Minnesota Twins News & Rumors Forum
Page 8 of 8 FirstFirst ... 678
Results 141 to 152 of 152

Thread: Projections have Twins headed for fourth-straight 90-loss season

  1. #141
    Senior Member MVP
    Posts
    5,625
    Like
    1,121
    Liked 535 Times in 354 Posts
    I think the offense was hard to fix this year....there were either bad players, or elite players available, or 1B/3B types. Apparently we don't ever need another OF here, as the minors will fix that for us for sure......so there weren't a lot of good offensive guys to sign. And, that's how it is going to be, imo. Medium guys will be signed to extensions. Only expensive, cheap, or old/injured guys will be in FA in the future (offensively).
    Lighten up Francis....

  2. #142
    Twins News Team MVP
    Posts
    6,569
    Like
    828
    Liked 814 Times in 518 Posts
    Blog Entries
    3
    Quote Originally Posted by jay View Post
    http://www.fangraphs.com/depthcharts...n=ALL&teamid=8

    C - .697
    1B - .811
    DH - .745

    I must have added wrong earlier. That actually averages to .751. Makes that .737 look even more "fairly conservative" and .725 certainly wouldn't qualify as "optimistic". It probably wouldn't be the first time you've heard this and I don't mean this as a personal attack, but maybe you're being a bit too negative?

    All figures are an average of ZIPS and Steamer multiplied out by some playing time projections from the staff at FG. http://www.fangraphs.com/blogs/intro...and-standings/
    Therein lies the rub. Color me thoroughly unconvinced with some flawed or problematic assumptions.

    And ....it also turns out that my own net .825 OPS projection from 1B was very optimistic. I would argue that their projections from .745 OPS/DH and .697 OPS/C are similarly even more optimistic and wildly more conditionally-based....RE-

    .697 C?
    unless Pinto moves into the full-time role sooner rather than later....AND....hits like he did last September. This is a very optimistic forecast, that to this point doesn't fully reflect the very potential reality of the team starting out the season with Pinto in AAA and Suzuki as the starter and either/or Fryer/Hermann as the backup(s).... I would argue that my number was more realistic.

    .745 DH?
    and Jason Kubel gets 400+ PAs at DH and somehow semi-miraculously beats his average projection from 3 services (Steamer/Oliver/ZIPS) of .706 OPS-----the only way this .745 number is realistic is if Willingham successfully rebounds AND gets significant PAs at DH, AND that Mauer is factored in for far more DH ABs than I project. The DH net number I projected- being closer to .700 OPS, again seems more realistic.
    Last edited by jokin; 02-03-2014 at 06:08 PM.

  3. #143
    Senior Member Big-Leaguer jay's Avatar
    Posts
    872
    Like
    12
    Liked 49 Times in 30 Posts
    Blog Entries
    2
    You're more than welcome to disagree with the playing time projections that Fangraphs is using. However, I don't think you have any basis at all to claim that ZIPS and Steamer are anything other than completely neutral.

    I'm not sure how ZIPS/Steamer having Pinto at a .723 is hitting "like he did last September", but let's not let hyperbole and selective data get in the way of making a point... certainly no shortage of that elsewhere.

    Let's compare, 1B/DH/C for 2014:
    Fangraphs - .751
    Me - .737
    You - .725 and that's "very generous" with copious usage of "optimistic"

    Okay...

  4. #144
    Senior Member Triple-A
    Posts
    204
    Like
    106
    Liked 22 Times in 15 Posts
    Blog Entries
    1
    If there are injuries to any of the FA pitchers then it will be a 90 loss season since the offense wasn't improved perhaps another 96 loss season.

  5. #145
    Twins News Team MVP
    Posts
    6,569
    Like
    828
    Liked 814 Times in 518 Posts
    Blog Entries
    3
    Quote Originally Posted by jay View Post
    You're more than welcome to disagree with the playing time projections that Fangraphs is using.

    However, I don't think you have any basis at all to claim that ZIPS and Steamer are anything other than "completely neutral".

    I'm not sure how ZIPS/Steamer having Pinto at a .723 is hitting "like he did last September", but let's not let hyperbole and selective data get in the way of making a point... certainly no shortage of that elsewhere.

    Let's compare, 1B/DH/C for 2014:
    Fangraphs - .751
    Me - .737
    You - .725 and that's "very generous" with copious usage of "optimistic"

    Okay...
    Okay...misusage of hyperbole or mistatement of facts much? Where did I say that Fangraphs wasn't neutral? I did say that that they are using projections that don't appear to match with how the Twins have stated that they intend to use their roster-- (FWIW, the main Roto-Fantasy-League projectors all have even lower production forecasts for the players involved in this discussion).

    To wit: Oliver and ZIPS have Pinto somehow playing 143 games. At this point, Steamer probably has taken the more realistic assumption with 81 games as their guess. Average the 3 together and you come up with a .716 OPS for Pinto.

