIG difference when you’re comparing his stats in 1,714 PAs to his stats with only an additional 127 PAs (1,841 PAs total). IF ya think about it, you’re trying to disprove a point by assuming its wrong.
And talk about not making any sense
I didn’t say stats are only
everythingwhen they
come in too small of a sample size
AGAIN, I was NOT saying that “small sample size = good”, I was saying that DESPITE the small sample size, you don’t completely ignore the stats.
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 3:37 PM PST reply actions 0 recs
+reply
I don't know about that, to the groin.
by howtheyscored on Dec 17, 2009 3:40 PM PST up reply actions 0 recs
Well, he made a Fanpost to reply to a comment on another one, so +reply would be a pretty big step.
I know you nerds know NOTHING about the real game of baseball, or any other athletic endeavor requiring teamwork under physical stress.
Mr. F! | comics | art | New Nattowear | Unofficial McImage Directory
by Natto on Dec 17, 2009 3:49 PM PST up reply actions 0 recs
A new FanPost is the mother of all +replies.
I don't know about that, to the groin.
by howtheyscored on Dec 17, 2009 3:49 PM PST up reply actions 0 recs
I was saying that DESPITE the small sample size, you don’t completely ignore the stats.
Maybe you should have saved the “stats aren’t everything!” diatribe for a fanpost that is not imploring us not to “look at teh stats!”
Please hit better, Randy Winn.
by oldjacket on Dec 17, 2009 3:46 PM PST up reply actions 0 recs
arg
not imploring us not to "look at teh stats!"
Please hit better, Randy Winn.
by oldjacket on Dec 17, 2009 3:47 PM PST up reply actions 0 recs
Non-ironic use of pretentious is ironic
Juan Carlos Perez, please start hitting.
by marcello on Dec 17, 2009 3:54 PM PST up reply actions 0 recs
I don’t think anybody ignores small sample sizes so much as they value the fact that small sample sizes are a part of bigger sample sizes.
I don't know about that, to the groin.
by howtheyscored on Dec 17, 2009 3:59 PM PST up reply actions 0 recs
/brain explodes
It's my blarg! Quick Pitch
And I tweet (more often than I blarg).
by can of corn on Dec 17, 2009 5:33 PM PST up reply actions 0 recs
Get that logic sh*t out of this thread. It obviously wasn’t asking for any.
"It's too late now."
by ResDog on Dec 17, 2009 8:42 PM PST up reply actions 0 recs
I just wanted a logic sandwich.
Joe Martinez: You are cool.
When it's all said and done, America will be remembered for three things: The Bill of Rights, jazz, and baseball.
by cornball on Dec 18, 2009 8:35 AM PST up reply actions 0 recs
14 point of OPS are never a big difference.
I was promised lasagna.
by Cookyman on Dec 17, 2009 5:57 PM PST up reply actions 0 recs
Really?
So a 15 year veteran has his AGGREGATE OPS slide by 14 points after one year and you don’t think its a big difference?
Aggregate OPS is what he was talking about.
He slid 168 points with the Giants. (granted. . . somewhat small sample size (about 1/3 of a season)
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 6:14 PM PST up reply actions 0 recs
I think the general consensus here is not that stats are everything, but that they are very useful for player evaluation. People who do not use them or understand them will have a much more difficult time successfully evaluating players. Sabean and Bochy do not understand modern stats, and largely because of this, they are not very good at player evaluation.
Osiris, Lord of the Dead, and... relief pitcher for the San Francisco Giants?
by neurofarm on Dec 17, 2009 3:43 PM PST reply actions 0 recs
Couldn't agree more.
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 3:58 PM PST up reply actions 0 recs
So, if you disregard small sample size, then how do you figure that he will keep on “sucking” for the Giants if they keep him? In order to adequately determine if a player is going to fit on your team, how in the world can you make that decision based on 127 at-bats?
You want some non-statistical factors? How about playing those 127 at-bats for a brand new team, facing pitchers you’ve never seen, playing in a park you’ve never played in, AND having the weight of Scott Barnes on your shoulders?
It’s not just that we expect him to get back to his career averages. It’s that they didn’t give him nearly enough time to decide if he was going to be a good piece of the Giants team.
STEVE HOLM! refuses to be the odd man out.
by UnleashTheGore on Dec 17, 2009 3:57 PM PST reply actions 0 recs
Ha ha. I don’t disregard small sample sizes. I not only have much regard for the smallness of small sample sizes (i.e. small sample size not as good as big sample size) but the whole POINT of the post was that small sample sizes shouldn’t completely be disregarded!
Again, (wow, again), I just DON’T think small sample sizes should be ignored.
There is a possibility that Garko would have continued to perform more like he did in the last 7% of his PAs and less like his previous PAs.
how in the world can you make that decision based on 127 at-bats?
just a factor Gore. as in a factor of many.
It’s that they didn’t give him nearly enough time to decide if he was going to be a good piece of the Giants team.
Sure. That would be obvious to even Bochy and Sabean. You think they wouldn’t want more time and a bigger sample size? They/we/you don’t have that option. Oh, unless of course you played him closer to 100% of the time he was here. But despite your hindsight, at the time the Giants were trying to win games and obviously didn’t think he was helping them do that.
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 4:16 PM PST up reply actions 0 recs
I'm confused...
He played like crap for the Giants. And despite the small sample size, that should still factor into the decision
Those 127 PAs gave us ALL some doubt about Garko compared to what we thought when the trade first occurred.
And that means something.
I think even a small sample size like that should factor into a decision
Even though you don’t directly say that you are disregarding the theory of SSS, you sure as heck are implying that Garko’s SSS struggles mean more than his career numbers. And what are these “factors of many” that made you realize Garko can’t make it back to his career numbers? The “feeling” you/Bochy/Sabean have when watching him play? Well, like I said before, there are any number of reasons he could have struggled, but none of them would keep a logical person from thinking he could or would get back to career averages.
But despite your hindsight, at the time the Giants were trying to win games and obviously didn’t think he was helping them do that.
So, they came to the realization he wasn’t going to help them win ball games by…what? Seeing how well he warmed the bench? It is their fault that he didn’t get the playing time he should have.
STEVE HOLM! refuses to be the odd man out.
by UnleashTheGore on Dec 17, 2009 4:51 PM PST up reply actions 0 recs
Oooooooook
Chris Dominguez: Bringing dingerz back to The Bay (In a while)
SCIENCE
by CB30 on Dec 17, 2009 4:26 PM PST reply actions 0 recs
seems like small sample size or early career stats make decisions for many here.
OK it was small sample size plus a loop in Todd Linden’s swing from one side of the plate. But we already have group think evaluation of Schierholtz/Bowker by their early career sample, and many players although not all, improve on their early career stats. It seems as if it is difficult for a player to keep improving for the majors at the speed of triple A, at some point during their critical development as a player they have to develop at the speed of the “show”. Sometimes too long a period of minor league development can stiffle a player’s development. Some folks seem to have to improve by phases, ie; A, AA, AAA, MLB etc. while others, Like Runzler seemed to breeze through all of them with similar success and probably could have skipped a few levels.
by bradleybear on Dec 17, 2009 4:31 PM PST up reply actions 0 recs
I do agree with this. Players need time to improve at the ML level. There are only a select few players that can come up & perform from the beginning. The poor SSS of the mediocre Giants prospect does not mean they can’t be productive hitters. But the Giants FO is more guilty of this than most McCoven.
"It's too late now."
by ResDog on Dec 18, 2009 7:57 AM PST up reply actions 0 recs
Obviously any links in the above post are probably NSFW
by jctGamer on Dec 17, 2009 4:33 PM PST reply actions 0 recs
Nice.
Officially disinterested in any high school outfielders from the state of Mississippi.
Proud adoptive parent of Sergio Romo. Looking forward to adopting Justin Smoak.
by Lyle on Dec 17, 2009 5:03 PM PST up reply actions 0 recs
Yeah, I said it. Whatcha gonna do about it?
Well, I read it. And now I’m gonna make a flip comment about how there seems to be a lot of "Me vs. the MCC" posts lately. Well, 3. But 3 is a lot compared to…um, a few months ago? Yeah, I’d say a few months ago.
