Announcement

Collapse
No announcement yet.

How strong is BPV?

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

  • #31
    Bunch of points:

    First, don't consider the "5" to "3" downgrade of Ruiz's PT a "40% reduction". That's not how those numbers work. He could lose one single AB, and if it lowered his projected PA from 450 to 449, that would cost him the same deduction, per the MM tables:

    PLAYING TIME
    PA MM Role
    0 - 99 0 Non-factors
    100 - 249 1 Fringe/bench players
    250 - 449 3 Mid-timers
    450+ 5 Potential full-timers
    But getting back to the Posey/Ruiz comparison, our disconnect is that MM is measuring skills, and you're countering with stats. We are using PX in the MM calculations, not projected HR totals. That means that your disagreement isn't with Mayberry as much as with the whole concept of component skills analysis, which is the underpinning of just about everything we do around here.

    Ruiz's 3-year PX levels: 102, 70, 135. Projection for 2013: 110
    Posey's 3-year PX levels: 121, 69, 137. Projection for 2013: 113

    Ruiz's 3-year HR totals: 30 in 1150 AB, or one every 38 AB.
    Posey's 3-year HR totals: 46 in 1100 AB, or one every 24 AB.

    No question, there's a noticeable advantage to Posey in terms of HR output. But as you can see from the PX lines, by that metric (which is all MM looks at) they are pretty much the same guy in terms of power skill. PX doesn't perfectly correlate with HR, but it does pretty well. So there are three possibilities to explain this:

    1. PX is just doing a bad job of reflecting the power of one or both of them.
    2. PX is a leading indicator that Ruiz has some latent power that we haven't seen manifest in terms of his HR production yet.
    3. PX is a leading indicator that Posey's HR output is a little soft as compared to his power skill and we should expect some regression.

    I know you're new to the forums, not sure how new you are to the site overall.... but it's basically our first commandment around here to trust the skills indicators.

    Now, PX isn't the only indicator of power, of course. In the Forecaster (and coming soon to the site) we also use xPX to validate power output. Here are the xPX levels of these two guys for the same time period:

    Ruiz's 3-year PX levels: 87, 95, 131.
    Posey's 3-year PX levels: 121, 84, 104.

    Based on that, combined with the above, the conclusion you should at least start to come to based on this data is that Posey's power skill is somewhat soft, certainly softer objectively than you would get from subjective, unmeasurable characterizations like "still-improving MVP".

    Again, not arguing that I'd rather own Ruiz than Posey. But I am arguing that this is exactly what MM is designed to do: measure players by their relative skills rather than by subjective perception.

    Comment


    • #32
      You've got me pegged, Ray. I am brand new to the forums and starting my second season with the site. I know I am seldom the smartest guy in the room and that bad luck is not why I have not won my league since 1999. I will also "confess" that we remain proudly 4x4 after 25 years. I am trying to embrace new ideas like Mayberry and BPV - not dispute them...BUT...

      I have to disagree with you on the 40% downgrade of Ruiz. If I use Shandler's formula to produce a single Mayberry rating for Ruiz, he is a 65 based on 3145 and a 33 based on 3143, because the last number is factored twice. That's actually a 50% downgrade, isn't it?

      You said that my "disagreement isn't with Mayberry as much as with the whole concept of component skills analysis". I agree with the first part. I love the simple 4-part system and embracing imprecision. I don't think I disagree with the concept of skills analysis as much as I have trouble swallowing some of the rusults this analysis produces. B. Posey and C. Ruiz have the same power rating? J. Lucroy and A. McCutchen have the same speed rating? N. Aoki and Josh Hamilton had the same 2012 Basic Player Value? Do you buy any of that?

      I have always felt that knowing nothing is difficult, but knowing things that aren't true is just plain dangerous. Help me believe, Ray!!
      Last edited by kjordan; 01-17-2013, 02:04 PM.

