Announcement

Collapse
No announcement yet.

What is the purpose of BARISP?

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

  • What is the purpose of BARISP?

    Please forgive a somewhat ignorant question, but I have a purpose for asking. What is the purpose of computing BARISP? Is it to determine the competence of a player using BARISP as the measuring criteria alone because BARISP is somehow a good measure of a player’s or teams “clutch” ability?
    The pitcher who’s afraid to throw strikes, will soon be standing in the shower with the hitter who's afraid to swing.

  • #2
    No, there is no purpose.

    Comment


    • #3
      Some thoughts. Since a little over half of PAs with RISP occur with first base empty, it is a situation where the harm from a walk (with GIDP possibilities, forceouts, and a walk not leading directly to a run) is relatively less compared to the harm from a single. In THOSE half of RISP situations at least, a single is worth something like 4-8 times as much as a walk, and a single is also worth more relative to doubles, triples and home runs.

      It would be more valuable to look at BA with runners in scoring position and first base empty, and even more telling to do so in situations with 2 outs, but that's not a lot of total situations-they are just often the ones that are remembered.

      Just checking a couple of guys, George Brett and Cal Ripken were both 10% better on a relative basis in OPS+ with RISP, ex. Brett has a 135 OPS+ overall relative to the league, but 10% better relative to the league (or 148-149) with RISP, but his IBBs are all concentrated in those plate apperarances, boosting his OPS+ in a situation where walks were the least valuable.
      Last edited by brett; 10-04-2012, 02:21 PM.

      Comment


      • #4
        But what is that actually telling you? We know a hit with RISP is valuable. Great players tend to be a little better when the opponent pitching/defense is a little worse (when they have allowed RISP)? Well, no kidding. Defensive teams are more likely to IBB a great player when his hit could hurt them most? No kidding.

        Individual player stats are meaningful when they describe a facet of that player's ability, not when they simply evince the nature of the game. Exactly what skill is being measured here?

        Comment


        • #5
          Originally posted by brett View Post
          Some thoughts. Since a little over half of PAs with RISP occur with first base empty, it is a situation where the harm from a walk (with GIDP possibilities, forceouts, and a walk not leading directly to a run) is relatively less compared to the harm from a single. In THOSE half of RISP situations at least, a single is worth something like 4-8 times as much as a walk, and a single is also worth more relative to doubles, triples and home runs.

          It would be more valuable to look at BA with runners in scoring position and first base empty, and even more telling to do so in situations with 2 outs, but that's not a lot of total situations-they are just often the ones that are remembered.

          Just checking a couple of guys, George Brett and Cal Ripken were both 10% better on a relative basis in OPS+ with RISP, ex. Brett has a 135 OPS+ overall relative to the league, but 10% better relative to the league (or 148-149) with RISP, but his IBBs are all concentrated in those plate apperarances, boosting his OPS+ in a situation where walks were the least valuable.

          Thanx Brett. You seem to have honed in on my thinking without me having to do much explaining.

          Let me give a bit more information, and hopefully you’ll stay with me. I’ve never been a big fan of anything having to do with BA, and over the years have become much more a fan of looking at how well a batter moves specifically the lead runner by whatever means, rather than getting a base hit with RISP. It’s just never made a lot of sense to me that a guy who gets a swinging bunt IF hit with a runner on 3rd but the runner doesn’t score, is given all kinds of accolades because is BARISP goes up, while a batter who gets a sac fly gets nothing because its no AB, or a guy who hits the ball to the right side and get the runner across but makes an out, actually sees his BARISP go down.

          What I’m saying is, the whole metric doesn’t make a great deal of sense to me. Well over the years I’ve spent a lot of time tracking what I call an MRU(Moved Runner Up), where whenever there’s a runner on base it becomes an “opportunity”, and if a batter’s PA is the direct cause for the lead runner moving up, it becomes an MRU. MRUs/MROs = a was to see how well a given batter does in moving the lead runners.

