Announcement

Collapse
No announcement yet.

Progressing Toward Better Stats Thread

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Originally posted by Sultan_1895-1948 View Post
    I'm willing to put in the work, just need to know what direction to go.

    I like the direction we were going with SAeff (post#26) but seems like it needs one more step.
    I'd be lying if I said I even know where to start. So many variables.
    1885 1886 1926 1931 1934 1942 1944 1946 1964 1967 1982 2006 2011

    1887 1888 1928 1930 1943 1968 1985 1987 2004 2013

    1996 2000 2001 2002 2005 2009 2012 2014 2015


    The Top 100 Pitchers In MLB History
    The Top 100 Position Players In MLB History

    Comment


    • #17
      Originally posted by Sultan_1895-1948 View Post
      Brett can you check my work on this?

      This just doesn't seem right, for a guy with

      Home - .361/.496/.652 - 106 tOPS+
      Away - .328/.467/.615 --- 94 tOPS+

      [ATTACH]138302[/ATTACH]

      Is this stat basically just saying, hey....I know he benefitted from his home park, but not as much as others throughout the league did? If that's the case, I need to question if I'm cut-out for this stat stuff. Context and perspective is important, but if we're just gonna completely throw out what THAT PLAYER ACTUALLY DID, and go 100% by league avgs....well Sammy Shortstop or Carl Catcher could be impacting the numbers. I'll hold off over-reacting until I hear about the min PA thing, and if the work is even right :silent:

      Well we know that Ted Williams was hurt in relative production by his home park. He played in a park that favored right handed hitters. That is the point of relative splits. I don't see the problem. Ted Williams road numbers WERE 192 OPS+ relative to what everyone else did on the road. If Williams WAS 192 relative to the road, and 190 overall, he has to be a little under 190 at home. This is not about being hurt or helped in stats, its about being hurt or helped in relative stats, and therefore in "value".
      Last edited by brett; 04-30-2014, 07:28 PM.

      Comment


      • #18
        Originally posted by Sultan_1895-1948 View Post
        Inbox full Brett.....

        Ok, I will do that. I only need to do the AL up until like '73 or so, right?

        Anyway, I'm doing Williams right now.

        Have a question.

        He's got a season with 5 PA and a 231 away sOPS. Will that throw off the results? Do we need a min PA? Or because it's a rate stat and not a counting stat, just roll with it? I'd say a minimum ONLY because, with such few PA, a player has less chance to regress toward the mean. What say you?
        The low PA seasons will not throw things off, because they are only getting weighted by the number of PAs.

        Comment


        • #19
          Originally posted by brett View Post
          Well we know that Ted Williams was hurt in relative production by his home park. He played in a park that favored right handed hitters. That is the point of relative splits. I don't see the problem. Ted Williams road numbers WERE 192 OPS+ relative to what everyone else did on the road. If Williams WAS 192 relative to the road, and 190 overall, he has to be a little under 190 at home. This is not about being hurt or helped in stats, its about being hurt or helped in relative stats, and therefore in "value".
          Ok thats a good explanation.

          would you say that if we were able to look at handedness, the results would more resemble his Home Road / difference?

          Comment


          • #20
            another thing I was going to say is that Williams was therefore hurt in WAR by playing in a high offensive park that favored right handed hitters (compared to usual splits). Because of his high offensive park, he gets fewer "runs" for each thing he did with the bat. It may sound odd, because I think they should have given him his actual runs produced, and converted them into fewer "wins" because it took more runs to win in Boston, but the way WAR does it, within a season, all players get their "runs" adjusted to a standard park of the season. From one season to another, the conversion of runs to wins is different depending on the run environment. High run environment might be 10.5 runs per win, low might be 9.5.

            Comment


            • #21
              Originally posted by Sultan_1895-1948 View Post
              Ok thats a good explanation.

              would you say that if we were able to look at handedness, the results would more resemble his Home Road / difference?

              Do you mean Williams versus left and right handed pitchers, or relative to left handed batters?

              I think Williams would have a much lower OPS+ relative to left handed pitchers than the league, but good for a lefty against lefties, probably similar to his overall rate.