And now I’m gonna feel obligated to say something about your post. So…um…this is about Garko and how…actually I’m not really sure what your point is. I keep getting lost upon reading it.
Okay, so I only read it once.
Anyway,
Those 127 PAs gave us ALL some doubt about Garko compared to what we thought when the trade first occurred.</blockquote
No. No it didn’t. Just like I didn’t doubt that Pablo Sandoval was a great hitter based on the beginning of his season.
The baseball Satanist
by thehavenot on Dec 17, 2009 4:54 PM PST reply actions 0 recs
Pablo’s case is a little different than garkos, so not much of a point, but. . .
look – nothing against MCC – but its ironic how defensive and unwelcoming MCC gets when minority or contrary views are raised.
The fringe seems to be the machine no?
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 5:06 PM PST up reply actions 0 recs
First, of all, any player who goes 127 ABs performing below their career levels is the same as Garko’s; you just have to cherry-pick stats that say they’re a bad player. I’m sure you could find 127 ABs where Bonds hit terribly, but the point is that 127 ABs are largley irrelevant compared to 1700 ABs when it comes to evaluating a player, so that stretch shouldn’t be isolated and used as an argument against retaining them.
The fact that Garko’s bad stretch happened to come while he was in a playoff push and after a trade isn’t significant when it comes to trying to evaluate his talent, because it’s just another cherry-picked sample size in the grand scheme of things. And that’s why giving up on him after such a small period is stupid (at the very least, they could have tendered him a contract and tried to trade him; non-tendering indicates they have almost zero confidence in his abilities, which of course brings up the problems with the FO that many have already pointed out).
And I don’t think MCC is being defensive or unwelcoming; you’re just wrong, and we’re pointing it out. Minority or contrary views are fine, but don’t expect everyone to lovingly attach themselves to the underbelly of your wrongness. In fact, you shouldn’t want that; if you’re wrong, it’s better to tell you so you can figure out why and fix it.
by quincy0191 on Dec 17, 2009 5:27 PM PST up reply actions 0 recs
Honestly, that really doesn’t happen the way you seem to think. I’ve read this place regularly for a long time, even if I don’t comment much. I’ve seen this argument come up before and it doesn’t hold water. The guy that drifts through, reads a bit, then throws up a huge fanpost loudly informing the rest of the commenters that he is more perceptive than they are, he is not a member of the lowing herd, and he knows that this stat thing isn’t so special – that guy doesn’t get received well. Is it because he’s right, and the folks here feel threatened by his swollen cranium and it’s uber-rightness?
No. What actually happened was the guy came in, read a bit, got steamed because their assumptions weren’t his assumptions, and he let off steam. Then he’s surprised when his obvious attitude of superiority is taken amiss.
So, yeah. Don’t be that guy.
I'm as tall as Mel - why can't I hit 500 home runs?
by Ott on Dec 17, 2009 5:29 PM PST up reply actions 1 recs
Pablo’s case is a little different in that there is even less information to go on than Garko, yes.
Anyway, the MCC doesn’t get defensive and unwelcoming when a minority or contrary view is expressed. The MCC gets unwelcoming when somebody shows either lack of thought or a disdain for the community or both. I think yours is a little bit of column B.
The baseball Satanist
by thehavenot on Dec 17, 2009 6:18 PM PST up reply actions 0 recs
no way. pablo’s case is different b/c he was essentially a rookie who showed huge potential in a significant number of MLB at bats (even though a “small” sample size), and rookies are expected to not be consistent, so the club has every reason to give him more time to work through his slumps. Also a lot cheaper. there are many reasons why that’s a horrible comparison.
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 6:27 PM PST up reply actions 0 recs
Hmm
If rookies are expected to not be consistent, I assume that means that veterans are supposed to be good all of the time. If this is the case, then someone should tell the Giants management because they have it backward.
Also, Garko was also cheap. Not as much as Pablo, but cheap nonetheless.
You gave me two bad reasons why that comparison is “horrible.” I think you might be overstating your case a bit.
The baseball Satanist
by thehavenot on Dec 17, 2009 6:35 PM PST up reply actions 0 recs
gotta run – very interested to hear why it’s a good comparison though
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 7:12 PM PST up reply actions 0 recs
Nope
Not gonna be sucked into this argument any further, thanks.
The baseball Satanist
by thehavenot on Dec 17, 2009 8:56 PM PST up reply actions 0 recs
Consistency is a myth. No player in the history of the game has ever hit at a similar level for an entire season, let alone his career. It doesn’t happen. Rookies are expected to improve their production, not their consistency.
Joe Martinez: You are cool.
When it's all said and done, America will be remembered for three things: The Bill of Rights, jazz, and baseball.
by cornball on Dec 18, 2009 8:39 AM PST up reply actions 0 recs
and I apologize for sounding superior or like my contrary view was better or revolutionary. Not my intention at all.
Also – no disdain toward the MCC community. I enjoy the MCC community and try not to speak up much b/c I actually assume that I’m NOT nearly as knowledgeable as some if not most out there.
but obviously you can’t expect all newcomers to spend hundreds of hours accumulating all the knowledge that has been disseminated on MCC right? That’s the exact kind of elitistcismism that I think is Ironic.
I thought I had a somewhat unique point of view that would make for an interesting discussion. Clearly I was wrong.
but definitely appreciate the constructive criticism
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 6:36 PM PST up reply actions 0 recs
but obviously you can’t expect all newcomers to spend hundreds of hours accumulating all the knowledge that has been disseminated on MCC right? That’s the exact kind of elitistcismism that I think is Ironic
I don’t think we do. When people sincerely question why somebody is dismissing their opinion and show an actual open mind instead of making it all about them, there’s no problem. Whenever ego gets involved, it turns into “You don’t like what I think, that means you think I’m stupid, that means you must think you are better, that means you are elitist/the machine/groupthinkers!” Obviously no one is going to be happy seeing that.
I think in your case Goofus’ post is a good indicator of the problem You took your disagreement with some people on another thread and made a fanpost about it. The only reason to make a fanpost about it, as far as I (and apparently others, too) can tell, is that you wanted to highlight how the MCC made you feel unwelcome. Of course we aren’t going to take such a fanpost very seriously.
Just think of it as water under the bridge and move forward. You can disagree. Just keep things like “If only you were able to understand better than Winn hit in 09” out of your posts. If you feel like you have been dismissed, be the bigger man (woman?).
The baseball Satanist
by thehavenot on Dec 17, 2009 6:50 PM PST up reply actions 0 recs
/hopes no one remembers his little exchange with E when he was a new member of the community…
The baseball Satanist
by thehavenot on Dec 17, 2009 6:57 PM PST up reply actions 0 recs
My first comment on these boards was telling all you jerks how stupid you were for saying that Matt Cain sucks.
I don't know about that, to the groin.
by howtheyscored on Dec 17, 2009 10:35 PM PST up reply actions 0 recs
My first post
was a rosterbation fanposts on a different account like 2 years ago,. Then I posted another fanpost this year about halfway through the year on this account, and actually started commenting after that.
The end.
#1 threat to America: Pandas
Also, Tim Lincecum
Official McPokeMaster
by GrahamCrakalaka on Dec 17, 2009 10:39 PM PST up reply actions 0 recs
My first post was a rosterbatory three-way deal involving Aaron Rowand, Milton Bradley, and Kyle Blanks.
Oopsie on the last player.
Lethargy
It has me
by dregarx on Dec 17, 2009 11:13 PM PST up reply actions 0 recs
My first post was a joke that had already run its course.
FORESHADOWING!
GROUGTHINK ALERT
The first Chester Arthur fanboy ever.
by groug on Dec 17, 2009 11:23 PM PST up reply actions 0 recs
I think my first post was arguing with rainman about Barry Bonds.
FORESHADOWING!
Matt Cain: He'll save children, but not the Dodger children.
"AT LAST I AM A PARENTS." - Buster
by jponry on Dec 17, 2009 11:37 PM PST up reply actions 0 recs
WHY DIDN’T THEY TRADE BONDS AND MORRIS FOR ADAM JONES AND WLADIMIR BALENTIN?
