Minnesota Twins News & Rumors Forum
Page 5 of 8 FirstFirst ... 34567 ... LastLast
Results 81 to 100 of 152

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

  1. #81
    Senior Member All-Star
    Posts
    1,416
    Like
    64
    Liked 40 Times in 28 Posts
    Quote Originally Posted by old nurse View Post
    Does the model measure what it is supposed to? I can see that 30% of the time the model is off by more than 10 games for team wins. So given the data they have they are grossly inaccurate. With an error bar that large they should lose credibility.
    This is something that has been bothering me re:statistical modeling both in regards to this application but more in relation to all modeling and metrics with a baseball slant whether that is WAR, SIERA, xFIP, PECOTA, Steamer or whatever.

    There is a certain segment of the baseball consuming public that latches onto these new mathematical analysis tools and thinks they are great. They are helping to enhance our understanding of the game. On the opposite end of the spectrum there is a segment of fans that seems to feel that these tools weren't needed by the greats of the past and they aren't needed now. There is nothing new that needs to be learned about the game and any attempt to replace RBI's and Wins and Runs Scored and whatever else with a new statistic is anathema. I am not trying to pigeon hole anyone here on this board into either of these groups, certainly not meaning to imply anything about Old Nurse so please don't be offended. There are as many positions on this subject as there are frequnecies in the EM spectrum, surely. These just seem to be the opposite ends of that pendulum.

    I certainly identify myself with the former group. I love using math as a way to view the world. I find that it helps to clarify things if I can use quantitative analysis. So, I freely admit that I am biased towards these new constructs because of my larger world view. OK enough rambling...to my point.

    In the quoted post the author states that this model is grossly wrong 30% of the time. For simplicities sake let's assume that means the metric is mostly correct 70% of the time. The implied message being that this is a worthless tool and shouldn't be given much credence.

    My question is what is the alternative? "Gut feeling"? There is a place in the world for "gut feelings" for certain, but by and large they aren't very accurate. Look no further than houses of gambling, NCAA tournament brackets, the lottery, or the ever popular "how many jelly beans are in this jar" game if you're unsure. In fact my guess is most statistical modeling is derived from a desire to get away from gut feeling guesses.

    So here are my questions: Is there an alternative to both statistical modeling and gut feelings? Is it more accurate? How accurate is human intuition? How well did Keith Law, Jayson Stark or Gleeman and the Geek do at predicting 2013's final standings? How good is your intuition? If you want me to believe that a computer model is crappy than show me what I should be paying attention to instead, please.

    Sorry, this turned into much more of a rant than I meant it to be and considerably longer. I obviously don't know what the word concise means and got a double helping of verbose instead. Again Old Nurse and everyone else, this was not directed at you specifically, your post was just the tipping point to something that had been bubbling up within me for a while.

    TL : DR My point was not to offend with this but rather to say, if you want me to believe you, show some evidence that there is a better alternative. 70% correct is clearly not perfect, but if the alternative is only correct 60% of the time, isn't that a step in the right direction?
    Last edited by Oxtung; 01-29-2014 at 12:30 AM. Reason: clarification

  2. #82
    Senior Member All-Star
    Posts
    1,416
    Like
    64
    Liked 40 Times in 28 Posts
    Quote Originally Posted by jay View Post
    Unless you're a Cuban first baseman for the White Sox.
    I wonder if he could be being comped to Puig, Cespedes, Ichiro, etc... My gut (oh god did I just say that after my stupid rant?!?) tells me that imports from other major leagues both succeed more frequently than prospects but also have much better first seasons. I have absolutely nothing to back that up with though.

  3. #83
    Senior Member All-Star Hosken Bombo Disco's Avatar
    Posts
    1,038
    Like
    1,635
    Liked 488 Times in 285 Posts
    Quote Originally Posted by Oxtung View Post
    This is something that has been bothering me re:statistical modeling both in regards to this application but more in relation to all modeling and metrics with a baseball slant whether that is WAR, SIERA, xFIP, PECOTA, Steamer or whatever.