              Compared to all left handed hitters, Williams would probably be lower than 190, but its hard to tell. The hitting edge to lefties basically raises the number of left handed hitters up to a proportion where they don't hit better than righties at the same position. There are about 10% lefties in the population but something like 30% in lineups, so the lefty edge pushes the amount of lefties up, but if we compared Williams to the top 1/3 of lefties (1/10th of the league and the population rate) he'd be lower.

              However, in sports in GENERAL, left handers are more likely to be elite athletes even when handedness is not a factor in the sport. About half of all elite athletes across all sports are left handed. Even things like sprinting.

              Comment


              • #22
                Originally posted by brett View Post
                another thing I was going to say is that Williams was therefore hurt in WAR by playing in a high offensive park that favored right handed hitters (compared to usual splits). Because of his high offensive park, he gets fewer "runs" for each thing he did with the bat. It may sound odd, because I think they should have given him his actual runs produced, and converted them into fewer "wins" because it took more runs to win in Boston, but the way WAR does it, within a season, all players get their "runs" adjusted to a standard park of the season. From one season to another, the conversion of runs to wins is different depending on the run environment. High run environment might be 10.5 runs per win, low might be 9.5.
                Does that same issue exist in WAA?

                Comment


                • #23
                  Ok, here's Mr. George Herman......all 1914 PA were at home, that's why it starts in 1915.

                  After picking a few more specific players, I'll have the full excel file with AL OPS+ and will post it.

                  RuthsOPS+.jpg

                  A couple quick questions Brett.....

                  1. I know we're removing the pitcher's batting by looking at league OPS+ but shouldn't we somehow attempt to replace that pitcher with a "DH" type batter to make a level playing field for later eras? Otherwise we have just eight hitters?

                  2. Would it be more revealing or necessary to look at sOPS/650 or /162 for shorter career guys, or since it's a rate thing, it doesn't matter?
                  Last edited by Sultan_1895-1948; 05-01-2014, 05:56 PM.

                  Comment


                  • #24
                    Originally posted by Sultan_1895-1948 View Post
                    Does that same issue exist in WAA?
                    The only difference between WAR and WAA is the baseline used. Take out "replacement runs" and you have WAA. So yes, the same issue will exist in WAA.
                    1885 1886 1926 1931 1934 1942 1944 1946 1964 1967 1982 2006 2011

                    1887 1888 1928 1930 1943 1968 1985 1987 2004 2013

                    1996 2000 2001 2002 2005 2009 2012 2014 2015


                    The Top 100 Pitchers In MLB History
                    The Top 100 Position Players In MLB History

                    Comment


                    • #25
                      Originally posted by Bothrops Atrox View Post
                      The only difference between WAR and WAA is the baseline used. Take out "replacement runs" and you have WAA. So yes, the same issue will exist in WAA.
                      So you're telling me that you can use WAR; but just change the baseline to value greatness a bit more than career value, longevity, or simply showing up with a uniform on, and it will essentially be WAA? THAT'S the only difference?
                      Last edited by Sultan_1895-1948; 05-01-2014, 09:02 PM.

                      Comment


                      • #26
                        Originally posted by Sultan_1895-1948 View Post
                        Does that same issue exist in WAA?

                        Yes, if you play in a 5 run park, versus another player in a 4 run park, you have to create or save 5 runs for every 4 that they do at home to get a win in WAA.

                        Comment


                        • #27
                          Originally posted by Sultan_1895-1948 View Post
                          So you're telling me that you can use WAR; but just change the baseline to value greatness a bit more than career value, longevity, or simply showing up with a uniform on, and it will essentially be WAA? THAT'S the only difference?
                          WAR IS WAA, plus you get a chunk of wins (2.5 per full season) for playing average. It is easy to lose WAA if you are below average for some seasons, but you have to be pretty bad to lose WAR.

                          Comment


                          • #28
                            Originally posted by Sultan_1895-1948 View Post
                            Ok, here's Mr. George Herman......all 1914 PA were at home, that's why it starts in 1915.

                            After picking a few more specific players, I'll have the full excel file with AL OPS+ and will post it.

                            [ATTACH]138325[/ATTACH]

                            A couple quick questions Brett.....