WHYYYYYYYYYYYY?
GROUGTHINK ALERT
The first Chester Arthur fanboy ever.
by groug on Dec 17, 2009 11:43 PM PST up reply actions 0 recs
Totally wrong
You took your disagreement with some people on another thread and made a fanpost about it
Actually, I think only 2 people disagreed with me, but no one was reading that post much and I thought it was interesting and wanted more comments/discussion.
When people sincerely question why somebody is dismissing their opinion and show an actual open mind instead of making it all about them, there’s no problem. Whenever ego gets involved
I’m actually very open minded to being proved wrong – that’s exactly why I wrote the post. I had a theory, and wanted to see what others thought.
But the first 7 posts weren’t exactly sincere questioning or criticism:
1.
Sense, You should make some.(and I’m the elitist closed minded person?)
2. “oh”
3.
Well then – how about that?
4.
I don’t get it…That’s the difference between his career line before coming to the Giants and after. If you think 14 points of OPS is that big of a difference then I guess that’s your choice.
Which I’ve explained I think is a really bad argument. disagree?
5. “lolwut?”
6. " Stats aren’t everything unless they’re come in too small of a sample size? What the hell kind of argument is that?" – total mis-read of my post
7. If you have to put a caveat in the thread title, you probably shouldn’t post it.
So yeah – I got a bit defensive and was definitely not the bigger man (I apologize Winn) but seriously, how many times do I have to repeat myself ?
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 7:04 PM PST up reply actions 0 recs
I read your whole post and had no clue what your point was. You’ve argued your case further in here and I’m still not sure what your point is. That’s on you.
Matt Cain: He'll save children, but not the Dodger children.
"AT LAST I AM A PARENTS." - Buster
by jponry on Dec 17, 2009 7:14 PM PST up reply actions 0 recs
For me it was that and the seeming incongruity of the poll attached to the post.
It's my blarg! Quick Pitch
And I tweet (more often than I blarg).
by can of corn on Dec 17, 2009 7:18 PM PST up reply actions 0 recs
From original post:
Note: I took most of this post from comments I made on another fanpost that were not warmly received.
From above:
You took your disagreement with some people on another thread and made a fanpost about it
Totally Wrong
Actually, I think only 2 people disagreed with me, but no one was reading that post much and I thought it was interesting and wanted more comments/discussion.
You took a disagreement from one place and made an entire fanpost about it, and addressed it to 99% of us.
There is something to be said for tone, but you could also work for more clarity. You seem to acknowledge that two posters is a SSS and shouldn’t apply to all the rest, but then the whole thing is addressed to 99% of the rest. You come off sounding combative and haven’t edited your thoughts in way that make it clear what exactly you are thinking/arguing. So the 99% of us are left only being able to respond to your tone, not your argument.
co-dad w/AfDC of
Ishikawa, the Topps Rookie All Star Team's First baseman. Does he get a chance in 2010?
by kennv on Dec 17, 2009 9:02 PM PST up reply actions 0 recs
But the first 7 posts weren’t exactly sincere questioning or criticism
Yeah, I know. But rather than examine yourself to see if maybe it was something you were doing to illicit such responses, you went on the attack. Rather than take a look at yourself to see if maybe you weren’t clear enough, you convinced yourself that it was everybody else who was wrong.
When you got seven responses showing confusion and/or misunderstanding, you determined that it was everyone else who had reading comprehension problems. Your post was confusing, and it had a somewhat combative tone.
Sorry if that doesn’t jibe with your worldview.
The baseball Satanist
by thehavenot on Dec 17, 2009 9:50 PM PST up reply actions 0 recs
not really – i already admitted i did a bad job w/ the post.
I was in a rush
crucify me.
oh wait.
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 18, 2009 7:49 AM PST up reply actions 0 recs
not really – i already admitted i did a bad job w/ the post.
Good. But not at first, which is when I was talking about. Anyway, glad you are someone who can look at himself and admit to some wrong here. You certainly have a lot more introspection than many people.
crucify me.
oh wait.
But you might want to stop being so sensitive.
You might want to stop being so sensitive.
The baseball Satanist
by thehavenot on Dec 18, 2009 9:45 AM PST up reply actions 0 recs
ha – just a joke.
(but yes – also a fault of mine. thank you ;)
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 18, 2009 11:55 AM PST up reply actions 0 recs
Okay
Well, apparently it flew right over my head. I’m sorry.
The baseball Satanist
by thehavenot on Dec 18, 2009 12:20 PM PST up reply actions 0 recs
but obviously you can’t expect all newcomers to spend hundreds of hours accumulating all the knowledge that has been disseminated on MCC right? That’s the exact kind of elitistcismism that I think is Ironic.
i didn’t spend hundreds of hours accumalating all that knowledge, hell, i barely cite stats. just lurk a bit more before posting
by sfoakbay on Dec 17, 2009 7:11 PM PST up reply actions 0 recs
this
I lurked for a few weeks before posting, and I picked up the stat stuff pretty quickly, and can apply it fairly well, if I do say so myself. And people rarely disagree strenuously with me, and never to the extent seen here.
by quincy0191 on Dec 18, 2009 1:22 AM PST up reply actions 0 recs
100% fake account. MCC gave it away
say hey nation is the Ralph Nader of McC.-Xanthan
by say hey nation on Dec 17, 2009 6:20 PM PST up reply actions 0 recs
not defensive and unwelcoming
just arguing your position with reasonable points, if anyone’s getting defensive, it’s you
by sfoakbay on Dec 17, 2009 7:08 PM PST up reply actions 0 recs
its ironic how defensive and unwelcoming MCC gets when minority or contrary views are raised.
The fringe seems to be the machine no?
It seems to me that you have worked actively to engender a response in which you are ridiculed rather than conversed with. Perhaps this entire fanpost was crafted with your desire to state the “irony” of how MCC is unwelcome to minority/contrary views. It would certainly explain your word choice:
Yeah, I said it. Whatcha gonna do about it?
But I’m a little surprised at the seemingly ubiquitous view here that blind statistics rule
…it seems like the “fringe” thinks that its almost guaranteed that with a large enough sample size, all players will regress to the mean, or fall back toward their career stats and that looking at other factors (other than age) is foolish.
Your purposefully inflammatory words engendered the expected ho-hum response, which allowed you to continue:
Pretentious and illiterate much?
Why do you all comment if you don’t even read?
And then there’s the small matter of your sig: “Every reform movement has a lunatic fringe.”
I find it ironic that you try to claim marginalization considering the fact that every post you make carries that deprecating addendum.
Perhaps you were trying to prove something about MCC. The only thing I saw proven in this sequence of comments was that the McCoven brook none of these kinds of shenanigans.
Lethargy
It has me
by dregarx on Dec 18, 2009 1:03 AM PST up reply actions 0 recs
Maybe it’s magic, the manager, the field, the seagulls, or just a lack of confidence brought on by a bad-luck slump.
All true. The point is nobody really knows. So we go back to look at and analyze records of actual events to see if we can spot trends, gain insights and increase knowledge. That is a solid, rational methodology. Carving up squirrels and sacrificing their entrails on an altar erected over second base to the gods of baseball is also a methodology. I don’t have proof that using advanced statistical analysis is better than that, but I think I can make a pretty good case that it is. The same holds true for using advanced statistical analysis over things like “gut,” “feel” and “baseball instinct,” all of which are unquantifiable and unverifiable. So, yeah, maybe seagull flight patterns are affecting baseball outcomes. But there’s no really strong, logical argument that can be made to indicate that it is, or that it’s superior to statistical analysis. Maybe magic exists. But there’s no compelling evidence (Trevor Hoffman’s change up notwithstanding) that it exists. Why should I believe in magic or the influence of seagulls over baseball outcomes or “knowing the game,” or “clutch hitting?” There’s no compelling, logical, tangible evidence supporting any of them.
Some players just play better in certain places and certain situations. Some players are better under pressure, some play better during the day, some are better in the playoffs, and others are worse.
