'Plus' Stats - Page 2 - SimLeague Baseball

Posts: 11631 (0)

Hall of Famer

411 649 54 1

Hall of Famer

Sorry for the formatting, I can't get it right.

In a nut shell; the bottom 4 average between 34-38 HR's a season and the top 4 are between 29-34. These #'s might be a little off, I did it from memory and I spent so much time trying to format the friggin table now I am going to be late for work!

6/8/2011 6:56 AM

mixtroy

Posts: 1057 (0)

Hall of Famer

120 3 1 2

Prospect

Posted by ncmusician_7 on 6/7/2011 11:57:00 PM (view original):
Posted by mixtroy on 6/7/2011 10:18:00 PM (view original):
Posted by ncmusician_7 on 6/7/2011 6:25:00 PM (view original):
The only things that matter are the specific advanced stats (1B/100AB#, BB/100PA#, etc). Whether or not the # numbers are above or below their real life numbers is irrelevant.
But given the choice with a few exceptions, and all else being equal, I'd gladly take the guy whose # numbers are higher than his RL numbers for the simple fact that an "over" player is more inclined to perform better than the guy whose # numbers are lower than his RL numbers.
Based on what evidence?

Are you kidding? Just look at Hauser and Freeman in the above list, and even 15 Gavvy. All of these guys' HR/100AB# are higher than their RL HR/100AB numbers and they pretty much ALWAYS will hit more HRs than they did in RL!!!

6/8/2011 10:27 AM

ncmusician_7

Posts: 1832 (0)

Hall of Famer

70 164 12 18

All Star

Posted by mixtroy on 6/8/2011 10:27:00 AM (view original):
Posted by ncmusician_7 on 6/7/2011 11:57:00 PM (view original):
Posted by mixtroy on 6/7/2011 10:18:00 PM (view original):
Posted by ncmusician_7 on 6/7/2011 6:25:00 PM (view original):
The only things that matter are the specific advanced stats (1B/100AB#, BB/100PA#, etc). Whether or not the # numbers are above or below their real life numbers is irrelevant.
But given the choice with a few exceptions, and all else being equal, I'd gladly take the guy whose # numbers are higher than his RL numbers for the simple fact that an "over" player is more inclined to perform better than the guy whose # numbers are lower than his RL numbers.
Based on what evidence?
Are you kidding? Just look at Hauser and Freeman in the above list, and even 15 Gavvy. All of these guys' HR/100AB# are higher than their RL HR/100AB numbers and they pretty much ALWAYS will hit more HRs than they did in RL!!!

Their raw real life stats aren't relevant. The # numbers are what is used by the sim. If a player has 6HR/100AB#, it doesn't matter if their real life numbers are 4HR/100AB or 8HR/100AB. Comparing sim performance to raw real life stats is pointless. Now keep one thing in mind, since the # numbers are truncated not all 6HR/100AB# are the same (could be anywhere from 6.00 to 6.99).

6/8/2011 12:47 PM

ncmusician_7

Posts: 1832 (0)

Hall of Famer

70 164 12 18

All Star

Posted by Trentonjoe on 6/8/2011 6:56:00 AM (view original):
Sorry for the formatting, I can't get it right.

In a nut shell; the bottom 4 average between 34-38 HR's a season and the top 4 are between 29-34. These #'s might be a little off, I did it from memory and I spent so much time trying to format the friggin table now I am going to be late for work!

2007 Fielder 5.16
1961 Killebrew 4.99
1964 Killebrew 5.44

1924 Hauser 5.33
1942 Laabs 5.32

Those performance history numbers (HR/100AB) support what I have said.

6/8/2011 1:18 PM (edited)

Trentonjoe

Posts: 11631 (0)

Hall of Famer

411 649 54 1

Hall of Famer

	Seasons	AVG. AB	AVG HR	AVG AB/HR
Fielder	41	601	31	19.3871
Killebrew 61	20	581	29	20.03448
Killebrew 64	25	625	34	18.38235
Ramirez	49	604	28	21.57143
		81482	4073	20.0054

Freeman	247	673	38	17.71053
Meusel	35	652	35	18.62857
Hauser	37	637	34	18.73529
Laabs	17	602	32	18.8125
		77401.88	4200.25	18.42792

For the totals I divided Freeman's totals by 8.

6/8/2011 1:56 PM (edited)

Trentonjoe

Posts: 11631 (0)

Hall of Famer

411 649 54 1

Hall of Famer

I think the bottom group are better home run hitters by about 8%.