    There is a certain segment of the baseball consuming public that latches onto these new mathematical analysis tools and thinks they are great. They are helping to enhance our understanding of the game. On the opposite end of the spectrum there is a segment of fans that seems to feel that these tools weren't needed by the greats of the past and they aren't needed now. There is nothing new that needs to be learned about the game and any attempt to replace RBI's and Wins and Runs Scored and whatever else with a new statistic is anathema. I am not trying to pigeon hole anyone here on this board into either of these groups, certainly not meaning to imply anything about Old Nurse so please don't be offended. There are as many positions on this subject as there are frequnecies in the EM spectrum, surely. These just seem to be the opposite ends of that pendulum.

    .....
    I loved reading this. To be brief in reply, I welcome the new detailed statistical tools, but I don't bury my head in them. I'm grateful that others do, because as you said, they can enhance our understanding, and I guess that makes me a freeloader to some degree, letting the rest of you study in detail what I am happy to know in the abstract. For example the stats tell us that Pelfrey and Hughes are fly ball pitchers with possibly high FB% (verify) and I wouldn't have known that without being on Twins Daily; and won't an outfield of Buxton & Hicks be nice then? And we are just a year away from that. So with this knowledge, I am optimistic Hughes and Pelfrey were good signings. Kudos to Ryan.

    But in the end, as a casual fan, BABIP doesn't tell me much more than simple batting average already tells me. There is a degree of chance, luck, funny hops and diving grabs that happen in every game that will skew any statistic in any direction.. Same with SIERA and ERA+ for pitchers. I do take heed to the advanced stats, but at the same time, basic ERA is simple, familiar, and usually tells its piece of the story accurately enough.

  4. This user likes Hosken Bombo Disco's post and wants to buy him/her a steak dinner:

    Halsey Hall (01-29-2014)

  5. #84
    Senior Member Big-Leaguer jay's Avatar
    Posts
    884
    Like
    13
    Liked 50 Times in 31 Posts
    Blog Entries
    2
    Quote Originally Posted by Hosken Bombo Disco View Post
    But in the end, as a casual fan, BABIP doesn't tell me much more than simple batting average already tells me. There is a degree of chance, luck, funny hops and diving grabs that happen in every game that will skew any statistic in any direction.. Same with SIERA and ERA+ for pitchers. I do take heed to the advanced stats, but at the same time, basic ERA is simple, familiar, and usually tells its piece of the story accurately enough.
    I can certainly respect that view. BABIP was greatly misused when it was first bandied about by saying so-and-so was lucky or unlucky with a simple glimpse at the number. It goes much deeper than that and is dependent on the hitter's profile and tools. I do like some of the new pitcher metrics because, frankly, ERA is rotten. The definition of earned runs is convoluted and it can't measure the huge impact of defense outside of that other convoluted measure, errors.

    My mind was blown when I was introduced to advanced metrics probably 5 years ago. It's amazing how far the studies and measures have come even since then. At some point, a certain level of consolidation and common acceptance needs to occur. There's also lots of work left to do on assigning individual's value. With so many variables, it'll be quite the quest to truly master and measure it all.

  6. #85
    Senior Member Big-Leaguer jay's Avatar
    Posts
    884
    Like
    13
    Liked 50 Times in 31 Posts
    Blog Entries
    2
    Quote Originally Posted by Oxtung View Post
    I wonder if he could be being comped to Puig, Cespedes, Ichiro, etc... My gut (oh god did I just say that after my stupid rant?!?) tells me...
    Likely so, but it still seems exorbitant. The overall success rate might be higher than a rook from MiLB, but there's been plenty of international imports that flopped... or at least that's what my gut tells me.