Again, perfectly true. But we don’t know why that happens or when it is going to happen again. Statistical analysis at least asks those questions and proffers intelligible, logical reasoning as to why those events do (or do not, as memory and perception is terribly skewed by emotion) occur, and gives verifiable hypotheses (if any trends or associations can be identified) as to when they will occur in the future. Traditional baseball methodology is basically a form of gnosticism – a privileged few individuals, set apart from the rest, that possess an illumination that allows them to see the truth. And I’m not a baseball gnostic.
Basically you’re right that stats aren’t everything. But at least they’re leading to a) interesting questions; and b) verifiable theories based on records of actual events (even if the data collection isn’t perfect).
Hector Sanchez: Underrated. Fighting body bias since the 2009 off season. I still love you, son, even if you're fat.
by tedfordfan on Dec 17, 2009 5:00 PM PST reply actions 0 recs
Thanks TFT – it’s nice to mix in a thoughtful response every once in a while. I was hoping to get some interesting discussion going (not that testicles aren’t interesting).
And while I very much enjoy the humor on MCC, I don’t really appreciate the insults and personal attacks (or at least not at me?). HA. No biggie.
But for the record, I actually love stats and even moreso advanced statistics (though I’m a bit of a neophite) b/c oldschool stats don’t account for nearly as many confounding variables. In fact, I was trying to offer some statistics to prove my point (I guess I shouldn’t have used the subtitle “Stats aren’t everything” because that’s not really the point, the point is that statistics taken from a small sample size aren’t necessarily worthless and we shouldn’t expect players to always regress to their mean.
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 5:50 PM PST up reply actions 0 recs
It’s cool. Just keep reading, learning and asking questions. If that comes across in posts, people respect you.
It doesn’t hurt to be able to laugh at yourself as well.
Hector Sanchez: Underrated. Fighting body bias since the 2009 off season. I still love you, son, even if you're fat.
by tedfordfan on Dec 17, 2009 6:07 PM PST up reply actions 0 recs
Traditional baseball methodology is basically a form of gnosticism – a privileged few individuals, set apart from the rest, that possess an illumination that allows them to see the truth. And I’m not a baseball gnostic.
I love this analogy. It’s weird, though, because I sort of saw it the other way around; but technically the “stat gospel” is teachable (though not to me; I can’t do math. at all. at all. i bungle basic arithmetic without paper and pencil), while the instincts are not.
OTOH I always saw it as everyone has instincts but only few people have math skills and the ability to think in datasets. But I must be wrong, because I see a lot of people around me nowadays who do think in empirical, rational terms, terms which are and always have been alien to me.
... null, void, invalid, iniquitous, unjust, damnable, reprobate, inane, empty of meaning and effect for all time
by shanghaijim on Dec 17, 2009 9:04 PM PST up reply actions 0 recs
So
You loved your comments from another thread so much that you turned therm into a thread of your own?
ok
"We're in this thing!" My adopted Giant: "Raptor Jesus" Guzman, "Sweet Jesus" Guzman and Jesus H. Guzman.
by Goofus on Dec 17, 2009 5:03 PM PST reply actions 0 recs
Where’s the knee Stan Conte in the nuts option?
Utter frustration and futility.
by Johnny Disaster on Dec 17, 2009 5:41 PM PST reply actions 0 recs
After all if it was all about general past performance, then the stock market and being a GM would be simple and pretty UNinteresting. What I think is interesting (like I’m sure a lot of other do) is trying to carve up the past performance to determine how a player might under perform or over perform their past statistics. And that’s why you obviously look at recent success, splits etc. despite the small sample sizes.
Pathetic that I have to keep defending that point, but I obviously wasn’t clear in my post. And the fact that I have to keep saying that is sort of one of my points. Despite the emphatic statement by Howtheyscored, I think a lot of comments on this blog DO reflect an OVERvaluation of general past performance and the mean and a general dismissal of the dreaded small sample size. At least in the last 6 mo. that I’ve been following MCC, I think 80% of Bochy/Sabean criticism is that they rely on small sample sizes. BUT, there’s only so many at bats to give out, so everyone in our system can’t get 300 MLB at bats a year right? In the meantime, decisions have to made (and critiqued by MCC) based on predictions by the Giants management. Sometimes those decisions have to be made based on small sample sizes.
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 5:52 PM PST reply actions 0 recs
Yes, sometimes decisions have to be made on a small sample size. I assure you that almost everyone is aware of that. But that is not the case with Ryan Garko. He’s been in the big leagues for a while. He’s not particularly old, nor is he even in a park that’s exceptionally cruel to right-handed hitters.
So take your best shot, why is 127 ABs more meaningful than the previous 1000?
Please hit better, Randy Winn.
by oldjacket on Dec 17, 2009 6:06 PM PST up reply actions 0 recs
"Yes, sometimes decisions have to be made on a small sample size"
Seriously, when? I can’t think of an example.
I guess it depends on how you define SSS. But if a sample is not statistically significant, it’s NOT STATISTICALLY SIGNIFICANT. It doesn’t suddenly become significant because you really want it to, or because it’s convenient, or because you have a hunch it is.
I was promised lasagna.
by Cookyman on Dec 17, 2009 6:14 PM PST up reply actions 0 recs
judging minor league relievers, for one. Drafting high school talent for another. Judging fielding is one of those things where you would almost always want more sample than you can get.
I feel like most of those cases, you just supplement your stats with secondary knowledge. That’s why sample size is not a OMG STATZ UBER ALLEZ rule, it’s more like a “hey lets look at something other than just that stat” rule.
Please hit better, Randy Winn.
by oldjacket on Dec 17, 2009 6:20 PM PST up reply actions 0 recs
Relief pitchers should never be judged based on one year, and, among statsy people, never are.
High school players are not drafted based on their stats.
Nobody makes any decisions based on SSS UZR.
I was promised lasagna.
by Cookyman on Dec 17, 2009 6:31 PM PST up reply actions 0 recs
If Pablo Sandoval were in a more statistically minded organization, his one year UZR at third base would certainly be being weighted right now. That none of those situations should be judged solely on stats, (though stats may figure into the discussion) is precisely the point I was making.
Please hit better, Randy Winn.
by oldjacket on Dec 17, 2009 11:09 PM PST up reply actions 0 recs
I get that, but that’s not making decisions “based on small sample size”. 1000 PA’s can be split into 10 groups of 100 PA’s. Each group in itself, is a small sample size. Together they’re a rather large one.
I was promised lasagna.
by Cookyman on Dec 18, 2009 2:45 PM PST up reply actions 0 recs
High school players are drafted based on their performances in national talent showcases, as well as scouting opinions. Those showcases are very small sample sizes, yet every year there are kids that shoot up scouting lists because of how they performed in a weekend-long tournament.
Joe Martinez: You are cool.
When it's all said and done, America will be remembered for three things: The Bill of Rights, jazz, and baseball.
by cornball on Dec 18, 2009 8:43 AM PST up reply actions 0 recs
Different things require different sample sizes.
I was promised lasagna.
by Cookyman on Dec 18, 2009 2:47 PM PST up reply actions 0 recs
also: Rule 5 drafts, trading for prospects or players with few MLB at bats, drafting anyone who doesn’t have a large sample size at whatever position they’re playing. drafting anyone who doesn’t have a large or relevant (i.e. against MLB talent) sample size.
Also, it’s been a while since my stats class but can you tell me how many PA it takes for it to be “significant” scientifically speaking?
I’m arguing that 127PAs IS significant (obviously, if I think its a factor) – but I may be wrong…
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 6:43 PM PST up reply actions 0 recs
Rule 5 drafts,
Minor league stats and scouting reports.
trading for prospects or players with few MLB at bats,
Minor league stats and scouting reports.
Drafting anyone who doesn’t have a large sample size at whatever position they’re playing.
Hmm…are we talking about defense here? If so, then it has nothing to do with stats.
Drafting anyone who doesn’t have a large or relevant (i.e. against MLB talent) sample size,
Not based on stats.
I was promised lasagna.
by Cookyman on Dec 17, 2009 7:08 PM PST up reply actions 0 recs
This guy doesn’t think minor league stats mean anything? He just sounds more and more like sabean.