      Comment


      • #33
        Let's forget about the Ruiz playing time discrepancy. The book's playing time projections are set in October, they're already dated, not just for Ruiz but for a ton of people. He was projected for 411 AB back then, it's 369 now. That moved him across the "5 vs. 3" threshold. A 3 seems entirely appropriate now. Can we agree on that?

        As for your comparisons:

        - I covered the the Posey/Ruiz power comparison above
        - Lucroy and McCutchen don't have the smae speed rating. McCutchen's a 4 and Lucroy's a 3.
        - one of my earlier posts talked about the perils of single-number comparison with BPV. Having the same BPV doesn't mean that two players have the same skill profile. They are comparably skilled if you were to rank players by BPV, but it doesn't mean they have the *same* skills.

        I've been around here for going on 15 years... so yes, I buy everything I'm telling you here. But we have newcomers every year who fight through the same things you're fighting through now. The fact that so many of them are still here and now touting how much they believe in this stuff now should tell you something. I know it's a leap of faith... but take this for a test drive, do some mock drafts with a MM-sorted cheat sheet or something. Put it through its paces.

        Comment


        • #34
          Originally posted by kjordan View Post
          You've got me pegged, Ray. I am brand new to the forums and starting my second season with the site. I know I am seldom the smartest guy in the room and that bad luck is not why I have not won my league since 1999. I will also "confess" that we remain proudly 4x4 after 25 years. I am trying to embrace new ideas like Mayberry and BPV - not dispute them...BUT...
          At the risk of butting into your conversation with Ray, the Mayberry Method isn't for everyone. It's been very gratifying to read all the compliments about MM in the forums this off-season, so unscientifically I'm ready to crown it a winner, but I don't think any of us claim that it's the right strategy for everyone.

          I personally take the opposite approach: I embrace precision. I tweak the projections considerably, recomputing R$ for everyone on the day of my league's auction. Give it to me in pennies, with a second column with 80% of R$ please. However, I think the result is sort of the same. When the auction starts, I no longer worry about projections at all. That stage of the game is in the past because all of that intuition and research is baked into the output. I'm free to concentrate only on the auction aspects of the game: tracking my money, thinking several players ahead about whether I'm spending too much or too little, penciling in multiple players per budgeted open slot on my roster to make sure I'm not getting shut out of a needed commodity, etc.

          If Mayberry Method frees you up to focus on the auction aspects of the competition on draft day, then hooray, but that's not the only path to that result.
          "If you torture data long enough, they will confess." -- Ronald Coase

          Comment


          • #35
            Good stuff Michael. Totally agree. We aren't a one-size fits all operation. kjordan is of course well within his rights to say "Mayberry's not for me", and will still hopefully get plenty of value from the site. If that point got lost here, I apologize.

            We all have our systems that work for us. For me, Rotolab is my crutch... let's me see MM, dollar values, whatever else all in one dashboard. I don't think I could draft without it. I've just trained my brain over the years to think along a draft that way.

            Comment


            • #36
              I'm gonna do that, Ray. Thanks for all your time & patience. Maybe we will talk again.

              Have a great season.

              Comment


              • #37
                Originally posted by RAY@HQ View Post
                The idea is to embrace imprecision.
                My life, in a few words.
                RIP Paco de Lucia.

                Comment


                • #38
                  Is there a Mayberry list in the Forecaster, like last year? I can't seem to find it.
                  What's the rumpus?

                  Comment


                  • #39
                    Originally posted by Old Tom View Post
                    Is there a Mayberry list in the Forecaster, like last year? I can't seem to find it.
                    Universal Draft Guide is playing this role. It's got the MM codes on it.

                    Comment


                    • #40
                      Have the Mayberry ranges changed? I'm just trying to understand because I sometimes find guys who RSpd doesn't fit a range and it comes out as a 5 when the table says it should be a 4. I'll have to find the examples again but for 2013, what are the correct ranges?

                      Comment


                      • #41
                        The ranges changed with the conversion to RSpd. Current ones are in the glossary.

                        Comment


                        • #42
                          Thanks as always Ray!

                          Comment

                          Working...
                          X