Does speed kill?

[rluzinski]

I'm not gonna dig into it, but I'm guessing that dude, being a stat nerd, had data to back this up, or more likely was referencing data well-known in the stat nerd community.

 

Being part of the nerd community, I have access to that data!

 

1003 - 2004, Groundballs

 Pull CF Opp Right Handers 38,382 12,828 10,320 Left Handers 6,195 20,894 8,472 ---------------------- 44,577 33,722 18,792 ---------------------- 45.9% 34.6% 19.4%

Between 2003 and 2004, over twice as many groundballs were pulled than taken the opposite way. It's interesting to see that lefthanders spray grounders pretty evenly around the field, however.

[RobertR]

Thinking about it some more, I'd like to see this comparison of data.

 

Score within 3 runs.

 

#2 hitter, 1 out, #1 hitter on first base

 

vs.

 

#2 hitter, 1 out, pitcher on first base

 

You'll probably need several years worth of data to come to any conclusions, but I tend to think that the study presented is more complicated than it needs to be.

 

Robert

[bbrauch2000]

why isnt this thread in the statistical analysis forum?

 

statheads dont want to hear old schoolers opinions on it anyways, so why not place it where it belongs?

[nsimps]

Quote:

---

Because, like all just about all statistical models, you have to make assumptions and simplifications. In your world, where everything has to be exactly the same and aggregate stats can never be used, you would never be able to study anything.

---

 

What are you talking about? There are assumptions, but as your favorite bastion of knowledge Wikipedia says

---

assumptions must be made carefully or they can lead to bad conclusions. Nobody says the assumptions have to be perfect, they just have to have some sembelance of reason behind them.

[Blazer25]

Quote:

---

Between 2003 and 2004, over twice as many groundballs were pulled than taken the opposite way.

---

 

Huh? Am I missing something here?

 

RLuz if you're going to make a point with stats, at least be accurate.

 

In the numbers you showed 54% of GB were NOT pulled...

 

How then can the author of this study say "Most GB are pulled"...oh that's right, that's the side of the arguement he needs to make in order for his study to work.

 

Who's not being objective?

[rluzinski]

How then can the author of this study say "Most GB are pulled"...oh that's right, that's the side of the arguement he needs to make in order for his study to work.

 

He clearly shows what effects he's talking about and proves his assertion with his own table:

 [b]Percentage of ground balls fielded, left/right splits[/b] +--------------+-------+------+------+------+------+------+------+ | Situation | GB | Bats | 1B | 2B | 3B | SS | OF | +--------------+-------+------+------+------+------+------+------+ | All GB | 41034 | L | 0.22 | 0.31 | 0.06 | 0.15 | 0.17 | | All GB | 54460 | R | 0.04 | 0.14 | 0.26 | 0.30 | 0.17 | +--------------+-------+------+------+------+------+------+------+

53% of a lefties groundballs were fielded by either 1B or 2B. 56% of righties groundballs were fielded by the SS or 3B. It's all explained clearly. You just skimmed the article looking for "inconsistencies" and grabbed onto his "pulled" statement. I guess you should have looked a little harder.

 

Nobody says the assumptions have to be perfect, they just have to have some sembelance of reason behind them.

 

So the study's assumptions have "no semblence of reason"? Alot of pretty knowlegable people felt many of his assumptions were pretty reasonable at Baseball Think Factory. Your criteria for "reasonable assumptions" continues to be unreasonable and inpractical.

[nsimps]

Quote:

---

Alot of pretty knowlegable people felt many of his assumptions were pretty reasonable at Baseball Think Factory. Your criteria for "reasonable assumptions" continues to be unreasonable and inpractical.

---

 

If I find "a lot" of people that are "pretty knowledgable" that agree with me can i just dismiss other people's opinions too? The fact is the first part of the study rests on the assumption that you can judge the affect of a base-stealer on first base by comparing it to all other batting situations the batter following them faces.

[rluzinski]

If I find "a lot" of people that are "pretty knowledgable" that agree with me can i just dismiss other people's opinions too?

 

What I'm dismissing is your characterization of the study as having "no semblance of reason". I respect the collective knowledge of the BTF community and the magnitude of your dismissal of the author?s techniques doesn?t seem to jive with the overall response to the article there.

 

It is quite possible that the adjustments the author makes still doesn't accurately model the problem at hand. As a result, your criticisms may be perfectly valid. My "issue" is that you reject any model that makes inferences based on aggregate data. By having that stance, you reject a statistical technique that, when done correctly, is widely accepted within the statistical community.

 

Unless you are a renowned statistics professor, I find it hard to take seriously that level of dissidence.

 

The fact is the first part of the study rests on the assumption that you can judge the affect of a base-stealer on first base by comparing it to all other batting situations the batter following them faces.

 

... after making several adjustments, correct. Part 2 of his study attempted to validate that methodology by comparing the relative improvement from having a prolific base stealer at first vs. a non-prolific base stealer. It certainly didn't disagree with the conclusions from part 1. Now, it used aggregate data to do so, so I'm sure you would reject part 2's conclusions as well.