Adopted Giant: Henry Sosa
by raisingcain on Dec 17, 2009 7:12 PM PST up reply actions 0 recs
And as for your question, depends what you’re trying to check. 127 PA’s are almost a decent sample size for K%, but a terrible one for, say, AVG. Different stats need different sample sizes.
I was promised lasagna.
by Cookyman on Dec 17, 2009 7:18 PM PST up reply actions 0 recs
OK then
so answer his question specifically. He’s asking just how many AB’s it takes for a sample to be “significant.” How many for:
- batting average
-slugging average
-K rate
-BB rate
I’d be curious to know as well.
Officially disinterested in any high school outfielders from the state of Mississippi.
Proud adoptive parent of Sergio Romo. Looking forward to adopting Justin Smoak.
by Lyle on Dec 17, 2009 8:42 PM PST up reply actions 0 recs
Juan Carlos Perez, please start hitting.
by marcello on Dec 17, 2009 9:34 PM PST up reply actions 0 recs
Thanks, Marcello. I assume most of you stat-folk would agree with this?
I’ll cut-and-paste a copy of it, just to have handy for future conversations.
Officially disinterested in any high school outfielders from the state of Mississippi.
Proud adoptive parent of Sergio Romo. Looking forward to adopting Justin Smoak.
by Lyle on Dec 18, 2009 1:38 PM PST up reply actions 0 recs
I'll take the liberty of speaking in the name of the stat-folk:
Yes, at least to an extent (there’s no true, objective point, where stats suddenly become stable).
It’s why we keep on mentioning peripheral stats (K’s, BB’s, GB’s). They’re simply more reliable and predictive than the basic ones, at least in non-huge sample sizes.
Also, we all think Cookyman should be made King of the Internet.
I was promised lasagna.
by Cookyman on Dec 18, 2009 3:00 PM PST up reply actions 0 recs
Officially disinterested in any high school outfielders from the state of Mississippi.
Proud adoptive parent of Sergio Romo. Looking forward to adopting Justin Smoak.
by Lyle on Dec 18, 2009 3:39 PM PST up reply actions 0 recs
If only you were able to understand better than Winn hit in 09
I never said 127 ABs should be a BIGGER factor than the bigger sample size. BUT here’s why I said it should be a factor and not ignored in predicting future performance with the Giants:
127 PAs:
recent performance
with the Giants
in NL
against NL pitching
1000 PAs
Not most recent
not w/ giants
not in NL
not against NL pitching
again, not freaking only factor, not the most important factor, but a factor.
other factors?
1. Garko not cheaper than other options that MIGHT end up performing just as well (ishi bowker etc.)
2. Garko has not been and is unlikely to ever be pujols or berkman etc.
3. There’s others on the market that are likely to perform better than Garko (albeit for more $)
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 6:22 PM PST up reply actions 0 recs
i know i'm just picking apart one point in your post
but ishi has little chance of ever being as good of a hitter as garko, and defintely not next year, where garko could have been had for 2-3 million
by sfoakbay on Dec 17, 2009 7:15 PM PST up reply actions 0 recs
I know you said hitter
Garko WAR: .8, 1.6, .6, .8
Ishikawa WAR: .8
Yeah, it’s all fielding and its a small sample for UZR for Ishi, but gotta stick up for my boy.
co-dad w/AfDC of
Ishikawa, the Topps Rookie All Star Team's First baseman. Does he get a chance in 2010?
by kennv on Dec 17, 2009 9:15 PM PST up reply actions 0 recs
Now I’m confused about what we are even talking about.
Do you, or do you not think that it’s a good idea to weight 127 ABs in French vanilla more than a guy’s entire career?
Please hit better, Randy Winn.
by oldjacket on Dec 18, 2009 5:38 AM PST up reply actions 0 recs
You can draw conclusions from small sample sizes. They’re just more unreliable as the sample size shrinks. The frustration with Bochy and Sabean is that they appear to give those samples much, much, much, much, much, much, much too much importance.
That and they say stupid things in the press which indicate that they don’t understand, nor do they want to understand, some of the most basic insights that good statistical analysis brings to the table.
Hector Sanchez: Underrated. Fighting body bias since the 2009 off season. I still love you, son, even if you're fat.
by tedfordfan on Dec 17, 2009 6:10 PM PST up reply actions 0 recs
+ 1
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 6:44 PM PST up reply actions 0 recs
If you agree with that comment, why are you okay with the decision to let Garko go? I don’t see a logical consistency to those statements.
Joe Martinez: You are cool.
When it's all said and done, America will be remembered for three things: The Bill of Rights, jazz, and baseball.
by cornball on Dec 18, 2009 8:45 AM PST up reply actions 0 recs
It may seem illogical at first glance, but it’s not.
I do think smaller sample sizes are less reliable on a whole. . . and as a rule.
That doesn’t mean there useless and should always be ignored.
That’s sort of the issue I had a problem w/ that made me want to write this post – it would be interesting to discuss when people think there are exceptions to the rules.
(as opposed to discussing only what the “rule” is (based on advanced statistics etc.) – which I think is a good decision, but should be wrapped up pretty quickly)
Also, as far as me agreeing that sabean and bochy sometimes rely in error on small sample sizes:
again, just b/c they do it too often doesn’t mean they shouldn’t ever do it. Some players decline (and some player’s minor league statistics DON’T translate to MLB)
so as the actual team-management, they have to look at a smaller (more recent) sample size and determine whether or not its a decline or an aberration. I think they should have actually relied MORE on the small sample size so in some cases (Winn) and less so in others (Bowker), but maybe got it right in others (Frandsen).
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 18, 2009 8:56 AM PST up reply actions 0 recs
Who’s saying that small sample size should always be ignored?
I think most folks are saying that small sample sizes should usually be taken with a grain of salt. There’s a big difference between the two.
by Wonderful Terrific Monds on Dec 18, 2009 10:05 AM PST up reply actions 0 recs
To be fair, the first time he made his comment, he did get the following response:
100 ABs means ABSOLUTELY NOTHING, period.
by Missing Barry on Dec 18, 2009 10:47 AM PST up reply actions 0 recs
But they judge some players based on small sample sizes, for no reason whatsoever, and then keep throwing crappy players out there over and over and over. What is the reasoning behind it?
Your overall point does have some truth to it. Sometimes things change. Sometimes Melvin Mora becomes an elite hitter…for a while. Sometimes Sammy Sosa becomes one of the greatest homerun hitters ever. Sometimes mediocre pitchers like Chris Carpenter turn into elite hitters. And, of course, sometimes the reverse is true.
However, there is no reason behind the small sample size judgments. Why was Garko only given 127 sporadic (a key part of this issue) at bats while Edgar Renterria was and is the starting shortstop until his contract runs out? Why was Winn out there every single day, but Bowker or Schierholtz or whoever wasn’t given much consideration at all?
It’s this that annoys the crap out of me.
The baseball Satanist
by thehavenot on Dec 17, 2009 6:13 PM PST up reply actions 0 recs
Me too
Leaving in the "veteran""professional hitter" “gamer” drives me MAD (thus this posting:)
And your implied reasoning is that the prospect like bowker/frandsen etc. at least has a SHOT at figuring it out right? whereas the veteran who is doing poorly (based on a small sample size) could very well just be in his waining phase.
I would have pulled Winn and Renteria also b/c of that reasoning.
And maybe Garko’s career is just on the short side and he’s already in his decline. No magic rule that you have to be over 33 for that to happen right?
But what would you do? I don’t get it. would you have given Garko more at bats when you had 2 other players at that position who ALSO had only a small sample size of less than stellar performance (ishi/bowker)?
Especially when those other two are cheaper and younger players?
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 17, 2009 6:49 PM PST up reply actions 0 recs
If I traded for Garko, I would have definitely let him have more time out there than he had. It’s such a non-question in my mind that it is shocking how he was treated after they traded a decent prospect for him.
If I felt that he wasn’t that much better than Ishikawa (and I assume you are lumping Bowker in here too) then I wouldn’t have traded for him to begin with.