  7. This user likes jay's post and wants to buy him/her a steak dinner:

    Oxtung (01-29-2014)

  8. #86
    Senior Member All-Star JB_Iowa's Avatar
    Posts
    3,316
    Like
    1,300
    Liked 1,279 Times in 740 Posts
    Quote Originally Posted by Oxtung View Post
    My question is what is the alternative? "Gut feeling"? There is a place in the world for "gut feelings" for certain, but by and large they aren't very accurate. Look no further than houses of gambling, NCAA tournament brackets, the lottery, or the ever popular "how many jelly beans are in this jar" game if you're unsure. In fact my guess is most statistical modeling is derived from a desire to get away from gut feeling guesses.
    Advanced metrics can enhance understanding of the game but for many fans (myself included), that is all they do --- enhance understanding. Some of us don't aspire to be general managers (even of a fantasy team) or to work for a baseball team in any other manner. Nor do we feel a compulsion to prove that we are the "smartest kid in the class".

    For me, while I learn something from the advanced stats and dig through the info when someone presents it, I don't particularly want to spend a lot of time or energy making understanding them a highlight of my life. Nor do I want to get drawn into long discussions of statistics.

    I really just want to sit and enjoy the game. And knowing some advanced stats can enhance that. And things like Parker's analysis of a player (one of my favorite things on TD) can enhance that because I continually learn more about the game. But sometimes just watching and enjoying those "gut feelings" is good enough, too.

    Baseball (especially in the winter/spring) is about HOPE. The possibility that THIS will be the year. The possibility that a player will be better than we expect (maybe even better than we dream). The possibility that a young player will have a breakout season. The possibility that everything will "go right".

    It isn't what I expect but can be what I hope for. And its why I kept watching the Twins even when all the losses mounted up the last few years. If I place too much emphasis on advanced stats, I have a "gut feeling" that I would conclude that "all hope is gone".

  9. #87
    Senior Member All-Star
    Posts
    1,284
    Like
    252
    Liked 288 Times in 150 Posts
    Using old line that follows the phrase "while the Twins are a bad team on paper", I will sum up my attitude toward projections by saying "but they don't play their games in the projects". Let's hope they surprise.

  10. This user likes big dog's post and wants to buy him/her a steak dinner:

    Willihammer (01-29-2014)

  11. #88
    Senior Member All-Star
    Posts
    1,662
    Like
    11
    Liked 54 Times in 35 Posts
    Quote Originally Posted by Oxtung View Post
    This is something that has been bothering me re:statistical modeling both in regards to this application but more in relation to all modeling and metrics with a baseball slant whether that is WAR, SIERA, xFIP, PECOTA, Steamer or whatever.

    There is a certain segment of the baseball consuming public that latches onto these new mathematical analysis tools and thinks they are great. They are helping to enhance our understanding of the game. On the opposite end of the spectrum there is a segment of fans that seems to feel that these tools weren't needed by the greats of the past and they aren't needed now. There is nothing new that needs to be learned about the game and any attempt to replace RBI's and Wins and Runs Scored and whatever else with a new statistic is anathema. I am not trying to pigeon hole anyone here on this board into either of these groups, certainly not meaning to imply anything about Old Nurse so please don't be offended. There are as many positions on this subject as there are frequnecies in the EM spectrum, surely. These just seem to be the opposite ends of that pendulum.

    I certainly identify myself with the former group. I love using math as a way to view the world. I find that it helps to clarify things if I can use quantitative analysis. So, I freely admit that I am biased towards these new constructs because of my larger world view. OK enough rambling...to my point.

    In the quoted post the author states that this model is grossly wrong 30% of the time. For simplicities sake let's assume that means the metric is mostly correct 70% of the time. The implied message being that this is a worthless tool and shouldn't be given much credence.

    My question is what is the alternative? "Gut feeling"? There is a place in the world for "gut feelings" for certain, but by and large they aren't very accurate. Look no further than houses of gambling, NCAA tournament brackets, the lottery, or the ever popular "how many jelly beans are in this jar" game if you're unsure. In fact my guess is most statistical modeling is derived from a desire to get away from gut feeling guesses.

    So here are my questions: Is there an alternative to both statistical modeling and gut feelings? Is it more accurate? How accurate is human intuition? How well did Keith Law, Jayson Stark or Gleeman and the Geek do at predicting 2013's final standings? How good is your intuition? If you want me to believe that a computer model is crappy than show me what I should be paying attention to instead, please.

