Two Myths I'm Ready to Debunk

[rolliefingers]

Something that was mentioned during last night's game might be a factor in the MiL to ML conversion. It was stated (by Dempsey, I think) that players coming out of the Internationl League, as opposed to the PCL, will be better prepared for the ML because IL parks are more in line with ML parks, and that stats put up by IL players are a better indicator of how they will perform in the ML.

I honestly don't know if this is correct or not.

It's true that PCL parks are more hitter-friendly than IL parks. That's what people mean when they say "adjust for league".

[rolliefingers]

Well, that's just it, isn't it? If you expect the whole class to fall into the averages of historical players, that's an assumption even a monkey could make.

I'm tired of comments recently like "the Os have five pitchers at AAA who could outpitch Paul Shuey" because the translations say it is so.

The rubber hits the road when the GM has to sign one or two particular replacement level guys for the major league club and a few to stash in AAA. And which ones get the nod in what order and how long you stick with each one.

Like Frobby, I'm tired of hearing how much of a fluke it is from the statheads that Doyne, Hoey, Knott, House and others SHOULD be doing much better and how Bynum has been a fluke relative to his past numbers.

NOBODY IS SAYING THAT!

As for the last part, none of those guys, including Bynum, has been given a real chance. Repeat after me: SMALL SAMPLE SIZE.

[RShack]

Individual minor league records, in context, are a valid predictor of how someone will do in the majors.

Is this really true? I agree that available stats are the best stats available. But that does NOT mean they are valid predictors. Something can be the "best available predictor" and still be a very lousy predictor.

Here's a useful question: For guys with "decent" (not overwhelming, but decent) MiL numbers, what percentage of them last 5 years in the bigs? That's the kind of thing we need to know if we're going to use these things properly.

[hoosiers]

NOBODY IS SAYING THAT!

As for the last part, none of those guys, including Bynum, has been given a real chance. Repeat after me: SMALL SAMPLE SIZE.

How many more times would you have pushed Doyne out to the pitcher's mound before sending him back down? Hoey?

These guys were practically admitting to the press they could not get major league hitter's out.

[geschinger]

To be fair, not all of those guys are "freely available"

Here are two guys we could use right now: Jack Hannahan and Jack Cust. Both were acquired for a song.

We had that type of player in the organization but gave up on him after <20 ABs. I don't know w/any certainty that he'd of been a productive player for us but we should of at least tried to find out which of course we did not.

[RShack]

Exactly. Shack, I think you're a little too enamored with semantic arguments. And by a little, I mean a lot.

But that's exactly the mistake: this is NOT primarily a semantic argument. It is a substantive argument about how people typically use stats around here. Saying it's "just semantics" is how people improperly dismiss something that matters a lot.

What happens around here routinely is that several people implicitly treat MiL stats as if they are reliable predictors of a given individual's performance, and then say they don't do that. But they do. Then, they play what *is* a semantic game by saying, "Hey, I never said it was 100% guaranteed".

[hoosiers]

NOBODY IS SAYING THAT!

As for the last part, none of those guys, including Bynum, has been given a real chance. Repeat after me: SMALL SAMPLE SIZE.

You just wrote above how unfair it would be if I took a few individual outliers to disprove your opinion. So, what is it? What can you predict and just how accurate is it?

The only thing an ignorant poster like myself knows is that Bynum, Trax and Guts all performed significantly better than was predicted and that Doyne, Hoey, Knott and House have been worse.

You keep pleading "small sample size", but what does that mean? You, as GM of the Orioles, would have kept pushing Doyne, Hoey and Birkins out to pitch? For how long EXACTLY?

The confidence with which you assert that the performance of these players would have been significantly better with MORE TIME belies the statistical certainty of your assertions. I'm 100% certain of that.

[Sports Guy]

How many more times would you have pushed Doyne out to the pitcher's mound before sending him back down? Hoey?

These guys were practically admitting to the press they could not get major league hitter's out.

Doyne is a bad example...Older guy who wasn't that great in the minors until this year.

You aren't taking his numbers(his career numbers) and age into the proper context.

And yes, by all means, let's keep sending out Baez(funny how he is still here...Didn't you say the FO was going to sign these FA relievers to turn around and trade them at the deadline this year?)....Baez is obviously out pitching these guys.

[Sports Guy]

You just wrote above how unfair it would be if I took a few individual outliers to disprove your opinion. So, what is it? What can you predict and just how accurate is it?

The only thing an ignorant poster like myself knows is that Bynum, Trax and Guts all performed significantly better than was predicted and that Doyne, Hoey, Knott and House have been worse.

You keep pleading "small sample size", but what does that mean? You, as GM of the Orioles, would have kept pushing Doyne, Hoey and Birkins out to pitch? For how long EXACTLY?

The confidence with which you assert that the performance of these players would have been significantly better with MORE TIME belies the statistical certainty of your assertions. I'm 100% certain of that.

Do House and Knott even had 50 ab's between them?

If you think that is a fair chance(and btw, those are sporadic ab's) then you having this conversation with you is pointless.

Boyne pitched 3.2 innings....Boy, what a sample size that is! Definitely didn't deserve more than that.