The baseball Satanist
by thehavenot on Dec 17, 2009 6:55 PM PST up reply actions 0 recs
God damn it! Who allowed Bochy an account?
say hey nation is the Ralph Nader of McC.-Xanthan
by say hey nation on Dec 17, 2009 6:09 PM PST reply actions 0 recs
I shutter to think...
…really eye dew…
"I don’t know why people feel the need to come up with reasons 'why' for everything..." - Missing Barry
by victor frankenstein on Dec 17, 2009 6:11 PM PST reply actions 0 recs
so much better than the shutter jokes I was coming up with…
Utter frustration and futility.
by Johnny Disaster on Dec 17, 2009 6:19 PM PST up reply actions 0 recs
Shutter you say?
I know you nerds know NOTHING about the real game of baseball, or any other athletic endeavor requiring teamwork under physical stress.
Mr. F! | comics | art | New Nattowear | Unofficial McImage Directory
by Natto on Dec 17, 2009 8:29 PM PST up reply actions 1 recs
Let's go over the wrong very carefully
First, it’s “shudder”, not “shutter”. I know victor frankenstein already pointed that out, but I technically wrote it first, and the novel that follows took some time.
Second, no one is using stats to the exclusion of everything else. Stats are simply more available, more detailed, easier to compare, and better at evaluating performance. That’s not to say that things like scouting, park factors, and psychology are totally irrelevant, but simply that they are far less important, because they are. Good hitters hit well in various environments, bad hitters hit badly in various environments. There are obviously exceptions, but it’s far better to base evaluation on the rule rather than the exception.
This leads to the third wrong, LaRoche>Garko offensively. Assuming that Garko is a bad Giant is a silly thing to do, because most players play the same irrespective of their team. For the sake of argument, let’s assume that it’s true, and we don’t know that because we’re assuming he’ll regress to his career levels. How do we go about finding out whether Garko is a bad Giant? Well, we ship him off to a lot of other teams, and if he does well with all of them, he’s just bad for the Giants. That is impractical, because it requires a high level of cooperation from other teams with little to no gain from them, as well as an overly complicated plan of action. So we can’t go that route. What if we assume that he’s just a bad Giant, and we know he’s doing badly because he’s a bad Giant? Well, what’s to say that LaRoche won’t be bad for the same reason? How can we be sure that any player will be a good Giant? We can’t, so signing free agents and bringing up minor league players introduces a great deal of risk, far more than is cost-effective for million-dollar contracts. We are therefore crippled, and unable to compete. Basically, assuming that players are or can be just bad for any given team is both illogical and impractical, so there’s no reason to do it.
Fourth, my fanpost is only a few days old. That’s not that old.
Fifth, doubt was not thrust upon all of us after 127 ABs. I certainly had small doubts as to whether Garko could be effective in a Giants uniform, but these were centered on the fact that he came from the AL, a more hitter-friendly league, and the lack of decent hitters in our lineup, which in turn could cause his production to drop. But I for one did not think that Garko had magically become worse after the trade, which ruins your theory right there. And while that still leaves open the possibility of many thinking Garko wasn’t the player here that he was there, I think the comments here show this to be false as well. So that doesn’t really mean anything.
Sixth, howtheyscored is right when he says that small sample size are relevant only in that they are part of large sample sizes, and judging a player on a small sample size when a large one is available is therefore silly, because larger sample sizes generally give a better indication of how a player will perform (again, not always, but let’s go by the rule, not the exception). Here’s the wOBA graph for Eugenio Velez:
As you can see, his wOBA generally sits around .300 (career .307). This is not a good thing for a bad defensive player, and definitely not for a corner outfielder. Eugenio Velez, over his career, has shown himself to be a bad player, exactly at replacement level over his career (0.0 WAR, -0.3 RAR). But if you were to cherry-pick sample sizes, say the end of 2007 and beginning of 2008, you’d find a much higher wOBA, over .350. That sample size says Velez is a good hitter, while his career numbers say he’s not. Given that career numbers>smaller samples when it comes to evaluating a player, a small sample size is not necessarily indicative of a player’s overall talent, and therefore should not be weighed over their career numbers, as smaller samples can vary wildly as we have just shown. (Yes, I know it’s fantastically ironic to use Velez as an example when Sabean/Bochy think he’s a starting LF, which is kind of the point).
Seventh, illogical=irrational. Reason=logic. They’re synonyms. So saying something is illogical but not necessarily irrational is just wrong.
Eighth, while the idea that players produce more or less under certain conditions is sound, this is typically due to things like park factors more than “being a bad Giant”. You then posted several discrepancies in OPS between home and away games for a lot of Giants, which are accounted for by park factors. Park-adjusted stats typically show players to be the same pretty much everywhere, though their stat lines can differ greatly in non-park-adjusted stats. The reason for this would seem obvious, and therefore we can say that AT&T was “made” for Ishikawa and not for Schierholtz. Ishikawa’s hitting strengths play better at AT&T than Schierholtz’s, so Ishikawa hit better at AT&T than Schierholtz. Saying "Don’t tell me that PacBell was “made for” Ishikawa but not Schierholtz." is like saying "Explain what makes a car go, but don’t tell me it’s the engine. It’s hard to explain something when a criteria is not using the explanation.
Ninth, players have to fall back to their career levels with a large enough sample size, because the larger the sample size, the more of their career is included, until their entire career is included and the stats match up perfectly because you’re comparing 10,000 ABs with 10,000 ABs. I think what you meant to say is that you shouldn’t assume that players will always continue to perform to their career averages, which is obvious, and no one is saying that we expected Garko to put up the exact same line here that he did with the Indians. But again, 127 ABs is not enough of a sample to indicate that he is regressing as significantly as he did here, and given what we know about other players, Garko shouldn’t be expected to continue his performance here last year. Now, if he were going from age 28 last season to age 42 this season, going from .285/.362/.464 to .235/.303/.330 wouldn’t be unreasonable, but the simple fact is that the vast majority of players do not regress that fast (Garret Atkins does, but again, we’re talking rule, not exception).
Tenth, your poll is clearly silly.
by quincy0191 on Dec 17, 2009 6:23 PM PST reply actions 6 recs
One of my favorite lines...
…from King’s Christine:
“Well, that just about covers the motherfucking waterfront, doesn’t it?”
"I don’t know why people feel the need to come up with reasons 'why' for everything..." - Missing Barry
by victor frankenstein on Dec 17, 2009 6:36 PM PST up reply actions 0 recs
8th
Do you really think Ishikawa’s home road splits are have that much to do with park factors rather than small sample size shenanigans (happened upon tougher pitchers on the road, slumped away, hot at home) and/or some psychological issue (comfort, routine, homesickness, nerves, etc).
There was a great post a few months into the season when Ishi was “struggling” by Fairweather Fan on what Ishi’s slugging would be if like three balls had been homers in almost any other park’s right center, where they instead turned into two outs and a triple. I don’t think ATT was “made for Ishi”.
co-dad w/AfDC of
Ishikawa, the Topps Rookie All Star Team's First baseman. Does he get a chance in 2010?
by kennv on Dec 17, 2009 9:25 PM PST up reply actions 0 recs
I think they have something to do with park factors and something to do with psychology and something to do with luck and something to do with hot streaks. I was just pointing out that saying that forcing an explanation that doesn’t have anything to do with park factors is an unreasonable demand, because park factors likely play a role.
by quincy0191 on Dec 17, 2009 11:22 PM PST up reply actions 0 recs
howtheyscored is right when he says that small sample size are relevant only in that they are part of large sample sizes, and judging a player on a small sample size when a large one is available is therefore silly, because larger sample sizes generally give a better indication of how a player will perform
This ignores the difference in time period and ability through our sample, though. Maybe you’re just oversimplifying things and you understand the concept, but obviously the more recent performance is more important. Small samples pretty much never lead us to conclusions, but they do tell us something – just not enough to make any real decisions on. So I guess my point is that it’s incorrect to say 127 PA’s don’t mean anything or must be part of a larger sample…even if practically that’s the case since 127 PA’s isn’t enough to give us enough confidence to really make a sound decision from….
by Missing Barry on Dec 18, 2009 12:21 AM PST up reply actions 0 recs
Yeah, I get that recent performances in smaller samples are valuable (obviously a 38 year old player’s age-37 season is more relevant than their 27-29 seasons, even though the latter contains more data), but I didn’t want to write every little nuance and exception; I figured it was already long enough. And 127 PAs mean something, just so little that non-tendering Garko based on his production in 127 PAs is kind of retarded. That data is significant, just a hell of a lot more significant as a part of his career numbers.
by quincy0191 on Dec 18, 2009 1:26 AM PST up reply actions 0 recs
Nice.