    Sorry, this turned into much more of a rant than I meant it to be and considerably longer. I obviously don't know what the word concise means and got a double helping of verbose instead. Again Old Nurse and everyone else, this was not directed at you specifically, your post was just the tipping point to something that had been bubbling up within me for a while.

    TL : DR My point was not to offend with this but rather to say, if you want me to believe you, show some evidence that there is a better alternative. 70% correct is clearly not perfect, but if the alternative is only correct 60% of the time, isn't that a step in the right direction?
    The question is, of what use is something that is grossly wrong 30% of the time? It does not provide useful material if it is that wrong. The argument that it is better than anything else is hollow. It is not anti statistic argument that I have. It is that the statical method used is grossly flawed. If your analysis of a 40 man roster leaves you that far in error, you have bad analysis. Figure out why you are wrong so often and try again. I said use it for entertainment value. That is all Davenport model is good for. It cannot predict improvements in player. It anticipates declines for individual players that may or may not be there. It has no way to determine the effect of changing teams will benefit/hurt a player. The statistical bias of a design is no different than a gut feeling of an individual that spends a lifetime involved in a game.

  12. #89
    Senior Member All-Star crarko's Avatar
    Posts
    1,220
    Twitter
    @crarko
    Like
    155
    Liked 775 Times in 407 Posts
    Quote Originally Posted by old nurse View Post
    The question is, of what use is something that is grossly wrong 30% of the time?
    Weather forecasting?

  13. #90
    Twins Moderator All-Star diehardtwinsfan's Avatar
    Posts
    4,373
    Like
    414
    Liked 810 Times in 510 Posts
    Blog Entries
    3
    Quote Originally Posted by Oxtung View Post
    This is something that has been bothering me re:statistical modeling both in regards to this application but more in relation to all modeling and metrics with a baseball slant whether that is WAR, SIERA, xFIP, PECOTA, Steamer or whatever.

    There is a certain segment of the baseball consuming public that latches onto these new mathematical analysis tools and thinks they are great. They are helping to enhance our understanding of the game. On the opposite end of the spectrum there is a segment of fans that seems to feel that these tools weren't needed by the greats of the past and they aren't needed now. There is nothing new that needs to be learned about the game and any attempt to replace RBI's and Wins and Runs Scored and whatever else with a new statistic is anathema. I am not trying to pigeon hole anyone here on this board into either of these groups, certainly not meaning to imply anything about Old Nurse so please don't be offended. There are as many positions on this subject as there are frequnecies in the EM spectrum, surely. These just seem to be the opposite ends of that pendulum.

    I certainly identify myself with the former group. I love using math as a way to view the world. I find that it helps to clarify things if I can use quantitative analysis. So, I freely admit that I am biased towards these new constructs because of my larger world view. OK enough rambling...to my point.

    In the quoted post the author states that this model is grossly wrong 30% of the time. For simplicities sake let's assume that means the metric is mostly correct 70% of the time. The implied message being that this is a worthless tool and shouldn't be given much credence.

    My question is what is the alternative? "Gut feeling"? There is a place in the world for "gut feelings" for certain, but by and large they aren't very accurate. Look no further than houses of gambling, NCAA tournament brackets, the lottery, or the ever popular "how many jelly beans are in this jar" game if you're unsure. In fact my guess is most statistical modeling is derived from a desire to get away from gut feeling guesses.

    So here are my questions: Is there an alternative to both statistical modeling and gut feelings? Is it more accurate? How accurate is human intuition? How well did Keith Law, Jayson Stark or Gleeman and the Geek do at predicting 2013's final standings? How good is your intuition? If you want me to believe that a computer model is crappy than show me what I should be paying attention to instead, please.

    Sorry, this turned into much more of a rant than I meant it to be and considerably longer. I obviously don't know what the word concise means and got a double helping of verbose instead. Again Old Nurse and everyone else, this was not directed at you specifically, your post was just the tipping point to something that had been bubbling up within me for a while.