          Unfortunately though, while I would have liked to track them by where the base runners were and of course the number of outs, it would have added so much more time to my generating the game stats, I just did what I could and let it go. Well, last year I converted my stat program to a scoring/stat program, and have been able to tell the old circuit board what I want done, and its been able to take that next step. Please see here. pbi4.pdf

          I have to apologize for not using ML data, but that’s not my thing. What you’re seeing is HS data for the 2012 season. As you look at the various players and the various situations, I hope you’ll agree that the MRUs are a much better way to judge whether or not a player is helping his team by moving runners closer to the plate. BTW, in case you haven’t guessed, an MRU with no runners on is always a HR.

          So that’s what I’ve been tinkering with now that I can fairly easily get the data relative to the runners and the number of outs. Any thoughts are appreciated.
          The pitcher who’s afraid to throw strikes, will soon be standing in the shower with the hitter who's afraid to swing.

          Comment


          • #6
            I'm not sure where you are going with your question, but we've discussed the idea of clutch hitting from time to time. My personal view, not shared by many, is that we are looking at random events. Great hitters and poor hitters have different skill levels of hitting, and luck impacts them equally. Great hitters will have higher averages on average as compared to poor hitters, but luck will be distributed randomly among all players.

            My bias is that I don't believe in clutch hitting. If every player hits as well as he can in every situation, some will end up hitting higher in certain situations than others do due to luck.

            If someone says that a certain player hits better 'because' it is a clutch situation, then I think it's looking at the data backwards. If he truly hit's better that way, then it's because he is not trying to hit well in other situations. That doesn't make him a clutch hitter: it makes him a very good hitter who tends to be lazy when he perceives it doesn't count. Same data, different interpretation, arguably, either is correct. Arguably, no one can tell which is the correct view.
            "It's better to look good, than be good."

            Comment


            • #7
              I fall into the category that firmly believes that some players are indeed better in "clutch" situations. However, there are so many obstacles placed [from many different angles and perspectives] AGAINST the very idea of "cluch," that's it's hard to make a case for it.

              Example:

              You have a player from the Comstock Mining Giants who bats .254 overall; but who exhibits certain situation patterns and production numbers that strongly suggest he's "better" in "clutch" situations. Let's say some of the evidence looks like this:

              He's 115 hits in 453 AB for his typical .254 average. His 201 Total Bases [67 S; 27 D; 4 T; 17 HR] are also typical for him. His consistency and his occasional power exhibitions have convinced coaches NOT to tinker too much with him to correct his occasional streaks of swinging for the fences. He strikes out about 80 times a year and walks a similar number of times.

              However, a study of his performance with RISP reveals that, on average, he goes 43 for 139 when RISP prevails. That comes to .309; so his performance without RISP 72 hits in 314 is .229.

              Another thing that comes to the batting coach's attention is that he appears to adjust his hitting style whenever there is any situation with a runner at third base
              and less than two out. While he K's overall 80 times in 453 AB [17.66%], in these situations [which average 22 AB per season, he K's only twice [9.09%]. There is also a slightly higher ratio of XBT to singles in RISP situations of any kind. The coach is convinced to leave this guy alone ... he's thinking ... and he's clutch.

              The manager and the front office are unconvinced because they are sabermetrically informed:

              -sample sizes are too small;
              -random luck factors into his hit distribution patterns;
              -this club has never been higher than 4th in league standings since we signed him - so how can anything he does be seen as "cluch?"
              -if he can hit better from his "runner on 3rd fixation," have him adapt that style permanently and stop screwing around swinging for the fences.

              Meanwhile, unbeknownst to either his hitting coach or the front office, the player HIMSELF follows this mantra, in which he has complete faith and psychological conviction:

              1. In these special situations, the pitcher loses his full windup and maximum force of momentum pushing off the rubber.
              2. The guy is at a disadvantage. He's distracted. His catcher is distracted. His manager is getting edgy. This is a good place for me to be.
              3. I, as Red Barber would have put it, [am] "in the catbird seat!"

