Stats-Based CPU Sliders [MLB 12 Version]

Collapse

Recommended Videos

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • steviegolfballs
    Rookie
    • Feb 2010
    • 243

    #316
    Re: Stats-Based CPU Sliders [MLB 12 Version]

    Nomo,

    Thanks for your feedback! As you know I try and combine your in-depth statistics with the "eye test". I just felt that at FR - 4 it did not have the flow that the game should have and many plays looked odd (balls rolling past infielders as they stand still watching them). Putting SH back to 4 has reduced the number of infield singles but also put back in the number of outfield singles so that was my reasoning for going back to FR - 5 and I think it might give a better look to the game and a bit more realism than dealing with the infield singles. Of course when I raised FR back to 5 the outfielders were cutting too many balls off in the gaps and down the lines so that was my reasoning for dropping the fielder run speed one notch down to 3. FRS doesn't seem to affect the infielders nearly as much as the outfielders. I watched two games this morning and it produced a good result although one of the games was in Detroit and there were two triples to right-center but that is more a product of the field than the sliders, that park is huge in that section of the outfield!

    One other thing I noticed in doing my testing from before with SH 4 vs SH 3....I had many more strikeouts with it set to SH 3, in the 40 game test at SH 3 I was up around 7.3-7.4 while I was closer to 6.9-7.0 with SH 4. Is this consistent with your findings?

    When I look at the replay vault I don't see too many inconsistencies with how they label the hits but that doesn't necessarily mean they are charting them in the pitcher analysis the same as they are labeling them for replay.

    Comment

    • nomo17k
      Permanently Banned
      • Feb 2011
      • 5735

      #317
      Re: Stats-Based CPU Sliders [MLB 12 Version]

      Originally posted by steviegolfballs
      Nomo,

      Thanks for your feedback! As you know I try and combine your in-depth statistics with the "eye test". I just felt that at FR - 4 it did not have the flow that the game should have and many plays looked odd (balls rolling past infielders as they stand still watching them). Putting SH back to 4 has reduced the number of infield singles but also put back in the number of outfield singles so that was my reasoning for going back to FR - 5 and I think it might give a better look to the game and a bit more realism than dealing with the infield singles. Of course when I raised FR back to 5 the outfielders were cutting too many balls off in the gaps and down the lines so that was my reasoning for dropping the fielder run speed one notch down to 3. FRS doesn't seem to affect the infielders nearly as much as the outfielders. I watched two games this morning and it produced a good result although one of the games was in Detroit and there were two triples to right-center but that is more a product of the field than the sliders, that park is huge in that section of the outfield!

      One other thing I noticed in doing my testing from before with SH 4 vs SH 3....I had many more strikeouts with it set to SH 3, in the 40 game test at SH 3 I was up around 7.3-7.4 while I was closer to 6.9-7.0 with SH 4. Is this consistent with your findings?

      When I look at the replay vault I don't see too many inconsistencies with how they label the hits but that doesn't necessarily mean they are charting them in the pitcher analysis the same as they are labeling them for replay.
      Yes your reasoning makes sense... I haven't felt weird about FR = 4 so I keep it there for now, but I'll keep an eye on it.... Yes I do notice that infielder reactions are slightly worse, but the hits going through in between infielders haven't bothered me much (partly because it didn't happen quite as often in previous years' games I suppose...).

      Thank you for visual feedback... they are much appreciated!

      Strikeouts are fine... it's right on consistently that I feel really lucky to have found that combination. SH shouldn't affect that aspect of the game anyways. Those numbers related to walks, strikeouts, etc., have been pretty solid, so I don't think any need for adjustment. Right now I'm only touching sliders that affect things once ball is hit for that reason... mostly.

      Also thanks for confirming the replay vault thing... I have more confidence in those numbers then!
      The Show CPU vs. CPU game stats: 2018,17,16,15,14,13,12,11

      Comment

      • steviegolfballs
        Rookie
        • Feb 2010
        • 243

        #318
        Re: Stats-Based CPU Sliders [MLB 12 Version]

        Originally posted by nomo17k
        Yes your reasoning makes sense... I haven't felt weird about FR = 4 so I keep it there for now, but I'll keep an eye on it.... Yes I do notice that infielder reactions are slightly worse, but the hits going through in between infielders haven't bothered me much (partly because it didn't happen quite as often in previous years' games I suppose...).