    TL : DR My point was not to offend with this but rather to say, if you want me to believe you, show some evidence that there is a better alternative. 70% correct is clearly not perfect, but if the alternative is only correct 60% of the time, isn't that a step in the right direction?
    I think these tools are neat and funn to talk about, but the one thing I don't see a lot of is people going back and looking at what they predicted to how they actually did and compared the tools to each other and to the "experts"... I don't think it's too much to ask that we look at the actual results of said tool before we start placing value in them.

  14. This user likes diehardtwinsfan's post and wants to buy him/her a steak dinner:

    Oxtung (01-29-2014)

  15. #91
    Twins News Team All-Star PseudoSABR's Avatar
    Posts
    1,954
    Like
    257
    Liked 206 Times in 115 Posts
    My question is what is the alternative? "Gut feeling"? There is a place in the world for "gut feelings" for certain, but by and large they aren't very accurate.
    What assurances do I have that the gut feelings have been objectively eliminated from these metrics? Each metric produces varying results because they weigh different aspects of different values--which is a reflection of their bias. Again, there's no mathematical innocence or objectivity at work here.

    For instance if you believe that walks are more valuable than hits in determining winning outcomes--how do you determine that difference? Do you look at runs scored? The same runs that we might dismiss in having value in determining player worth?

    Objectivity is a tenuous goal when we are looking at a phenomenons that are invariably abstract and lack reducible modalities. There's non-mathematical biases at work when you start dividing the field into zones or when you attempt to combine defense and hitting and baserunning into the same metric. As useful as it might be, the need to quantify an abstract phenomenon changes the nature of the phenomenon.

    I worry that the quality assurance of such metrics simpy confirms the makers' deep-seated biases (Well these numbers look right because I generally value speed and defense than average). How do you assure the accuracy of these models if not by the models themselves? These kind statistical models night only be self-affirming.

    For my own part, I like to look at individual pieces of data--but I like to do my own 'math' when determining a players total value or rather do my own thinking in which aspects of the game have more weight than others.
    Last edited by PseudoSABR; 01-29-2014 at 01:06 PM.

  16. These 2 users like PseudoSABR's post and want to buy him/her a steak dinner:

    savvyspy (01-29-2014), SydneyTwinsFan (01-29-2014)

  17. #92
    Senior Member Double-A
    Posts
    107
    Like
    50
    Liked 14 Times in 11 Posts
    If the offense regresses (very possible) its not likely to be made up for by slightly improved (Likely) pitching and the fact that defensively they are below average. They are probably right where they usually are at 68-72 wins.

    I'm not sure about why they are predicting Mauer to regress. That seems a bit odd. I think its much more likely that Dozier regresses (he's never had 2 good seasons in a row) and that Suzuki gives you nothing. Overall I don't see how this team scores runs. I don't see anyone but Mauer and Arcia hitting over .250, they have ZERO power, and they don't walk a ton.

  18. #93
    Senior Member Big-Leaguer jay's Avatar
    Posts
    884
    Like
    13
    Liked 50 Times in 31 Posts
    Blog Entries
    2
    Quote Originally Posted by savvyspy View Post
    If the offense regresses (very possible) its not likely to be made up for by slightly improved (Likely) pitching and the fact that defensively they are below average. They are probably right where they usually are at 68-72 wins.
    Everything I've seen has the Twins offense regressing, but that regression is towards the mean (aka... better). Regression describes returning towards the mean whether the starting point is bad or good. It doesn't describe going from bad to even worse, which is what I think you're saying?

    Help me understand which positions you think will put up worse production than last year? I'm not seeing many ways for the offense to be worse (barring significant injury).

  19. #94
    Senior Member All-Star crarko's Avatar
    Posts
    1,220
    Twitter
    @crarko
    Like
    155
    Liked 775 Times in 407 Posts
    Quote Originally Posted by jay View Post
    Everything I've seen has the Twins offense regressing, but that regression is towards the mean (aka... better). Regression describes returning towards the mean whether the starting point is bad or good. It doesn't describe going from bad to even worse, which is what I think you're saying?

