Study #1 - What is clutch worth in OOTP4? - Page 2

Henry · 11-26-2003, 03:29 PM

Quote:

Originally posted by jaxmagicman
I think it was seasons. The win-loss records were seasons over .500 as wins, and season under .500 as losses and the ties were seasons at .500.

Ah... after a 2nd read, I see it

Vajracchedika · 11-26-2003, 07:04 PM

A bit of advice to those running 'tests' of OOTP. You need to be running multiple seasons and assembling much larger data sets before you can draw any conclusions. This is routine practice in any form of research. Simming 10 games is completely unreliable and inconclusive. You have to sim hundreds and hundreds of games at the very least, and preferably thousands, before you can start making proper assessments. Otherwise, you're prone to statistical anomalies in a short-term sample that can throw off your data. This is obivious from looking at the record for 'great' clutch performance vs. 'normal' clutch performance. If you simmed multiple seasons, I am sure that you would find that 'great' clutch performance would generate a much greater difference than what was found in just a 10-game sample.

sporr · 11-26-2003, 07:07 PM

Actually, he simmed 10 seasons.

Henry · 11-26-2003, 07:39 PM

Quote:

Originally posted by Vajracchedika
A bit of advice to those running 'tests' of OOTP. You need to be running multiple seasons and assembling much larger data sets before you can draw any conclusions. This is routine practice in any form of research. Simming 10 games is completely unreliable and inconclusive. You have to sim hundreds and hundreds of games at the very least, and preferably thousands, before you can start making proper assessments. Otherwise, you're prone to statistical anomalies in a short-term sample that can throw off your data. This is obivious from looking at the record for 'great' clutch performance vs. 'normal' clutch performance. If you simmed multiple seasons, I am sure that you would find that 'great' clutch performance would generate a much greater difference than what was found in just a 10-game sample.

We concluded he did run 10 seasons, so his results are valid.

clarnzz · 11-26-2003, 07:41 PM

So when is Markus going to add a "productive outs" rating? Or is that just included in "clutch"?

Skipaway · 11-27-2003, 01:04 AM

You post count whores just tried too hard to ignore that Jason simmed 10 seasons.

And productive outs is already partially in game, it's called sacrifice bunt rating.

I wonder how much of those extra wins come from clutch pitching or clutch hitting? I believe the formula behind those two must be different, and the impact might also be different.

Vajracchedika · 11-27-2003, 03:11 AM

Well, if he simmed 10 seasons, he certainly never made that clear in his posting. It's completely confusing. If he simmed 10 seasons, then why does he refer to 1-9 as "a win loss record"? Maybe he means one winning season out of 10 seasons. And why does he say that clutch hitting, if it exists, is negligible, if he's just shown that clutch performance is worth a 10-game difference between "great" and "suffers"?

I would need more clarification from him before I conclude anything from his posting. And another issue is what defines "clutch" in OOTP terms? Is it situations late in a game, with a tying or winning run at stake? Or is it used even more so in post-season play? We need to see sim results from playoffs as well. Maybe the way to do it would be to sim a season with two teams that are all great in the clutch, but edit one of them to be suffers before they meet in the World Series or something like that.

Skipaway · 11-27-2003, 03:31 AM

It's not clear? How can a team have 81.5 wins in 10 games then?

And how did you intepret this sentense,"However, notice the sub-.500 win/loss record?"

Vajracchedika · 11-27-2003, 05:25 AM

All I know is that it has to be one of the most poorly worded postings I've ever seen. I suppose that I should assume that he simmed 10 seasons, but if the guy can't even string together a few coherent sentences, then I don't know how safe it is to assume anything.

Skipaway · 11-27-2003, 05:28 AM

Quote:

Originally posted by Vajracchedika
All I know is that it has to be one of the most poorly worded postings I've ever seen. I suppose that I should assume that he simmed 10 seasons, but if the guy can't even string together a few coherent sentences, then I don't know how safe it is to assume anything.

EJ · 11-27-2003, 02:41 PM

1) The original post was *not* perfectly clear.

2) Why was this done using OOTP4, and to parrot jaxmagicman, can someone do this with v5?

3) While 10 seasons is obviously vastly better than 10 games, I would still like to see 50 seasons before making any definitive conclusions.

So as to #s 2 & 3, why am I not contributing my own time, you ask??? Barely enough time to take a leak these days. So thank you, Jason, for getting this study started!

twins15 · 11-27-2003, 03:40 PM

2) This is using OOTP4 because Jason did this sometime in 2002, before OOTP5 came out.