Completely agree. Stats aren’t everything, but they are the easiest and most reliable bits of information to gather.
Proud member of the "Don't Trade Marleau" Club
by jwizzle241 on Dec 18, 2009 4:06 AM PST up reply actions 0 recs
Well put, sir. As cogent a defense of statistical analysis as I’ve seen in a while.
Joe Martinez: You are cool.
When it's all said and done, America will be remembered for three things: The Bill of Rights, jazz, and baseball.
by cornball on Dec 18, 2009 8:49 AM PST up reply actions 0 recs
Wait….so……wha…..???
Merry Christmas, McCoven. YOU DRINK YOUR DAMN EGGNOG
by goGSW24 on Dec 17, 2009 6:24 PM PST reply actions 0 recs
something about shuttering, i suppose
by sfoakbay on Dec 17, 2009 7:17 PM PST up reply actions 0 recs
SHUTTER UP, YOU’RE HURTING HIS FEELINGS!
by quincy0191 on Dec 17, 2009 8:10 PM PST up reply actions 0 recs
Those 127 PAs gave us ALL some doubt about Garko compared to what we thought when the trade first occurred.
All? Don’t be so sure. Not me.
/dislikes this post
Adopted Giant: Henry Sosa
by raisingcain on Dec 17, 2009 6:48 PM PST reply actions 0 recs
why are you here?
you don’t even like baseball!
Petey and Fresh always dunking on Spartans,
Biscuits in baskets from Heatley and Thornton,
Bam-Bam and Gore on the way to some rings,
These are a few of my favorite things.
by beat_la_25 on Dec 17, 2009 8:17 PM PST up reply actions 0 recs
Or cod.
It's my blarg! Quick Pitch
And I tweet (more often than I blarg).
by can of corn on Dec 17, 2009 8:47 PM PST up reply actions 0 recs
/points to chart
As you can see, his dislike for the Giants far overweighs his dislike for cod. Our research has shown that the disdain toward Gordon Beckham’s hair is increasing, and God Himself could very well make his way off of xanthan’s shit list. But the latest census tells us that ‘Tim’ is becoming quite a popular baby name specifically in the San Francisco Bay area, while the number of Tims leaving this dimension is decreasing. So ‘Tims’ is rising. We’ll have a full report & new updated chart at the next quarterly meeting. Are there any questions?
"It's too late now."
by ResDog on Dec 17, 2009 9:04 PM PST up reply actions 0 recs
is the triangle angry with the circle thingy?
Petey and Fresh always dunking on Spartans,
Biscuits in baskets from Heatley and Thornton,
Bam-Bam and Gore on the way to some rings,
These are a few of my favorite things.
by beat_la_25 on Dec 17, 2009 9:15 PM PST up reply actions 0 recs
This rivals Jason Reitman’s pie chart.
Joe Martinez: You are cool.
When it's all said and done, America will be remembered for three things: The Bill of Rights, jazz, and baseball.
by cornball on Dec 18, 2009 8:49 AM PST up reply actions 0 recs
just an impartial observer, then? a computer cannot love. or eat cha-cha bowls.
Petey and Fresh always dunking on Spartans,
Biscuits in baskets from Heatley and Thornton,
Bam-Bam and Gore on the way to some rings,
These are a few of my favorite things.
by beat_la_25 on Dec 17, 2009 8:47 PM PST up reply actions 0 recs
WTF?
FTW
Matt Cain: throwing complete game shutouts since 06'. No big deal.
by cain1rstballothof on Dec 17, 2009 7:04 PM PST reply actions 0 recs
I was waiting for supporting argument, but I read all the way to the poll and saw none.
Despite the irrelevant stats and douche catcher reference, I am not satisfied.
by DesertFox on Dec 17, 2009 7:10 PM PST reply actions 0 recs
You're just bitter cuz you failed calc 2.
Petey and Fresh always dunking on Spartans,
Biscuits in baskets from Heatley and Thornton,
Bam-Bam and Gore on the way to some rings,
These are a few of my favorite things.
by beat_la_25 on Dec 17, 2009 8:07 PM PST reply actions 0 recs
Well, I hate math the same way I hate basketball: because I can’t do it and it leads me to humiliation when I try.
And this place does get pretty “point-and-laugh” quick, but I figure, that’s how life is. I don’t like it, but you got to pretend to get along.
... null, void, invalid, iniquitous, unjust, damnable, reprobate, inane, empty of meaning and effect for all time
by shanghaijim on Dec 17, 2009 9:11 PM PST reply actions 0 recs
/points and laughs
WHY IS BOCOCK?!
by Lars The Wanderer on Dec 17, 2009 9:22 PM PST up reply actions 0 recs
LAUGH THIS POINTY-HATTED SUCKA
... null, void, invalid, iniquitous, unjust, damnable, reprobate, inane, empty of meaning and effect for all time
by shanghaijim on Dec 17, 2009 9:23 PM PST up reply actions 0 recs
In general, this place gets pretty point-and-laugh when something should be pointed and laughed at. You could argue that you should never point and laugh, but fuck that noise.
Juan Carlos Perez, please start hitting.
by marcello on Dec 17, 2009 9:40 PM PST up reply actions 0 recs
I’m not so sure I’d agree with you, but I’ll pretend in order to get along. /inserts amusing reference to being gay
... null, void, invalid, iniquitous, unjust, damnable, reprobate, inane, empty of meaning and effect for all time
by shanghaijim on Dec 17, 2009 9:44 PM PST up reply actions 0 recs
/points
WITH MY PENIS!!
WHY IS BOCOCK?!
by Lars The Wanderer on Dec 17, 2009 9:47 PM PST up reply actions 0 recs
BELTRE"S TESTICLES
... null, void, invalid, iniquitous, unjust, damnable, reprobate, inane, empty of meaning and effect for all time
by shanghaijim on Dec 17, 2009 9:50 PM PST up reply actions 0 recs
Yeah, but it’s we who are doing the laughing when you point with your penis.
The baseball Satanist
by thehavenot on Dec 17, 2009 9:52 PM PST up reply actions 0 recs
i’m pretty sure natto is pointing back with his penis right now
SWORD FIGHT!!!
by sfoakbay on Dec 17, 2009 11:27 PM PST up reply actions 0 recs
Mmmm… penis fencing. My favorite method of copulation in worms.
Lethargy
It has me
by dregarx on Dec 17, 2009 11:48 PM PST up reply actions 0 recs
For the last time
We do NOT wag our genitalia to make a point Lars.
"It's too late now."
by ResDog on Dec 18, 2009 6:04 AM PST up reply actions 0 recs
I was calling for a 6-fingered spoogeball.
WHY IS BOCOCK?!
by Lars The Wanderer on Dec 18, 2009 6:13 AM PST up reply actions 0 recs
I don’t know, man. Not everybody has to be into stats to enjoy baseball.
The only thing annoys me about fanposts like this are the “STATS ARE DUMB, HERE ARE SOME STATS.” And I’ve never seen you write anything like that.
For those who just want to enjoy the game without ever busting out the spreadsheet, more power to them.