    Help me understand which positions you think will put up worse production than last year? I'm not seeing many ways for the offense to be worse (barring significant injury).
    Yes, I've also noticed that people throw out the term "regression" without realizing how statistical regression differs from the common usage. And not just here, of course. Gauss is rolling over in his tomb...

  20. #95
    Senior Member Big-Leaguer jay's Avatar
    Posts
    884
    Like
    13
    Liked 50 Times in 31 Posts
    Blog Entries
    2
    Quote Originally Posted by crarko View Post
    Yes, I've also noticed that people throw out the term "regression" without realizing how statistical regression differs from the common usage. And not just here, of course. Gauss is rolling over in his tomb...
    Quite tricky definitions... regress is to go backwards while regression is to go back to a previous state/place (the mean).

    savvyspy, wasn't trying to jump on you. Your usage of 'regresses' was correct as well, just slightly confusing given the discussion on regression to the mean.

  21. #96
    Twins News Team All-Star TheLeviathan's Avatar
    Posts
    4,823
    Like
    176
    Liked 662 Times in 374 Posts
    Quote Originally Posted by jay View Post
    Help me understand which positions you think will put up worse production than last year? I'm not seeing many ways for the offense to be worse (barring significant injury).
    I think the combination we had at C/1B/DH will be worse this year than it was last year. People cite Mauer's injury - but forget Pinto hit out of his damn mind when he took over. Arguably better than Mauer in fact. I see a dip there.

    I see a very solid chance of a dip from Dozier.

    I see virtually no depth on the team in case of injury.

    And I see a team with very limited speed and very limited power. Not generally a good recipe. I don't necessarily think they regress....but I don't see them as being a whole hell of a lot better either.

  22. This user likes TheLeviathan's post and wants to buy him/her a steak dinner:

    notoriousgod71 (01-29-2014)

  23. #97
    Senior Member Big-Leaguer
    Posts
    993
    Like
    195
    Liked 221 Times in 96 Posts
    Quote Originally Posted by TheLeviathan View Post
    I think the combination we had at C/1B/DH will be worse this year than it was last year. People cite Mauer's injury - but forget Pinto hit out of his damn mind when he took over. Arguably better than Mauer in fact. I see a dip there.

    I see a very solid chance of a dip from Dozier.

    I see virtually no depth on the team in case of injury.

    And I see a team with very limited speed and very limited power. Not generally a good recipe. I don't necessarily think they regress....but I don't see them as being a whole hell of a lot better either.
    We're also forgetting that as mediocre as Morneau was he was STILL our second best hitter last year.

  24. #98
    Senior Member All-Star
    Posts
    1,416
    Like
    64
    Liked 40 Times in 28 Posts
    I appreciate the good replies to my previous comments. The civility displayed was appreciated. I like it when we can all act like adults and bring our differing viewpoints to the table without getting into shouting matches. I will try and touch on many of the points you made but don't want to be as long winded this time.

    First let me try and clarify my position a bit, again remember that I'm not trying to pick on any person in particular and this isn't really limited to just this thread.

    I think what really perturbs me is when a poster enters a thread that is clearly about a mathematical model or metric and makes an argument along the lines of "Well my gut says this can't be true therefore this is a terrible and worthless model and deserves no consideration." Ok, perhaps a bit of hyperbole in that last sentence. I'm not sure how that helps to further the discussion.

    I have no problem with people pointing out flaws in the methodology like Psuedo did earlier. He is absolutely right that often times there are some biases involved. Though sometimes those are base off of sound statistical analysis. For instance it is true, as a whole, that players tend to regress beginning in their late 20's or early 30's. Now, that isn't player specific, it would be great if we could narrow down the causes and be more selective in the application of regression, but statistical analysis is clearly not at that point yet. Expecting it to be is being unrealistic.