11-27-2003, 01:04 AM	#26
Skipaway Hall Of Famer Join Date: Apr 2003 Location: Where you live Posts: 11,017	You post count whores just tried too hard to ignore that Jason simmed 10 seasons. And productive outs is already partially in game, it's called sacrifice bunt rating. I wonder how much of those extra wins come from clutch pitching or clutch hitting? I believe the formula behind those two must be different, and the impact might also be different. __________________ Jonathan Haidt: Moral reasoning is really just a servant masquerading as a high priest.

11-27-2003, 03:31 AM	#28
Skipaway Hall Of Famer Join Date: Apr 2003 Location: Where you live Posts: 11,017	It's not clear? How can a team have 81.5 wins in 10 games then? And how did you intepret this sentense,"However, notice the sub-.500 win/loss record?" __________________ Jonathan Haidt: Moral reasoning is really just a servant masquerading as a high priest. Last edited by Skipaway; 11-27-2003 at 03:34 AM.

11-27-2003, 03:40 PM	#32
twins15 Hall Of Famer Join Date: Aug 2002 Location: Area 51 Posts: 4,792	2) This is using OOTP4 because Jason did this sometime in 2002, before OOTP5 came out. __________________ "Ah man we're just hungry man" - Dovonte Edwards Bismarck Boy Scouts of the OTBL - league yes-man Ross Gload at baseball-reference.com Book Quotes and Book Lists

11-26-2003, 07:04 PM	#22
Vajracchedika Minors (Rookie Ball) Join Date: May 2003 Posts: 36	A bit of advice to those running 'tests' of OOTP. You need to be running multiple seasons and assembling much larger data sets before you can draw any conclusions. This is routine practice in any form of research. Simming 10 games is completely unreliable and inconclusive. You have to sim hundreds and hundreds of games at the very least, and preferably thousands, before you can start making proper assessments. Otherwise, you're prone to statistical anomalies in a short-term sample that can throw off your data. This is obivious from looking at the record for 'great' clutch performance vs. 'normal' clutch performance. If you simmed multiple seasons, I am sure that you would find that 'great' clutch performance would generate a much greater difference than what was found in just a 10-game sample.

11-26-2003, 07:07 PM	#23
sporr Global Moderator Join Date: Dec 2001 Location: Muscatine, IA Posts: 8,277	Actually, he simmed 10 seasons.

11-26-2003, 07:41 PM	#25
clarnzz All Star Starter Join Date: Jun 2002 Location: Following everyone off a cliff. Posts: 1,522	So when is Markus going to add a "productive outs" rating? Or is that just included in "clutch"?

11-27-2003, 03:11 AM	#27
Vajracchedika Minors (Rookie Ball) Join Date: May 2003 Posts: 36	Well, if he simmed 10 seasons, he certainly never made that clear in his posting. It's completely confusing. If he simmed 10 seasons, then why does he refer to 1-9 as "a win loss record"? Maybe he means one winning season out of 10 seasons. And why does he say that clutch hitting, if it exists, is negligible, if he's just shown that clutch performance is worth a 10-game difference between "great" and "suffers"? I would need more clarification from him before I conclude anything from his posting. And another issue is what defines "clutch" in OOTP terms? Is it situations late in a game, with a tying or winning run at stake? Or is it used even more so in post-season play? We need to see sim results from playoffs as well. Maybe the way to do it would be to sim a season with two teams that are all great in the clutch, but edit one of them to be suffers before they meet in the World Series or something like that.

11-27-2003, 05:25 AM	#29
Vajracchedika Minors (Rookie Ball) Join Date: May 2003 Posts: 36	All I know is that it has to be one of the most poorly worded postings I've ever seen. I suppose that I should assume that he simmed 10 seasons, but if the guy can't even string together a few coherent sentences, then I don't know how safe it is to assume anything.

11-27-2003, 02:41 PM	#31
EJ Minors (Double A) Join Date: Dec 2001 Location: Pacific NW, USA Posts: 145	1) The original post was not perfectly clear. 2) Why was this done using OOTP4, and to parrot jaxmagicman, can someone do this with v5? 3) While 10 seasons is obviously vastly better than 10 games, I would still like to see 50 seasons before making any definitive conclusions. So as to #s 2 & 3, why am I not contributing my own time, you ask??? Barely enough time to take a leak these days. So thank you, Jason, for getting this study started!