        Thank you for visual feedback... they are much appreciated!

        Strikeouts are fine... it's right on consistently that I feel really lucky to have found that combination. SH shouldn't affect that aspect of the game anyways. Those numbers related to walks, strikeouts, etc., have been pretty solid, so I don't think any need for adjustment. Right now I'm only touching sliders that affect things once ball is hit for that reason... mostly.

        Also thanks for confirming the replay vault thing... I have more confidence in those numbers then!
        Agreed on the SH feedback, I was hoping you would say that! I am really happy with where the K's & BB's are and was hoping the SH was not the cause for the slight rise in my totals. I will get the games played and give the results in a few days.

        Comment

        • nomo17k
          Permanently Banned
          • Feb 2011
          • 5735

          #319
          Re: Stats-Based CPU Sliders [MLB 12 Version]

          Oh well my work computer is down so I'll just do a quick update...

          After 36 games, with the test set below.

          GIDPs stayed up, line drives fraction is noticeably down. Fly ball / line drive / ground ball fractions are:

          38.6% / 14.4% / 47.0%

          with this set. With default sliders they were:

          37.7% / 17.6% / 44.7%

          which pretty consistently stayed around through all the testing.

          So a few line drives ended up becoming either fly or ground balls, but tipped more toward grounders. That may have ended up increasing GIDPs quite a bit.


          After this I'll go back to Solid Hit 4 and Arm Strength 2.


          ### Apr. 1 test set ###

          Contact: 4
          Power: 5
          Timing: 6
          Foul Frequency: 4
          Solid Hits: 3
          Starter Stamina: 10
          Reliever Stamina: 5
          Pitcher Control: 5
          Pitcher Consistency: 4
          Strike Frequency: 5
          Manager Hook: 5
          Pickoffs: 5
          Pitch Speed: 10
          Fielding Errors: 7
          Throwing Errors: 5
          Fielder Run Speed: 4
          Fielder Reaction: 4
          Fielder Arm Strength: 3
          BR Speed: 6
          BR Steal Ability: 4
          BR Steal Frequency: 8
          Wind: 4
          Injury Frequency: 5
          Attached Files
          The Show CPU vs. CPU game stats: 2018,17,16,15,14,13,12,11

          Comment

          • nomo17k
            Permanently Banned
            • Feb 2011
            • 5735

            #320
            Re: Stats-Based CPU Sliders [MLB 12 Version]

            Here's Apr. 4 test set I'm using now.

            Contact: 4
            Power: 5
            Timing: 6
            Foul Frequency: 4
            Solid Hits: 4
            Starter Stamina: 10
            Reliever Stamina: 5
            Pitcher Control: 5
            Pitcher Consistency: 4
            Strike Frequency: 5
            Manager Hook: 5
            Pickoffs: 5
            Pitch Speed: 10
            Fielding Errors: 7
            Throwing Errors: 5
            Fielder Run Speed: 4
            Fielder Reaction: 4
            Fielder Arm Strength: 2
            BR Speed: 6
            BR Steal Ability: 4
            BR Steal Frequency: 8
            Wind: 4
            Injury Frequency: 5
            The Show CPU vs. CPU game stats: 2018,17,16,15,14,13,12,11

            Comment

            • steviegolfballs
              Rookie
              • Feb 2010
              • 243

              #321
              Re: Stats-Based CPU Sliders [MLB 12 Version]

              Originally posted by nomo17k
              Here's Apr. 4 test set I'm using now.

              Contact: 4
              Power: 5
              Timing: 6
              Foul Frequency: 4
              Solid Hits: 4
              Starter Stamina: 10
              Reliever Stamina: 5
              Pitcher Control: 5
              Pitcher Consistency: 4
              Strike Frequency: 5
              Manager Hook: 5
              Pickoffs: 5
              Pitch Speed: 10
              Fielding Errors: 7
              Throwing Errors: 5
              Fielder Run Speed: 4
              Fielder Reaction: 4
              Fielder Arm Strength: 2
              BR Speed: 6
              BR Steal Ability: 4
              BR Steal Frequency: 8
              Wind: 4
              Injury Frequency: 5
              What is your reasoning for starter stamina 10?

