Defensive Metrics Topic

MikeT23
Posts: 58012 (0)
Hall of Famer
87 1 47 1 189
Prospect
Posted by bad_luck on 9/5/2013 10:56:00 PM (view original):
Posted by MikeT23 on 9/5/2013 10:14:00 PM (view original):
WAR, ************!!!!
100% certain that I've seen you bring up WAR in an argument.
Yep.   Because you like it.    But, as I just said, WAR doesn't tell me anything I can't see by looking at "traditional" stats. 
9/6/2013 6:50 AM
tecwrg
Posts: 18995 (1)
Hall of Famer
4 1 152 32 129
Prospect
Great moments in sports broadcasting history:



"I DON'T BELIEVE WHAT I JUST SAW (because I have an advanced stat that says it shouldn't happen)"
9/6/2013 8:56 AM
bad_luck
Posts: 21761 (0)
Hall of Famer
48 1 10
Prospect
Posted by MikeT23 on 9/6/2013 6:50:00 AM (view original):
Posted by bad_luck on 9/5/2013 10:56:00 PM (view original):
Posted by MikeT23 on 9/5/2013 10:14:00 PM (view original):
WAR, ************!!!!
100% certain that I've seen you bring up WAR in an argument.
Yep.   Because you like it.    But, as I just said, WAR doesn't tell me anything I can't see by looking at "traditional" stats. 
A) since when do you attempt to cater to others?
B) you can't use it when you think it helps your argument and then disregard it as useless and flawed when it doesn't (are you tec?)
9/6/2013 9:08 AM
MikeT23
Posts: 58012 (0)
Hall of Famer
87 1 47 1 189
Prospect
A)  If I'm going to argue baseball with you, I have to use WAR because you ALWAYS refer to it. 

B)  I've never argued against offensive WAR, as I just said "WAR doesn't tell me anything I can't see by looking at "traditional" stats", but you like to add defensive WAR and I think that's bullshit.   
9/6/2013 9:30 AM
bad_luck
Posts: 21761 (0)
Hall of Famer
48 1 10
Prospect
You realize that when you use WAR, that includes defense, right?
9/6/2013 9:34 AM
MikeT23
Posts: 58012 (0)
Hall of Famer
87 1 47 1 189
Prospect
Pretty sure it can be split between offense and defense. 
9/6/2013 9:48 AM
bad_luck
Posts: 21761 (0)
Hall of Famer
48 1 10
Prospect
It can be, but you used (several times) regular WAR, which includes defense.
9/6/2013 9:50 AM
MikeT23
Posts: 58012 (0)
Hall of Famer
87 1 47 1 189
Prospect

I assume one of the two:

1.  It was during the HOF discussion.    Which makes me lazy because I didn't bother to split them.
2.  I was just parroting the numbers you posted.  Which also makes me lazy.

All in all, I'm just like everyone else, including you, I like the stats that make me right and ignore the same stat if it doesn't help my argument.

9/6/2013 9:57 AM
bad_luck
Posts: 21761 (0)
Hall of Famer
48 1 10
Prospect
I always look at the same group of stats.
9/6/2013 9:59 AM
MikeT23
Posts: 58012 (0)
Hall of Famer
87 1 47 1 189
Prospect
Yet, when it's pointed out that one stat you've used is bullshit, you come up with "It takes roughly three seasons of data to get good numbers" or some **** like that.   Then, when it's pointed out that you're fine with 700 PA for a hitter and maybe some positions get more opportunities in the field, you waffle on what sort of sample size is required. 
9/6/2013 10:09 AM
bad_luck
Posts: 21761 (0)
Hall of Famer
48 1 10
Prospect
How is that hard to understand? You need three seasons worth of defensive opportunities to establish the defensive talent level of a player using UZR.
9/6/2013 10:13 AM
MikeT23
Posts: 58012 (0)
Hall of Famer
87 1 47 1 189
Prospect
Is it opportunities that drives your sample size?   If so, why does a SS need the same amount of seasons as a RF?
9/6/2013 10:34 AM
bad_luck
Posts: 21761 (0)
Hall of Famer
48 1 10
Prospect
Posted by MikeT23 on 9/6/2013 10:34:00 AM (view original):
Is it opportunities that drives your sample size?   If so, why does a SS need the same amount of seasons as a RF?
Of course it's opportunities, what the **** else would it be?

There is no hard number. Any stat becomes more reliable when the sample gets larger. You can draw a conclusion from a players batting average after 300 plate appearances, but his true talent level is likely to vary signigicantly. You get a better idea at 600 plate appearances, but the year to year correlation is still really low. At 900 plate appearances the true talent level is much easier to discern. But that's roughly a year and a half for batting average and you should still regress it somewhat when making a prediction for the next year.
9/6/2013 11:14 AM
MikeT23
Posts: 58012 (0)
Hall of Famer
87 1 47 1 189
Prospect

OK, humor me.

Just ballparking numbers.    A SS gets roughly 630 chances(PO/A/E) per season.   A RF gets roughly 210 per season.

Does it take "roughly three seasons" to get a good read on their defense?   So the SS needs almost 1900 chances while the RF gets the same number over three seasons that a SS gets in one.

9/6/2013 11:45 AM
bad_luck
Posts: 21761 (0)
Hall of Famer
48 1 10
Prospect
It's actually closer to 2-1. 2b, SS, and CF get about twice as many defensive opportunities as the other positions.

The more opportunities, the more reliable the stat becomes. So, one year of 3B or LF stats aren't as reliable as one year of 2B or SS stats. Just like one year of offensive stats from the 8th hitter who only plays 120 games aren't as reliable as one year of offensive stats from the leadoff hitter who plays 162.
9/6/2013 11:52 AM
◂ Prev 1...14|15|16|17|18...23 Next ▸
Defensive Metrics Topic
△Top△Bottom

Search Criteria

Terms of Use Customer Support Privacy Statement

© 1999-2024 WhatIfSports.com, Inc. All rights reserved. WhatIfSports is a trademark of WhatIfSports.com, Inc. SimLeague, SimMatchup and iSimNow are trademarks or registered trademarks of Electronic Arts, Inc. Used under license. The names of actual companies and products mentioned herein may be the trademarks of their respective owners.