              To me, that's the foundation of clutch.

              There is a statistic that has not been formalized for all seasons; but it is on B-R in the various guises of REW 24, etc., the before and after math of situational hitting. It is based on the 24 base-out situation grid and puts precise values on each PA "outcome" from a before and after perspective.

              It strikes me that application of that measure of performance would go a long way in identifying "clutch," as long as the entire study is not strangled by mathematical prohibitions, like close and late; standings of teams in each game; event frequency barred by infrequency, etc.

              Comment


              • #8
                According to BB REF's 2012 league batting splits pages, the AL had about a 106 tOPS with RISP, 107 with runners on; the NL, 103, 106 respectively.
                Indeed the first step toward finding out is to acknowledge you do not satisfactorily know already; so that no blight can so surely arrest all intellectual growth as the blight of cocksureness.--CS Peirce

                Comment


                • #9
                  Originally posted by scorekeeper View Post
                  Thanx Brett. You seem to have honed in on my thinking without me having to do much explaining.

                  Let me give a bit more information, and hopefully you’ll stay with me. I’ve never been a big fan of anything having to do with BA, and over the years have become much more a fan of looking at how well a batter moves specifically the lead runner by whatever means, rather than getting a base hit with RISP. It’s just never made a lot of sense to me that a guy who gets a swinging bunt IF hit with a runner on 3rd but the runner doesn’t score, is given all kinds of accolades because is BARISP goes up, while a batter who gets a sac fly gets nothing because its no AB, or a guy who hits the ball to the right side and get the runner across but makes an out, actually sees his BARISP go down.

                  What I’m saying is, the whole metric doesn’t make a great deal of sense to me. Well over the years I’ve spent a lot of time tracking what I call an MRU(Moved Runner Up), where whenever there’s a runner on base it becomes an “opportunity”, and if a batter’s PA is the direct cause for the lead runner moving up, it becomes an MRU. MRUs/MROs = a was to see how well a given batter does in moving the lead runners.

                  Unfortunately though, while I would have liked to track them by where the base runners were and of course the number of outs, it would have added so much more time to my generating the game stats, I just did what I could and let it go. Well, last year I converted my stat program to a scoring/stat program, and have been able to tell the old circuit board what I want done, and its been able to take that next step. Please see here. [ATTACH]115510[/ATTACH]

                  I have to apologize for not using ML data, but that’s not my thing. What you’re seeing is HS data for the 2012 season. As you look at the various players and the various situations, I hope you’ll agree that the MRUs are a much better way to judge whether or not a player is helping his team by moving runners closer to the plate. BTW, in case you haven’t guessed, an MRU with no runners on is always a HR.

                  So that’s what I’ve been tinkering with now that I can fairly easily get the data relative to the runners and the number of outs. Any thoughts are appreciated.
                  One of the first things I look at with splits is if a guy cuts down on his strikeouts with RISP. A second thing is his 3 rates with RISP and first base empty. BA and slugging tend to drop because pitchers will walk guys if they fall behind, but if you also don't get a boost in OB% and you get a big drop in BA and slugging, it may mean something. Joe Morgan for example was horrible in BA and slugging with RISP and first base empty, (like .229/.350) AND he didn't even make up for it with an increase OB%.

                  Comment


                  • #10
                    Originally posted by brett View Post
                    One of the first things I look at with splits is if a guy cuts down on his strikeouts with RISP. A second thing is his 3 rates with RISP and first base empty. BA and slugging tend to drop because pitchers will walk guys if they fall behind, but if you also don't get a boost in OB% and you get a big drop in BA and slugging, it may mean something. Joe Morgan for example was horrible in BA and slugging with RISP and first base empty, (like .229/.350) AND he didn't even make up for it with an increase OB%.
                    I noticed that about Morgan too. I thought maybe he just lost BB as an offensive tool in that situation because he knew it wasn't as valuable, and the pitchers knew he knew.
                    Indeed the first step toward finding out is to acknowledge you do not satisfactorily know already; so that no blight can so surely arrest all intellectual growth as the blight of cocksureness.--CS Peirce

                    Comment


                    • #11
                      drstrangelove,

                      Where I’m going is, trying to figger out why all the “talking heads” go loopy over BARISP as though it were some indication of clutch hitting. Going on and on about how this player has such a great BARISP or that team is lousy in that situation doesn’t make sense to me. What makes sense is that runners are either being moved or they aren’t. if they are it’s a good thing, nd if they aren’t it’s a bad one. But base hits really don’t have a lot to do with it other than that’s the type of BIP that moved the runner.