[DrungoHazewood]

Here's something to ponder:

	        Age	Year	Tm	G	GS	IP	HR	BB	K	W	L	SV	ERAJim_Johnson	24	2007	NOR	26	25	134.3	25	56	75	5	10	0	6.57Garrett_Olson	23	2007	NOR	22	22	120	21	45	85	5	8	0	5.95Tim_Kester	35	2007	NOR	28	23	116.7	17	36	46	4	9	0	6.97Andy_Mitchell	28	2007	NOR	39	12	107	11	46	47	4	8	0	6.16Kurt_Birkins	26	2007	NOR	20	19	96.7	11	44	69	5	6	0	4.87Craig_Anderson	26	2007	NOR	15	15	91	12	21	39	4	6	0	5.68Rob_Bell	30	2007	NOR	10	10	61.3	9	20	42	3	4	0	5.13Jon_Leicester	28	2007	NOR	13	11	61.3	9	25	38	3	4	0	5.53Steve_Green	29	2007	NOR	52	0	60	10	41	49	2	5	2	6.54Victor_Moreno	28	2007	NOR	39	1	58.3	7	36	36	2	4	2	7.08Francis_Beltran	27	2007	NOR	47	0	52.7	4	18	33	2	4	6	5.71Cory_Doyne	25	2007	NOR	42	0	42.3	1	18	35	2	3	22	4.64Jim_Miller	25	2007	NOR	22	0	25.3	5	18	21	1	2	2	6.49Jim_Hoey	24	2007	NOR	20	0	25	2	11	28	2	1	1	2.94Nick_McCurdy	27	2007	NOR	17	1	24.3	11	8	14	1	2	0	9.51Paul_Shuey	36	2007	NOR	21	0	19.7	3	10	17	1	1	0	6.10Hayden_Penn	22	2007	NOR	4	4	18.7	3	6	14	1	1	0	5.91

Those are the Davenport translations of everyone who pitched 15 or more innings for Norfolk this year.

What looks out of place? What looks odd to you? Which of those translated performances appear to be wildly optimistic or unrealistically pessimistic if you'd give the guy 40 or 50 major league innings?

IMO, they all look pretty good. The total translated ERA for that group is 6.02. Yes, their actual MLB performance has been worse than that, but I think most of us can agree that their performance isn't typical or even expected, and is mostly in chunks of small samples.

[Sports Guy]

But that's exactly the mistake: this is NOT primarily a semantic argument. It is a substantive argument about how people typically use stats around here. Saying it's "just semantics" is how people improperly dismiss something that matters a lot.

What happens around here routinely is that several people implicitly treat MiL stats as if they are reliable predictors of a given individual's performance, and then say they don't do that. But they do. Then, they play what *is* a semantic game by saying, "Hey, I never said it was 100% guaranteed".

They are reliable but not 100%.

[BRobinsonfan]

Myth 1: there is a class of freely available "replacement level" players out there, who can easily replace the bottom few guys on a team's roster with almost no detriment to the team's performance.

Myth 2: you can look at a player's minor league stats and translate them into how they will do/would have done if they were playing in the majors.

Look at what has happened over the last few weeks and tell me again how these two myths are true.

You said you're ready to "debunk" ... when will that start?

[DrungoHazewood]

But that's exactly the mistake: this is NOT primarily a semantic argument. It is a substantive argument about how people typically use stats around here. Saying it's "just semantics" is how people improperly dismiss something that matters a lot.

What happens around here routinely is that several people implicitly treat MiL stats as if they are reliable predictors of a given individual's performance, and then say they don't do that. But they do. Then, they play what *is* a semantic game by saying, "Hey, I never said it was 100% guaranteed".

Future performance is a bell curve, a range of possibilities. Within that range they're very reliable, but we freely admit that some individuals will be out in the tails of the bell.

I don't see how this is such a hard thing to grasp, or that no matter what the standard deviation of that bell curve is (within reason) that this gives us a valuable tool.

[rolliefingers]

You just wrote above how unfair it would be if I took a few individual outliers to disprove your opinion. So, what is it? What can you predict and just how accurate is it?

The only thing an ignorant poster like myself knows is that Bynum, Trax and Guts all performed significantly better than was predicted and that Doyne, Hoey, Knott and House have been worse.

You keep pleading "small sample size", but what does that mean? You, as GM of the Orioles, would have kept pushing Doyne, Hoey and Birkins out to pitch? For how long EXACTLY?

The confidence with which you assert that the performance of these players would have been significantly better with MORE TIME belies the statistical certainty of your assertions. I'm 100% certain of that.

That means that using House's 3-for-20 stint - spread out over a month, no less - to prove that he's not an ML player is the height of madness.

Lots of guys struggle when they first come up, and lots of guys have slumps at inopportune times. You have to give them more than a handful of opportunities before drawing conclusions.

BTW, Bynum and Trax were pretty much exactly what we expected they'd be. Guthrie is the only guy who really supports your argument.

[Sports Guy]

That means that using House's 3-for-20 stint - spread out over a month, no less - to prove that he's not an ML player is the height of madness.

Lots of guys struggle when they first come up, and lots of guys have slumps at inopportune times. You have to give them more than a handful of opportunities before drawing conclusions.

BTW, Bynum and Trax were pretty much exactly what we expected they'd be. Guthrie is the only guy who really supports your argument.

And even Guthrie has been the pitcher you thought he would be for a majority of the season.