                            1. I know we're removing the pitcher's batting by looking at league OPS+ but shouldn't we somehow attempt to replace that pitcher with a "DH" type batter to make a level playing field for later eras? Otherwise we have just eight hitters?

                            2. Would it be more revealing or necessary to look at sOPS/650 or /162 for shorter career guys, or since it's a rate thing, it doesn't matter?
                            2) Doesn't matter, its a rate. OPS+ is the same no matter how many PAs you have.

                            1) Putting in a DH would have a very small effect to bring down national leaguers OPS+, by maybe a little over 1%. But its very complicated. In the NL we get pinch hitters who might be better situationally than the DH. They count, as far as I can tell (though I have asked before if a pinch hitter bats for the pitcher does it go into the league rates for OPS+, or is considered to be a pitcher at bat?) The DH might have a 110 OPS+. Pinch hitters are certainly worse, but maybe the guy who DH's in the AL is starting in the NL. There are still as many roster spots, so there is somebody with value somewhere. And WAR/WAA accounts for it by slightly lowering baselines in the AL. I would be interested in seeing AL versus NL batting stats excluding pitchers.

                            Comment


                            • #29
                              Originally posted by brett View Post
                              I am posting Dimaggio's composite sOPS+ for home and away, independent of home offensive environment (ie park factor).

                              This shows differences in Dimaggio's production at home versus all players in their home park, and with their home offensive setting put at the league average, and his road production versus all players on the road.

                              In other words, when road OPS+ is higher than home OPS+ it strongly suggests that the player SPECIFICALLY benefitted in relative production from his home park. It cancels out effects that a park has on ALL players such as Coors field just boosting offense as a whole.

                              I have also removed pitchers to make it consistent with OPS+ total. The only things it doesn't do is account for a player not having to hit in their home park on the road, and not facing their own pitching staff.



                              Code:
                              Road Pas	Road sOPS+	PAxsOPS+
                              362	        163	                59006
                              339	        201	                68139
                              334	        145	                48430
                              259	        242	                62678
                              305	        180	                54900
                              289	        221	                63869
                              346	        186	                64356
                              284	        192	                54528
                              282	        155	                43710
                              340	        189	                64260
                              173	        213	                36849
                              328	        196	                64288
                              217	        125	                27125
                              
                                              3858	                 712138
                              712138/3858=184.6 Career road sOPS+ including pitchers

                              League OPS+ with pitchers is 94 in his seasons so to remove pitchers multiply his OPS+ by .94 to get:

                              173.51 sOPS+

                              His overall was 155 so to get the home sOPS+ take 155 x 2 minus 173.5 to get 136.49

                              So Dimaggio had a 173.5 road relative OPS+ and a 136.5 home OPS+. Technically we should round them both to the 1s place, which would be 174 and 136. Quite telling of what his park did.
                              Is it .94? I get a better fit with the data using .92. The adjustment isn't a constant is it? Wouldn't it vary over time with pitchers worse today, and better hitters back then?

                              Also, since I am using the data, it's a small thing but the 1947 road PA was 307.
                              Last edited by drstrangelove; 05-02-2014, 04:41 PM.
                              "It's better to look good, than be good."

                              Comment


                              • #30
                                Originally posted by Sultan_1895-1948 View Post
                                So you're telling me that you can use WAR; but just change the baseline to value greatness a bit more than career value, longevity, or simply showing up with a uniform on, and it will essentially be WAA? THAT'S the only difference?
                                Yes. Any issues somebody has with how WAR deals with position, defense, ANYTHING will be the same issues with WAA. The ONLY difference is the comparison baseline. ONLY. . You can take the same formula and compare to any baseline you want to. Wins Above Good. Wins Above All Star. Wins Above Cobb. Whatever you want.
                                1885 1886 1926 1931 1934 1942 1944 1946 1964 1967 1982 2006 2011

                                1887 1888 1928 1930 1943 1968 1985 1987 2004 2013

                                1996 2000 2001 2002 2005 2009 2012 2014 2015


                                The Top 100 Pitchers In MLB History
                                The Top 100 Position Players In MLB History

                                Comment

                                Ad Widget

                                Collapse
                                Working...
                                X