                      In the end, whether there’s a way to interpret clutch hitting or not, I maintain using whether a BIP happened to be in the right place to just get by a fielder’s glove or just be too slow to have a play made on it is immaterial when compared to the fact that it was simply put in play.
                      The pitcher who’s afraid to throw strikes, will soon be standing in the shower with the hitter who's afraid to swing.

                      Comment


                      • #12
                        Originally posted by leewileyfan View Post
                        I fall into the category that firmly believes that some players are indeed better in "clutch" situations. However, there are so many obstacles placed [from many different angles and perspectives] AGAINST the very idea of "cluch," that's it's hard to make a case for it.
                        I guess my main problem is, there’s no universal definition for “clutch”. It appears that a RISP is what you want to use, and that’s fine, but it still makes no sense to me that a single with no outs and a runner on 3rd that doesn’t push the run across is more “clutch” than a triple with 2 outs and a runner on 1st.

                        Example:

                        You have a player from the Comstock Mining Giants who bats .254 overall; but who exhibits certain situation patterns and production numbers that strongly suggest he's "better" in "clutch" situations. Let's say some of the evidence looks like this:

                        He's 115 hits in 453 AB for his typical .254 average. His 201 Total Bases [67 S; 27 D; 4 T; 17 HR] are also typical for him. His consistency and his occasional power exhibitions have convinced coaches NOT to tinker too much with him to correct his occasional streaks of swinging for the fences. He strikes out about 80 times a year and walks a similar number of times.

                        However, a study of his performance with RISP reveals that, on average, he goes 43 for 139 when RISP prevails. That comes to .309; so his performance without RISP 72 hits in 314 is .229.

                        Another thing that comes to the batting coach's attention is that he appears to adjust his hitting style whenever there is any situation with a runner at third base and less than two out. While he K's overall 80 times in 453 AB [17.66%], in these situations [which average 22 AB per season, he K's only twice [9.09%]. There is also a slightly higher ratio of XBT to singles in RISP situations of any kind. The coach is convinced to leave this guy alone ... he's thinking ... and he's clutch.

                        The manager and the front office are unconvinced because they are sabermetrically informed:

                        -sample sizes are too small;
                        -random luck factors into his hit distribution patterns;
                        -this club has never been higher than 4th in league standings since we signed him - so how can anything he does be seen as "cluch?"
                        -if he can hit better from his "runner on 3rd fixation," have him adapt that style permanently and stop screwing around swinging for the fences.
                        The problem isn’t that the front office is sabermetrically informed, its that they aren’t sabermetrically enough.

                        Meanwhile, unbeknownst to either his hitting coach or the front office, the player HIMSELF follows this mantra, in which he has complete faith and psychological conviction:

                        1. In these special situations, the pitcher loses his full windup and maximum force of momentum pushing off the rubber.
                        2. The guy is at a disadvantage. He's distracted. His catcher is distracted. His manager is getting edgy. This is a good place for me to be.
                        3. I, as Red Barber would have put it, [am] "in the catbird seat!"

                        To me, that's the foundation of clutch.
                        I agree, but it makes no difference where the runner(s) may be. Why only on 2nd or 3rd?

                        There is a statistic that has not been formalized for all seasons; but it is on B-R in the various guises of REW 24, etc., the before and after math of situational hitting. It is based on the 24 base-out situation grid and puts precise values on each PA "outcome" from a before and after perspective.