6/8/2011 1:56 PM

Trentonjoe

Posts: 11631 (0)

Hall of Famer

411 649 54 1

Hall of Famer

I also did some #'s for the guys with Hr/100AB# of 7.

They do not support my theory.

	RL Hr/100AB	Seasons	AVG. AB	AVG HR	AVG AB/HR
Sosa 99	10.08	15	656	39	16.82051
Howard 06	9.98	48	601	34	17.67647
Ortiz o6	9.68	13	596	32	18.625
Williams 94	9.66	80	657	43	15.27907


Cravath 15	4.6	596	611	36	16.97222
York 43	5.95	34	630	38	16.57895
Keller 43	6.05	21	588	34	17.29412
Ott 36	6.18	12	598	34	17.58824

6/8/2011 2:10 PM

mixtroy

Posts: 1057 (0)

Hall of Famer

120 3 1 2

Prospect

Posted by ncmusician_7 on 6/8/2011 12:47:00 PM (view original):
Posted by mixtroy on 6/8/2011 10:27:00 AM (view original):
Posted by ncmusician_7 on 6/7/2011 11:57:00 PM (view original):
Posted by mixtroy on 6/7/2011 10:18:00 PM (view original):
Posted by ncmusician_7 on 6/7/2011 6:25:00 PM (view original):
The only things that matter are the specific advanced stats (1B/100AB#, BB/100PA#, etc). Whether or not the # numbers are above or below their real life numbers is irrelevant.
But given the choice with a few exceptions, and all else being equal, I'd gladly take the guy whose # numbers are higher than his RL numbers for the simple fact that an "over" player is more inclined to perform better than the guy whose # numbers are lower than his RL numbers.
Based on what evidence?
Are you kidding? Just look at Hauser and Freeman in the above list, and even 15 Gavvy. All of these guys' HR/100AB# are higher than their RL HR/100AB numbers and they pretty much ALWAYS will hit more HRs than they did in RL!!!
Their raw real life stats aren't relevant. The # numbers are what is used by the sim. If a player has 6HR/100AB#, it doesn't matter if their real life numbers are 4HR/100AB or 8HR/100AB. Comparing sim performance to raw real life stats is pointless. Now keep one thing in mind, since the # numbers are truncated not all 6HR/100AB# are the same (could be anywhere from 6.00 to 6.99).

Regardless of truncation, I'd take 6.00 to 6.99 any day over 4.4444233289 or whatever the truncated number may be. And for the life of me, considering your experience, I find it difficult to believe that you feel that whether or not a player's normalized numbers are above or below his RL numbers. Yes, there are some guys who have out of this world normalized numbers which, in fact, are higher than their respective RL numbers, but yet never come close to those sparkling numbers. But, for the most part, and I would think many would agree, hitters who normalize "over" their RL numbers will generally give you more than what they did in RL. For example, take a couple of guys that I always use: 1888 John Reilly and 1906 Sammy Strang. On their face, although not overwhelming, their normalized numbers are higher than their RL numbers. And guess what, you would be quite surprised at the production you get from them. i.e., WAY more than what you might think. And Reilly will hit more than 20 HRs even in minus HR parks, along with a .330 -.340 AVG, 80 SBs and a slugging pct which will approach nearly .600.

6/8/2011 3:18 PM

schwarze

Posts: 44838 (1)

Hall of Famer

1849 74 4 1 1

Hall of Famer

I have lots of statistics from the recently completed Greatest Franchise Mega Tourney ($120M theme). So I decided to look at this a little closer. I looked at all hitters that were selected at least 5 times (132 players, 1056 seasons). I calculated an average stat line for each player. I compared this average stat line to their corresponding # numbers. In a 120M league, you would expect their sim performance to be worse than their real life # performance. And this is indeed the case. I am looking at AVG, OBP and SLG. (100% means they hit their # numbers exactly.)

The sim AVG relative to their AVG# is 84.1%
The sim OBP relative to their OBP# is 84.1%
The sim SLG relative to their SLG# is 73.8% (this isn't surprising since HRs get depressed more than batting average or walks).

Now, I then compared AVG# to real life AVG and sorted by difference.

The top 20 players who's AVG# is greater than real life AVG showed a relative percentage of 85.0%... marginally higher than 84.1%
The bottom 20 players who's AVG# is less than real life AVG showed a relative percentage of 84.2%... just a hair above 84.1%

What does this all mean? The difference between the supposedly best normalized AVG hitters and the worst is minimal and could possibly be explained by ballpark choice or simply randomness.

Now,t his is just one stat (AVG). Home runs could be different. But I don't have the HR9# numbers for batters in my database.