              Comment

              • nomo17k
                Permanently Banned
                • Feb 2011
                • 5735

                #322
                Re: Stats-Based CPU Sliders [MLB 12 Version]

                With the OSFM Spring Training roster out last night, many of you guys presumably started your franchise last night. You must have spent an unhealthy amount of hours occupying PS3 since then, making your wife/GF very, very unhappy.

                Tonight, instead of cuddling your DS3 controller, you need to spend time putting effort into pleasing your lady to save the relationship from totally falling apart.

                I had to do my share by sliding.... BR Speed slider.

                I just wanted to see base runners run at a realistic speed on base path. To do so, I made a bunch of LHBs with Speed ratings at 0, 50, and 99, make them bunt for hit, and measured how fast it took for them between the bat contact and reaching the first base. Each iteration, I changed BR Speed slider to 0, 5, and 10 to see how much the slider changes their speeds.

                The MLB average is about 4.3 seconds. But that's for actually hitting, and not (regular) bunting that I used, so presumably my batter runners perhaps got better jumps.

                What do I take from this?

                I think the default (5) is fine. The median Speed rating for SCEA roster is about 40, and if you look at the lines that crosses Speed rating of 40, BR Speed slider at 5 cuts the MLB average line, perhaps slightly faster.

                Given my guys perhaps got a better jump (and therefore yielded faster homt-to-1B times), maybe the slider value at 5 - 6 would replicate most realistic physical speed.

                I'm going with 6 in hope of reducing GIDPs while maintaining visual authenticity.
                Attached Files
                The Show CPU vs. CPU game stats: 2018,17,16,15,14,13,12,11

                Comment

                • nomo17k
                  Permanently Banned
                  • Feb 2011
                  • 5735

                  #323
                  Re: Stats-Based CPU Sliders [MLB 12 Version]

                  Originally posted by steviegolfballs
                  What is your reasoning for starter stamina 10?
                  It's just a personal testing to see if we can do anything about the "6th inning scoring hike."

                  The stats indicate starters pitch a bit longer but not much, meaning situations and Manager Hook probably are much more important factors in when starters get pulled... I'm just getting more games in to compare with inning-by-inning scores that I calculated before.

                  But overall, the stats haven't changed much by that choice. So in the end I'll just adjust Starter Stamina by other means... (likely to 7, since an earlier discussion indicates 7 gives most realistic innings pitched over a season when you sim all the games...)
                  The Show CPU vs. CPU game stats: 2018,17,16,15,14,13,12,11

                  Comment

                  • ralphieboy11
                    Pro
                    • Jul 2005
                    • 543

                    #324
                    Re: Stats-Based CPU Sliders [MLB 12 Version]

                    Pretty good stuff on the chart.

                    I think you are really getting close to optimal stats combined with visual aspect.

                    I'm hoping this latest move with solid hits and arm strength brings down those GIDPs.

                    Are we worried lowering fielder arm strength will affect stolen bases or is that more tied to stolen base ability? I haven't really felt comfortable with the stolen bases yet. It's tough to get slower runners to get a few without letting the faster players run too much.

                    Comment

                    • nomo17k
                      Permanently Banned
                      • Feb 2011
                      • 5735

                      #325
                      Re: Stats-Based CPU Sliders [MLB 12 Version]

                      Originally posted by ralphieboy11
                      Pretty good stuff on the chart.

                      I think you are really getting close to optimal stats combined with visual aspect.

                      I'm hoping this latest move with solid hits and arm strength brings down those GIDPs.

                      Are we worried lowering fielder arm strength will affect stolen bases or is that more tied to stolen base ability? I haven't really felt comfortable with the stolen bases yet. It's tough to get slower runners to get a few without letting the faster players run too much.

                      Yes aside from GIDPs I'm pretty much trying to adjust visual feel right now. I'm curious how FR = 5, FRS = 3 turns out. My bet is that won't help GIDPs, but I just went with FR = 4 by feel so don't really know what's going to happen.

                      Fielder Arm Strength doesn't affect catcher on steals, so yes you'd be adjusting BR Speed and BR Steal Ability for that.

                      SB/CS have been fluctuating quite a lot, so I'm staying with my current setting to get more games in to see where it lands on average over a longer run. Last set, the success rate was very high, but one previous to it, it was low... it depends on situations and speediness of teams, so it's not easy to nail down by a couple dozen games.
                      The Show CPU vs. CPU game stats: 2018,17,16,15,14,13,12,11

                      Comment

                      • ralphieboy11
                        Pro
                        • Jul 2005
                        • 543

                        #326
                        Re: Stats-Based CPU Sliders [MLB 12 Version]

                        Ok. I understand on the stolen base figures. Not enough statistical evidence yet. I think when you do find the optimal average, we may have a situation where the league leaders have more than normal, while some of the guys that may have normally had 5 to 10 might on a year might not have any.