                        It strikes me that application of that measure of performance would go a long way in identifying "clutch," as long as the entire study is not strangled by mathematical prohibitions, like close and late; standings of teams in each game; event frequency barred by infrequency, etc.
                        WHEW! Most of that shot right over my head, but one thing didn’t. It sounds as though the data is ML data which wouldn’t suit my purposes at all, since all I’m interested in is HS data. However, I would be very interested in seen the 24 base-out situation grid and the precise values on each PA "outcome" from a before and after perspective. I may be able to use it to establish a “base” to use to compare to.
                        The pitcher who’s afraid to throw strikes, will soon be standing in the shower with the hitter who's afraid to swing.

                        Comment


                        • #13
                          Originally posted by leewileyfan View Post
                          I fall into the category that firmly believes that some players are indeed better in "clutch" situations. However, there are so many obstacles placed [from many different angles and perspectives] AGAINST the very idea of "cluch," that's it's hard to make a case for it.
                          I guess my main problem is, there’s no universal definition for “clutch”. It appears that a RISP is what you want to use, and that’s fine, but it still makes no sense to me that a single with no outs and a runner on 3rd that doesn’t push the run across is more “clutch” than a triple with 2 outs and a runner on 1st.

                          Example:

                          You have a player from the Comstock Mining Giants who bats .254 overall; but who exhibits certain situation patterns and production numbers that strongly suggest he's "better" in "clutch" situations. Let's say some of the evidence looks like this:

                          He's 115 hits in 453 AB for his typical .254 average. His 201 Total Bases [67 S; 27 D; 4 T; 17 HR] are also typical for him. His consistency and his occasional power exhibitions have convinced coaches NOT to tinker too much with him to correct his occasional streaks of swinging for the fences. He strikes out about 80 times a year and walks a similar number of times.

                          However, a study of his performance with RISP reveals that, on average, he goes 43 for 139 when RISP prevails. That comes to .309; so his performance without RISP 72 hits in 314 is .229.

                          Another thing that comes to the batting coach's attention is that he appears to adjust his hitting style whenever there is any situation with a runner at third base and less than two out. While he K's overall 80 times in 453 AB [17.66%], in these situations [which average 22 AB per season, he K's only twice [9.09%]. There is also a slightly higher ratio of XBT to singles in RISP situations of any kind. The coach is convinced to leave this guy alone ... he's thinking ... and he's clutch.

                          The manager and the front office are unconvinced because they are sabermetrically informed:

                          -sample sizes are too small;
                          -random luck factors into his hit distribution patterns;
                          -this club has never been higher than 4th in league standings since we signed him - so how can anything he does be seen as "cluch?"
                          -if he can hit better from his "runner on 3rd fixation," have him adapt that style permanently and stop screwing around swinging for the fences.
                          The problem isn’t that the front office is sabermetrically informed, its that they aren’t sabermetrically enough.

                          Meanwhile, unbeknownst to either his hitting coach or the front office, the player HIMSELF follows this mantra, in which he has complete faith and psychological conviction:

                          1. In these special situations, the pitcher loses his full windup and maximum force of momentum pushing off the rubber.
                          2. The guy is at a disadvantage. He's distracted. His catcher is distracted. His manager is getting edgy. This is a good place for me to be.
                          3. I, as Red Barber would have put it, [am] "in the catbird seat!"

                          To me, that's the foundation of clutch.
                          I agree, but it makes no difference where the runner(s) may be. Why only on 2nd or 3rd?

                          There is a statistic that has not been formalized for all seasons; but it is on B-R in the various guises of REW 24, etc., the before and after math of situational hitting. It is based on the 24 base-out situation grid and puts precise values on each PA "outcome" from a before and after perspective.