Please hit better, Randy Winn.
by oldjacket on Dec 18, 2009 5:48 AM PST up reply actions 0 recs
MY STATS PROVE THAT STATS ARE WORTHLESS
Giant Dirtbags: John Bowker, Steve Hammond. MIA List: Todd Jennings, Brian Anderson
Jeremy Affeldt induces DP's
by Giant among Angels on Dec 18, 2009 6:26 AM PST up reply actions 0 recs
lolwut
Adopted Giant: Clayton Tanner
by walkoff baltimore chop on Dec 17, 2009 11:28 PM PST reply actions 0 recs
One simple concept that’s missing from this discussion – it’s about distributions. Players tend to revert back towards their career norms (or whatever their expected level production in the current period is) once the sample size increases, and small samples are subject to huge amounts of sample error. So Garko had a really bad 127 PA’s. We expect him to do better in the future. It’s not that every player reverts back to their expected production – statistics tell us they don’t – rather we get a distribution where some players revert back to their expected level, some players surpass it, and some do worse in the future. We don’t know which Garko is going to fall under – we certainly recognize there’s a chance he never hits at the same rate he used to. The best we can do is assume he’ll be at the average of the distribution, because that will maximize the amount of time we’re right. So decision making needs to be based on that. All the factors you can come up with that may be in play for Garko causing that performance, by the way, are factors that are accounted for in the distribution.
The exception to this is if you have good reason to think there was some unusual factor affecting Garko that most of the others in the distribution are not affected by. If you think there’s a legitimate reason – not just wild speculation that maybe he doesn’t fit in at AT&T, but more something like an injury that likely affected his play in a significantly negative way, then you can factor that into play….but otherwise, the best you can do is play the odds based on the best distribution you can come up with for Garko’s future expected level of performance.
by Missing Barry on Dec 18, 2009 12:29 AM PST reply actions 0 recs
That may be the best we can do, but its sort of like saying the just invest in the S&P b/c you can’t beat it.
The best we can do is assume he’ll be at the average of the distribution, because that will maximize the amount of time we’re right.
I want my gm to beat the S&P; see who will be better than their total averages indicate (more often than not – obviously not all the time). Whether by knowing this guy was injured or this guy’s a head case, or this guy needs a fresh scene to produce etc.
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 18, 2009 7:40 AM PST up reply actions 0 recs
Well, you can’t beat the S&P because Wall St steals your profit margin.
As for beating the averages, the way you do that is by coming up with a more accurate distribution than the other guy. The principles behind the distribution still remain the same. Also, as I noted, there will be exceptions so I am open to a good baseball/scouting eye being able to combine with statistical analysis to find those exceptions, and that can help them beat the averages*. Just keep in mind exceptions are few and far between – you have to have a good reason to think a player is the exception, and I’m not seeing a plausible reason with Garko.
*Unfortunately, Sabean gives us the likes of Renteria instead. :(
by Missing Barry on Dec 18, 2009 9:12 AM PST up reply actions 0 recs
why the confusion then? (see below)
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 18, 2009 7:46 AM PST up reply actions 0 recs
Anyway – sorry for the badly written post – caused all of us much frustration (and me to slack way too much on my work – so I gots to get back to that today to remove my boss from my ass) – but I do think it’s an interesting topic.
I just shouldn’t have said “stats weren’t everything” b/c my main point was that career stats aren’t everything – trying to use some reasoning (including stats) to predict (at least one possible outcome) based ON a small sample size – which I thought was being devalued too much.
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 18, 2009 7:42 AM PST reply actions 0 recs
I do think to a degree you’re onto a discussion worthy topic. Unfortunately it seems like your delivery didn’t do a good job of creating real discussion….
by Missing Barry on Dec 18, 2009 9:13 AM PST up reply actions 0 recs
ha! clearly. my bad for sure
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 18, 2009 9:45 AM PST up reply actions 0 recs
I guess to go back to the topic – I do agree somewhat with your overall point. Small samples do tell us something (it just doesn’t tell us much), and certainly should be used in the evaluation – they just are unlikely to make much of a difference, since they don’t tell us much. I do think there are times when knowledge beyond statistical knowledge can be used to figure out which guys are more likely to perform above or below the mean – things like injuries, scouting evaluations (like if a guy, let’s just say he has the completely random name of….John Bowker…has a huge down and in hole on his swing on breaking balls that minor league pitchers can’t exploit effectively but MLB ones can, that is a good thing to know), and whatever else. I also would like a GM that can use statistical knowledge properly, AND combine that with other forms of knowledge to do even better.
That said, I don’t think Garko specifically fits this situation. I’m just not seeing anything that applies to him specifically that might make us expect his expected level of production to differ from a distribution we get from historical data. Something like park factor could affect him – like if we had an extreme pull LH HR hitter (who’s HR’s barely clear the fence, see: Damon, Johnny), it would make a lot of sense that he just doesn’t fit well in our park. But our park plays pretty normal for RHH, so that doesn’t seem to make sense. Garko wasn’t hurt. It’s supposed to be easier to hit in the NL…so it just seems to me in this case specifically we should expect him to behave the way other players do, so no adjustments are needed – he just doesn’t seem to fit the bill of an exception (or one that we can predict, at least)…
by Missing Barry on Dec 18, 2009 10:59 AM PST up reply actions 0 recs
I’m not sure if I follow but let me try to address this with my limited knowledge.
If you take all of Ryan Garko’s Career in terms of 127 PA samples sizes and plotted them on a graph (X axis for the stat value, Y access how common that value was) you would most likely end up with some form of a normal curve. Like the one below:
Now lets say that the 0 X value is his average line for 127 PA ( 279/351/441) and -1 was his value for his 127 PA with the Giants (235/307/330/). That means that we would just as to put a +1(323/395/552) in his 127 PA as he would be to put up a -1.
say hey nation is the Ralph Nader of McC.-Xanthan
by say hey nation on Dec 18, 2009 8:03 AM PST reply actions 0 recs
Was there a point to this point, other than small sample sizes are a figment of our imaginations?
Joe Martinez: You are cool.
When it's all said and done, America will be remembered for three things: The Bill of Rights, jazz, and baseball.
by cornball on Dec 18, 2009 8:32 AM PST reply actions 0 recs
Updated Post
Sorry – I did a rush job originally and wrote a bad (and combative) post in retrospect. . .
don’t have a lot of time to fix now, but I took a quick cut and edited quite a bit (see above) to try to clarify. hopefully better than leaving it alone,
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 18, 2009 8:48 AM PST reply actions 0 recs
While I appreciate your effort in modifying the original post, it kinda renders the entire thread useless. If someone reads your opening post, and then the ensuing comments, they’re going to be completely confused.
It's my blarg! Quick Pitch
And I tweet (more often than I blarg).
by can of corn on Dec 18, 2009 9:14 AM PST up reply actions 0 recs
Make a new post then?
Every reform movement has a lunatic fringe.
by Cannonballden on Dec 18, 2009 9:46 AM PST up reply actions 0 recs
No please.
I think the edit did its job.
co-dad w/AfDC of
Ishikawa, the Topps Rookie All Star Team's First baseman. Does he get a chance in 2010?
by kennv on Dec 18, 2009 10:52 AM PST up reply actions 0 recs
No
But in general in this type of forum, if you edit your post significantly, it is best to leave the original in place and insert a new section labeled UPDATE or some such so that readers who come upon the thread after many comments have been made on the original don’t think that the commenters are all nuts.
by taliesin on Dec 18, 2009 11:44 AM PST up reply actions 0 recs
We are kind of nuts, though
The baseball Satanist
by thehavenot on Dec 18, 2009 12:21 PM PST up reply actions 0 recs
Beltre's
Giant Dirtbags: John Bowker, Steve Hammond. MIA List: Todd Jennings, Brian Anderson
Jeremy Affeldt induces DP's
by Giant among Angels on Dec 18, 2009 2:27 PM PST up reply actions 0 recs
What kind?
Because I hate pistachios.
by quincy0191 on Dec 19, 2009 12:10 PM PST up reply actions 0 recs
Comments For This Post Are Closed
Secondary Sidebar
FanPosts
Community blog posts and discussion.
Recommended FanPosts
Recent FanPosts
Big League Stew - MLB - Yahoo! Sports
- Joe Nathan's elbow tells him it's time for Tommy John surgery
- Small Twitter account creates big drama in White Sox camp
- John Danks sticks with the script — no matter the consequence
- Spring Snapshot: Will fresh faces produce a White Sox winner?
- Vin Scully OK after tumble but scares bejeezus out of everybody
Certain photos copyright © 2010 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.