    I'm going with Steamer and the indication of the Twins intent, plus Pinto's injury and Winter ball shutdown (Ryan has announced that Suzuki is the Twins opening day catcher and Pinto is indicated to be headed to AAA, at this point, anyway)----

    while you're going with the other 2 services out of at least a half a dozen more that disagree with both you and them.

    Suzuki, among the 3 services I used, averages to a projected OPS of .656, with Steamer giving Suzuki the extreme outlier, higher projection, of .677 playing half-time. Every other projection out there, except Steamer, has Suzuki OPSing between .623 and .645.

    What was Suzuki's combined OPS for 2012 and 2013, you ask? .614!
    Again, who is being optimistic here, who is being hyperbolic and who is being realistic?

    If you assume that Pinto stays down until the All Star break, that would mean that Fryer/Hermann play in 20 games for around 80 PAs at a .611 OPS.

    Assuming an 80/20 split of the PAs to the break, the combined OPS at Catcher would be .645. With makeup games added there will be about 70-72 games left after the All Star game, so let's say that Suzuki is then traded- that leaves 60 games for Pinto? Or...let's assume, very generously and optimstically and probably unrealistically, that Pinto gets called up even sooner than the All Star game, say in mid- June, Fryer is DFAd instead, and Suzuki is basically relegated to one or two days a week duty for the rest of the season- and Pinto STILL ends up playing 60 games at Catcher. According to the projections, that would mean that the combined OPS at the position would be around ~.655- 670 OPS.

    This generous and optimistic forecast at the Catcher position would mean that, in order to hit your own combined OPS number at C/1B/DH, that Kubel has to come through above the lofty projections despite his very questionable status, and/or that Willingham also:

    comes back strong from his injury-plagued-year,
    gets significant DH PAs,
    AND is not traded away in June or July.

    A lot of things have to hit just right for your scenario to come to fruition, I just don't see it. And the combined OPS number comes in right around mine if we hit these rosy forecasts and only end up getting .811 OPS combined from the 1B postion as is projected in your own scenario.
    Last edited by ashburyjohn; 02-06-2014 at 11:21 PM.

  6. #146
    Senior Member All-Star
    Posts
    1,556
    Like
    3
    Liked 310 Times in 194 Posts
    Blog Entries
    1
    Quote Originally Posted by jokin View Post
    To wit: Oliver and ZIPS have Pinto somehow playing 143 games.
    For the record, ZiPS doesn't "project" playing time. (Oliver probably doesn't either.) Most projection systems just use a player's historical MLB/MiLB playing time as guide to present counting stats, with no regard to a player's current health or role. The rate stats are what's important there.

    Then, to project standings, Fangraphs or Clay Davenport or whoever will make subjective estimates on playing time. Davenport has him at 176 PA, and Suzuki at 358:

    http://www.claydavenport.com/projections/MIN.shtml

    Fangraphs has Pinto at 320 PA, Suzuki 288:

    http://www.fangraphs.com/depthcharts...n=ALL&teamid=8

    I might guess the former split is more likely, but referencing the 563 ZiPS number in any critique is completely meaningless.

  7. #147
    Senior Member Triple-A
    Posts
    287
    Like
    0
    Liked 43 Times in 32 Posts
    So, two "experts" disagree on how many PA's the two catchers get, how really important is that in the whole production or lack thereof in the entire lineup? I would think PA's plus production in traditional hitting spots in the entire batting order is much more important than wondering what two rather pedestrian catchers give to the lineup. If the so-called "big guns" don't produce 90+ losses again are staring us in the face.

  8. #148
    Senior Member Big-Leaguer jay's Avatar
    Posts
    872
    Like
    12
    Liked 49 Times in 30 Posts
    Blog Entries
    2
    Quote Originally Posted by Kwak View Post
    This has been a spirited discussion about Twins players and OPS. I am curious to learn what is the impact on the team won/lost record given the various conclusions of OPS? Is it like 1 or maybe two wins?--or a substantial number?
    There's a neat little formula where you can find that.
    Win Ratio = 4*(OPS/OPS Allowed) - 3

    So, a team that is perfectly average in both OPS for and against has a win ratio of 1 with an expectation of 81 wins. We've been debating 3 positions summed together, so a third of the lineup. If you change the production from those 3 spots by .025 (for example, .725 vs .750), it would affect the team OPS by a third of that. Let's call .725 the average and see what impact it has:

    WR = 4*(.733/.725) - 3 = 1.046

    Win Ratio can be converted to a more understandable winning percentage easily. W% = WR/(WR+1)