6/8/2011 5:09 PM

footballmm11

Posts: 4984 (0)

Hall of Famer

546 5 11

Veteran

Very interesting. schwarze's research is almost exactly what I'd do.

To echo what someone else said, I don't care if they outperform their RL numbers. No **** the better normalization guys will outperform those. What really matters is whether or not two guys with the same # numbers will perform differently based on which one normalizes better. My hypothesis was that they wouldn't.

The second point that schwarze makes is that two guys with the same OPS# will perform differently based on the makeup of their profile, but that is a completely separate question. And it does seem that the more value you derive from HRs the, worse you'll perofrm in the sim since they are depressed the most.

6/8/2011 6:09 PM

mixtroy

Posts: 1057 (0)

Hall of Famer

120 3 1 2

Prospect

Posted by schwarze on 6/8/2011 5:10:00 PM (view original):
I have lots of statistics from the recently completed Greatest Franchise Mega Tourney ($120M theme). So I decided to look at this a little closer. I looked at all hitters that were selected at least 5 times (132 players, 1056 seasons). I calculated an average stat line for each player. I compared this average stat line to their corresponding # numbers. In a 120M league, you would expect their sim performance to be worse than their real life # performance. And this is indeed the case. I am looking at AVG, OBP and SLG. (100% means they hit their # numbers exactly.)

The sim AVG relative to their AVG# is 84.1%
The sim OBP relative to their OBP# is 84.1%
The sim SLG relative to their SLG# is 73.8% (this isn't surprising since HRs get depressed more than batting average or walks).

Now, I then compared AVG# to real life AVG and sorted by difference.

The top 20 players who's AVG# is greater than real life AVG showed a relative percentage of 85.0%... marginally higher than 84.1%
The bottom 20 players who's AVG# is less than real life AVG showed a relative percentage of 84.2%... just a hair above 84.1%

What does this all mean? The difference between the supposedly best normalized AVG hitters and the worst is minimal and could possibly be explained by ballpark choice or simply randomness.

Now,t his is just one stat (AVG). Home runs could be different. But I don't have the HR9# numbers for batters in my database.

It could also vary depending on whom the hitters are.

6/8/2011 8:33 PM

Trentonjoe

Posts: 11631 (0)

Hall of Famer

411 649 54 1

Hall of Famer

Performance and Value are different things.

It does look a like the #'s are pretty accurate at predicting performance. A .310 BA# hitter is more or less a .310# hitter. Or at least he'll perform like other .310# hitters.

6/8/2011 8:39 PM

boogerlips

Posts: 3731 (0)

Hall of Famer

205 45

Hall of Famer

When I do a search (lord its been a while) I use both the # and + numbers and weigh them equaly. A 100+ / .300# guy will outperform a 100+ / .292# guy or a 95+ / .300# guy. My version of awesome scientific research is looking at performance histories. Even though I pretty well know what I'm getting at this point, I always double check the P.H. to confirm my thinkingness.

6/9/2011 1:31 AM

just4me

Posts: 4771 (1)

Hall of Famer

936 1 4 2 2

Hall of Famer

I also use both the + & # stats, and have to say, I mostly use the performance history as a guide, but I generally don't draft players whose # stats are worse than their RL stats as I was also under the impression that players whose # stats are better than their RL stats tend to perform better relative to their # stats than those that have worse # stats than RL. Especially in terms of HRs. I tend to only value the # stats as an indicator of performance at $60m or if their # stats are better than RL. If their # stats are worse than RL then I assume they will not perform to that level above $70m; whereas, a player with better # stats than RL I expect to exceed the # stats at $60-70m and equal or approach them at $80-90m.

As the cap gets lower than $60m or above $90m I value the + stats more and more relative to the # stats -- as the # stats are based on an average opponent (and $60-70m caps are the closest to average you'll find on WIS and the performance histories will give a good indicator of $80-90m performance) -- since the predictive value of the # stats and the performance history become less predictive as you separate from RL average, and sim average, opponents.

6/9/2011 4:51 AM

mixtroy

Posts: 1057 (0)

Hall of Famer

120 3 1 2

Prospect

Just4me, you and I are on the same page. And I'm with you and Boog in the performance history thing, because I will never, ever draft a player without conisdering his history, irrespective of their # numbers. And even if his #s are off the charts, if he's been used multiple times, but yet, his history says he's a dud, then he is what his history says he is, and I won't touch him. However, there are times when I will draft a player who has very little history (i.e., less than five times used), but who had put up good numbers in the times he was used.

6/9/2011 8:36 AM

Forums

'Plus' Stats Topic

Search Criteria