    I also fully accept and understand that not all people are interested in statistical analysis either because they don't have the time or the desire to dive into it. That's just fine. When my engineering friends start talking about how the Boeing Dreamliner decided to use titanium alloy A instead of alloy B my eyes start to glaze over. BTW JB, totally agree on Parker's analysis. On the other hand just because you don't understand the analysis is not a good reason to claim that a statistical model is wrong or not worth using. If you don't understand it then ask a question to clarify. Why does this model predict such a big drop in OPS for Mauer? It turns out that there are reasons to question his ability to maintain his great hitting. He is striking out much more frequently than earlier in his career. The more you strike out the lower your batting average and OBP. On a side note I highly recommend you look into ERA+ (or ERA- if you prefer fangraphs) Hosken. It essentially balances ERA for ballpark and league. After all a 4.00 ERA in Coors Field isn't the same as a 4.00 ERA at Target field and when you're facing a pitcher every 9th batter that's very different than facing David Ortiz. ERA+ adjusts for these things to make it easier to compare players.

    I recognize that not all statistical analysis is particularly well done. And if there are better alternatives out there it would be great if you would pass those along. For instance, bringing this discussion back to the topic at hand, you know of an alternative model or human prognosticator that is correct more than 70% of the time than great. We could all agree than that this model needs more work to be up to the highest standards. Again though, just saying being correct 70% of the time is terrible therefore this is useless doesn't really add much to the conversation. If the alternatives are your gut, which we have no idea how accurate it is, or another model that is only correct 60% of the time, well then 70% is looking pretty darn good.

    Psuedo, when it comes to determining the accuracy of models that predict future results it is pretty easy to determine whether their biases are accurate or not. After the season you just have to go back and see how often they were correct. That's the beauty of quantitative analysis. As I said earlier if you think STEAMER, ZIPS, PECOTA, etc... is a better choice, great. Let us know why that is better (or why you think you're a better prognosticator ).

    Statistical models or metrics take A LOT of time to create. Somebody has taken their off work hours and worked their butts off trying to create a tool of value. So when I see people glance at one aspect of the analysis and determine it doesn't fit their world view or gut feeling and then dismiss the whole project as being worthless it frustrates me.

    Perhaps we should have a thread dedicated to predicting the standings at the end of the season. Then we can compare our intuition to the statistical analysis. Perhaps the mods could even make a special title for a person that beats all the rest of us and the statistical models. They could be called The Prognosticator King. We could do something similar for OPS or even BA/OBP/SLG, though that seems harder to judge after the season.

    Well I've created another opus. Truly I need to learn how to be more concise.

  25. #99
    Senior Member MVP
    Posts
    5,713
    Like
    1,159
    Liked 564 Times in 369 Posts
    Think about this......MORE money and effort is applied to predict the stock market, and they are wrong a lot more often than 30% of the time.....70% accurate, given the variables. that's actually quite good imo.
    Lighten up Francis....

  26. #100
    Senior Member Big-Leaguer jay's Avatar
    Posts
    884
    Like
    13
    Liked 50 Times in 31 Posts
    Blog Entries
    2
    Quote Originally Posted by TheLeviathan View Post
    I think the combination we had at C/1B/DH will be worse this year than it was last year. People cite Mauer's injury - but forget Pinto hit out of his damn mind when he took over. Arguably better than Mauer in fact. I see a dip there.

    I see a very solid chance of a dip from Dozier.

    I see virtually no depth on the team in case of injury.
    Levi, I'll admit I was surprised when I looked at the numbers. We got a combined OPS of like .739 out of C/1B/DH last year. I plugged in some fairly conservative production and playing time projections for this year and ended up with .737 combined. There's some upside in that number, but I thought it would be higher before I started.

    On Dozier, if you mean a dip from his post-May numbers (like .790), I'd buy in more. His April-May was so bad though, he was only at .726 for the season. I can't see him much lower than that.

    If we hold steady at C, 1B, DH, 2B (and 3B?) that still leaves the inevitable improvement from the outfield production that was so atrocious. I do agree that a couple key injuries could ruin any chances of improvement in a hurry, but that's the case across most of MLB.

Page 5 of 8 FirstFirst ... 34567 ... LastLast

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.