                          It strikes me that application of that measure of performance would go a long way in identifying "clutch," as long as the entire study is not strangled by mathematical prohibitions, like close and late; standings of teams in each game; event frequency barred by infrequency, etc.
                          WHEW! Most of that shot right over my head, but one thing didn’t. It sounds as though the data is ML data which wouldn’t suit my purposes at all, since all I’m interested in is HS data. However, I would be very interested in seen the 24 base-out situation grid and the precise values on each PA "outcome" from a before and after perspective. I may be able to use it to establish a “base” to use to compare to.
                          The pitcher who’s afraid to throw strikes, will soon be standing in the shower with the hitter who's afraid to swing.

                          Comment


                          • #14
                            Why not look at the percentage of RISP driven in and even exclude walks if desired.

                            Lets take a look at Joe Morgan and George Brett who basically were equal saber value rate hitters.

                            Morgan was about 23.83 batting runs above average per 600 plate appearances and a 132 OPS+
                            Brett was about actually 22.04 batting runs above average per 600 PAs with a 135 OPS+


                            Brett saw a total of 3800 runners in scoring position in his career
                            and 2980 of those in "at bats"
                            and drove in 1207 in those situations (some RBI included runners on first and home runs)
                            That would mean that he drove in 1207 given 2980 total RISP excluding non-at bat PAs, or .405 per AB (and .318 per PA)

                            Morgan saw a total of 3108 runners in scoring position in his career
                            and 2376 of those in "at bats"
                            and drove in just 828 in those situations.
                            That would put him at just .348 of an RBI per at bat and .266 per PA

                            If Morgan had driven in at Brett's rate per at bat, he would have driven in 135 more runs in his career.

                            RBI per RISP seen (in RISP at bats) seems to perhaps be a decent looking stat for me.

                            Joe Dimaggio comes in a .494/.417 per at bat RISP and PA RISP for the splits we have
                            Ted Williams at .514/.357
                            Last edited by brett; 10-05-2012, 08:28 AM.

                            Comment


                            • #15
                              Originally posted by scorekeeper View Post
                              I guess my main problem is, there’s no universal definition for “clutch”. It appears that a RISP is what you want to use, and that’s fine, but it still makes no sense to me that a single with no outs and a runner on 3rd that doesn’t push the run across is more “clutch” than a triple with 2 outs and a runner on 1st. [
                              I used several examples, NONE of which suggested that a runner on third with no outs and followed by a single OR a triple with two outs.

                              :I agree, but it makes no difference where the runner(s) may be. Why only on 2nd or 3rd?
                              Again. This is YOUR scenario, not mine. I cited SITUATIONS often considered, RISP [specifying no particular placement] and the SPECIAL situation of a runner on 3B and less than 2 outs.

                              That last one, for me, speaks volumes about batter psychology if certain patterns are revealed. I don't have the spreadsheet I prepared several years ago [may have sent it to the RECYCLE BIN]; but I believe I recall at least one player who made WORSE contact in that sitiuation than in others.

                              I believe it was Alfonso Soriano who fanned MORE with 3B and , 2 Outs [almost 30%] than in other situations. As I recall, Kendall and Pujols exhibited very large declines in K's below their individual norms.

                              The biggest single factor for CLUTCH, to me, is the psychological mindset of the batter who sees the pitcher as the one with all the pressure on him [rather than the batter himself]. If one enters a "clutch" study simple acknowledging that point, I am sure he will find statistical evidence of that "attitude" in the batting lines.

                              If one accepts certain mathematical prohibitions on sample size, regressed win expectation "pressure" situations, "late and close" pronouncements, the mere iteration of subtractions will leave him high and dry ... concluding that there is no such thing as clutch.

                              That conclusion may "scientifically" pronounce that it's all random chance and luck. To me, that's a cop out.



                              WHEW! Most of that shot right over my head, but one thing didn’t. It sounds as though the data is ML data which wouldn’t suit my purposes at all, since all I’m interested in is HS data. However, I would be very interested in seen the 24 base-out situation grid and the precise values on each PA "outcome" from a before and after perspective. I may be able to use it to establish a “base” to use to compare to.[/QUOTE]

                              Comment

                              Ad Widget

                              Collapse
                              Working...
                              X