    W% = 1.046/(1.046+1) = .511
    162 * .511 = 82.8

    So, the team that was perfectly average saw their expected wins go up by 1.8 by getting a .750 OPS instead of .725 from a third of their lineup.
    Last edited by jay; 02-07-2014 at 09:26 AM. Reason: convert to W%

  9. These 4 users like jay's post and want to buy him/her a steak dinner:

    ericchri (02-07-2014), Oldgoat_MN (02-07-2014), PseudoSABR (02-07-2014), Willihammer (02-07-2014)

  10. #149
    Twins News Team MVP
    Posts
    6,569
    Like
    828
    Liked 814 Times in 518 Posts
    Blog Entries
    3
    Quote Originally Posted by spycake View Post
    For the record, ZiPS doesn't "project" playing time. (Oliver probably doesn't either.) Most projection systems just use a player's historical MLB/MiLB playing time as guide to present counting stats, with no regard to a player's current health or role. The rate stats are what's important there.

    Then, to project standings, Fangraphs or Clay Davenport or whoever will make subjective estimates on playing time. Davenport has him at 176 PA, and Suzuki at 358:

    http://www.claydavenport.com/projections/MIN.shtml

    Fangraphs has Pinto at 320 PA, Suzuki 288:

    http://www.fangraphs.com/depthcharts...n=ALL&teamid=8

    I might guess the former split is more likely, but referencing the 563 ZiPS number in any critique is completely meaningless.
    I'm not following you completely here. Why does ZiPS use the 563 PAs and 143 games for Pinto if they're completely meaningless numbers? Why publish all of the expected "counting stats"?

    Why not just release the rate stats without all of the further detail if it's so meaningless?

  11. #150
    Senior Member All-Star
    Posts
    1,219
    Like
    100
    Liked 95 Times in 69 Posts
    Quote Originally Posted by jay View Post
    There's a neat little formula where you can find that.
    Win Ratio = 4*(OPS/OPS Allowed) - 3

    So, a team that is perfectly average in both OPS for and against has a win ratio of 1 with an expectation of 81 wins. We've been debating 3 positions summed together, so a third of the lineup. If you change the production from those 3 spots by .025 (for example, .725 vs .750), it would affect the team OPS by a third of that. Let's call .725 the average and see what impact it has:

    WR = 4*(.733/.725) - 3 = 1.046

    Win Ratio can be converted to a more understandable winning percentage easily. W% = WR/(WR+1)

    W% = 1.046/(1.046+1) = .511
    162 * .511 = 82.8

    So, the team that was perfectly average saw their expected wins go up by 1.8 by getting a .750 OPS instead of .725 from a third of their lineup.

    Thank you for responding.

    The referenced formula seems simplistic and has no qualifiers--I'll use it for discussion purposes. The expected 1.8 additional wins generated by this incremental OPS would be well within the error margin of the original estimate of wins postulated in the thread and likely within the error limit of the cited formula itself! Ergo, noise.

  12. #151
    Senior Member Big-Leaguer jay's Avatar
    Posts
    872
    Like
    12
    Liked 49 Times in 30 Posts
    Blog Entries
    2
    Quote Originally Posted by Kwak View Post
    Thank you for responding.

    The referenced formula seems simplistic and has no qualifiers--I'll use it for discussion purposes. The expected 1.8 additional wins generated by this incremental OPS would be well within the error margin of the original estimate of wins postulated in the thread and likely within the error limit of the cited formula itself! Ergo, noise.
    Well, I didn't want to define the entire concept. Here's an excellent link if you'd like to read more about it: http://walksaber.blogspot.com/2009/0...ps-tricks.html

    You're probably right that either result is well within the error range for both scenarios. However, if that's your definition of what makes it noise, there wouldn't be a single move you could make that wouldn't qualify. Even a big free agent signing and the expected gain in wins would be well within the error range of what could actually happen without that FA given all the dynamics of an individual season.

    If you played the season a million times, you'd find these teams would get extremely close to averaging the 81 and 82.8 numbers respectively. Although individual seasons are within error range of each other, that expected number doesn't change.

  13. #152
    Senior Member All-Star
    Posts
    1,556
    Like
    3
    Liked 310 Times in 194 Posts
    Blog Entries
    1
    Quote Originally Posted by jokin View Post
    I'm not following you completely here. Why does ZiPS use the 563 PAs and 143 games for Pinto if they're completely meaningless numbers? Why publish all of the expected "counting stats"?

    Why not just release the rate stats without all of the further detail if it's so meaningless?
    Illustrative purposes only. Good for vets like Mauer, not so much for rookies like Pinto who will be battling for playing time (or even a roster spot).

    Basically, depth charts, roster composition, and health reports are not direct inputs into ZiPS. I think this is covered in most overviews of the system.

Page 8 of 8 FirstFirst ... 678

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
©2014 TwinsCentric, LLC. All Rights Reserved.
Interested in advertising with Twins Daily? Click here.