                        I guess that's a problem that can happen with any stat though.

                        Comment

                        • nomo17k
                          Permanently Banned
                          • Feb 2011
                          • 5735

                          #327
                          Re: Stats-Based CPU Sliders [MLB 12 Version]

                          Originally posted by ralphieboy11
                          Ok. I understand on the stolen base figures. Not enough statistical evidence yet. I think when you do find the optimal average, we may have a situation where the league leaders have more than normal, while some of the guys that may have normally had 5 to 10 might on a year might not have any.

                          I guess that's a problem that can happen with any stat though.
                          What is your issue with steals? I don't feel there's any anomaly in how individuals steal. I've seen slow runners swipe on a surprise attempt, and fast ones are swiping as usual. Just curious.
                          The Show CPU vs. CPU game stats: 2018,17,16,15,14,13,12,11

                          Comment

                          • ralphieboy11
                            Pro
                            • Jul 2005
                            • 543

                            #328
                            Re: Stats-Based CPU Sliders [MLB 12 Version]

                            I don't have any statistical evidence to back up what I'm saying, so just ignore my comment. I just don't visually remember seeing slow runners steal, and some speedier guys seem to go every time.

                            Comment

                            • steviegolfballs
                              Rookie
                              • Feb 2010
                              • 243

                              #329
                              Re: Stats-Based CPU Sliders [MLB 12 Version]

                              Originally posted by nomo17k
                              With the OSFM Spring Training roster out last night, many of you guys presumably started your franchise last night. You must have spent an unhealthy amount of hours occupying PS3 since then, making your wife/GF very, very unhappy.

                              Tonight, instead of cuddling your DS3 controller, you need to spend time putting effort into pleasing your lady to save the relationship from totally falling apart.

                              I had to do my share by sliding.... BR Speed slider.

                              I just wanted to see base runners run at a realistic speed on base path. To do so, I made a bunch of LHBs with Speed ratings at 0, 50, and 99, make them bunt for hit, and measured how fast it took for them between the bat contact and reaching the first base. Each iteration, I changed BR Speed slider to 0, 5, and 10 to see how much the slider changes their speeds.

                              The MLB average is about 4.3 seconds. But that's for actually hitting, and not (regular) bunting that I used, so presumably my batter runners perhaps got better jumps.

                              What do I take from this?

                              I think the default (5) is fine. The median Speed rating for SCEA roster is about 40, and if you look at the lines that crosses Speed rating of 40, BR Speed slider at 5 cuts the MLB average line, perhaps slightly faster.

                              Given my guys perhaps got a better jump (and therefore yielded faster homt-to-1B times), maybe the slider value at 5 - 6 would replicate most realistic physical speed.

                              I'm going with 6 in hope of reducing GIDPs while maintaining visual authenticity.
                              Thanks for the insight.....and I am taking the girlfriend to see Les Miserables tonight....my punishment for pulling an all-nighter testing games last night

                              Comment

                              • nomo17k
                                Permanently Banned
                                • Feb 2011
                                • 5735

                                #330
                                Re: Stats-Based CPU Sliders [MLB 12 Version]

                                Originally posted by ralphieboy11
                                I don't have any statistical evidence to back up what I'm saying, so just ignore my comment. I just don't visually remember seeing slow runners steal, and some speedier guys seem to go every time.
                                That's actually normal. A vast majority of steals are attempted by established base stealers, and the rest are more of "surprise" kind... or failed hit-and-run turned successful steals...

                                A list of players that are not so fast (by Speed rating) but with more than one SB:

                                Matt Diaz
                                Chase Utley
                                Wilson Valdez
                                J.D. Martinez
                                Lance Berkman
                                Aaron Hill
                                Ryan Theriot
                                Mike Aviles
                                Brett Lawrie
                                Alex Gordon
                                Danny Valencia
                                Brian Dinkelman
                                David Murphy
                                Josh Hamilton
                                The Show CPU vs. CPU game stats: 2018,17,16,15,14,13,12,11

                                Comment

                                Working...