Comments on: Can Bronson Catch Bert? http://www.baseball-reference.com/blog/archives/15160 This and that about baseball stats. Tue, 16 Jul 2013 17:01:55 +0000 hourly 1 https://wordpress.org/?v=4.6 By: Brian K http://www.baseball-reference.com/blog/archives/15160/comment-page-1#comment-155722 Wed, 21 Sep 2011 23:19:21 +0000 http://www.baseball-reference.com/blog/?p=15160#comment-155722 So much for that. Complete game 2:12 minute shutout for Bronson today. The Astros really must want to get the season over with.

]]>
By: Mike Felber http://www.baseball-reference.com/blog/archives/15160/comment-page-1#comment-154734 Tue, 20 Sep 2011 07:03:52 +0000 http://www.baseball-reference.com/blog/?p=15160#comment-154734 Fair enough JT, & thanks for the advanced stats info. But we can do much better by asking the questions in specific cases & figuring out the causes of the biggest distinctions: & a good chance we can come to a general consensus as to how to take each system with a grain of salt, when, & in which directions.

Is it really almost all defense that accounts for differences in players as disparate as Killer, Simmons, Brock, & Santo, as mentioned in post 14? I seem to recall walks being cited as a reason Stargell was rated much higher on FG, something that was not defense. Park, base running, other things?

IF it is most all defense measured differently, I would state Offensive WAR, either system, & separately the career defensive #s of both systems, so we would know & could take into consideration the arguments for each system.

]]>
By: Johnny Twisto http://www.baseball-reference.com/blog/archives/15160/comment-page-1#comment-154664 Tue, 20 Sep 2011 04:04:08 +0000 http://www.baseball-reference.com/blog/?p=15160#comment-154664 there seems to be a consensus that WAR, regardless of who calculates it, while useful, isn't the last word

It's definitely not the last word. Its developer would say that. It's an argument-starter, not an argument-ender. It's a wonderful tool for broad-based comparisons. You want to find the best players who have been left out of the HOF? Look up players with more than, say, 40 WAR. It's better than searching for players with 2000+ hits, or a 120 OPS+, etc. But it is not not not intended to answer all questions. Unfortunately people take it that way and disparage it for that very reason.

]]>
By: Johnny Twisto http://www.baseball-reference.com/blog/archives/15160/comment-page-1#comment-154658 Tue, 20 Sep 2011 03:56:22 +0000 http://www.baseball-reference.com/blog/?p=15160#comment-154658 But in the seven advanced stats at the end of this B-R page, Maddux does better overall! Do not 6 of these, all but ERA +, ADJUST for defense? http://www.baseball-reference.com/leagues/NL/1993-pitching-leaders.shtml. How is this internally consistent?

No, they do not.

]]>
By: Jeff Rose http://www.baseball-reference.com/blog/archives/15160/comment-page-1#comment-154439 Mon, 19 Sep 2011 20:44:26 +0000 http://www.baseball-reference.com/blog/?p=15160#comment-154439 Yet another backdoor way to gloss Blyleven!

]]>
By: Santos http://www.baseball-reference.com/blog/archives/15160/comment-page-1#comment-154426 Mon, 19 Sep 2011 20:17:06 +0000 http://www.baseball-reference.com/blog/?p=15160#comment-154426 @25

I think that's a sensible route. I think the major difference between the WAR here and Fangraphs WAR (for position players) is the defensive metric used. This site uses Total Zone while Fangraphs uses UZR (for seasons from 2002 to the present). That is probably the major reason for the variation, and as you said, can account for massive differences over a career. I'm curious to see what you find out when you look into it though. If you publish your findings anywhere be sure to make it known if you can.

]]>
By: Santos http://www.baseball-reference.com/blog/archives/15160/comment-page-1#comment-154424 Mon, 19 Sep 2011 20:11:59 +0000 http://www.baseball-reference.com/blog/?p=15160#comment-154424 @24

Point well taken. Who is responsible for categorizing the batted ball types? Is it a third party company that looks at all this, or is it an official scorer type capacity at each ballpark? Just curious to know. I agree FIP is a happy medium, and I think it's probably the most well known of the DIPS stats available today, and the differences between the two aren't all that big to begin with. I guess what I was trying to get across in reference to the original question, is that right now there is no concrete right or wrong when it comes to WAR, so there won't be one single version of it (most notably for pitching). But it's not necessary for them to agree since the underlying stats are based off two different philosophies. But as long as one understand the ingredients that go into making them, then it doesn't matter.

]]>
By: Mike Felber http://www.baseball-reference.com/blog/archives/15160/comment-page-1#comment-154419 Mon, 19 Sep 2011 20:05:10 +0000 http://www.baseball-reference.com/blog/?p=15160#comment-154419 Useful information above, thanks. Though the trouble with a smell test is that it so So prone to selection bias, what you think before, & the usually small sample size you see, received wisdom, & other biases.

Also, I appreciate that Fangraphs is often higher due to replacement level differences, but very often the total scores are about the same also! And often enough they are 10, 20, even 30% or so variable from B-R WAR in career value! And I am just talking about position players here, not even pitchers, who it is pointed out tend to vary more.

So while it might be useful to advocate for one pitching evaluation system over another, at least it would be very useful to look at the causes for all the variation on how most guys, position players, are rated over a career.

Best would be to take folks like those I show above who have a high % of variation, see WHY that is so, & compare it to the many players who rate about the same in both. That will isolate what factors we think should be adjusted in which systems.

But it is different enough with players that it only makes sense to brainstorm & hear many opinions & arguments what each system does best. THEn we will be in a better position to independently evaluate how much to accept each version, & in which instances.

]]>
By: Dr. Doom http://www.baseball-reference.com/blog/archives/15160/comment-page-1#comment-154417 Mon, 19 Sep 2011 20:02:27 +0000 http://www.baseball-reference.com/blog/?p=15160#comment-154417 @21

The problem with SIERA, of course, is that you have to believe/agree with the scoring of GB/LD/FB. It has been noted that those are notoriously inaccurate from park to park and scorer to scorer when compared to actual ball-speed and angle data. Personally, I think FIP does a good job of finding a happy medium between a total projection system like SIERA and a traditional system like ERA. B-R's version of WAR is also a "happy medium" type of thing, but it puts a lot more weight on sequencing. Really, it's probably best to just average the two and see what you get. But that's just my theory.

Additionally, as to the whole WAR discussion, Baseball Prospectus has WARP and The Baseball Gauge also uses a WAR metric that's similar to Fangraphs for pitching, but more similar to B-R for offense. Both of these sites are also worth checking. And, of course, if we're talking MVP or Cy Young voting, one should never go into it without at least considering WPA and REW, and then considering WPA/LI. It's a lot to think about, that's for sure.

]]>
By: Kahuna Tuna http://www.baseball-reference.com/blog/archives/15160/comment-page-1#comment-154377 Mon, 19 Sep 2011 18:21:49 +0000 http://www.baseball-reference.com/blog/?p=15160#comment-154377 "P.S. Note that the 50 homers I allowed accounted for 80 (59.7%) of the 134 total runs scored off me in 1986, the highest percentage of any of the 45 pitcher seasons shown above. Basically, when I wanted to give up a run or two, I grooved a homer pitch — much more efficient and less stressful than holding a bunch of runners on. Jack and Bill, I couldn't help noticing that your percentages were well below 50%. As veterans, don't you think that's kind of irresponsible?"

]]>