Jump to content

Providing independent coverage of the Minnesota Twins.

MinnCentric Forums


Subscribe to Twins Daily Email

Photo

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

  • Please log in to reply
151 replies to this topic

#41 PseudoSABR

PseudoSABR

    Twins News Team

  • Twins News Team
  • 2,115 posts

Posted 27 January 2014 - 09:39 PM

To be fair, everybody kind of does. So while we make take issue with the stats of some individuals, the end results are probably not as far off as we'd like to think.

I get that. But really, if you go look at each player on the Twins roster, even for the atrocious pitchers last year, their projections drift towards the mean. Which is likely in the aggregate, but when I see it applied to every individual on an entire roster, well that looks formulaic if not a bit lazy.

I just question the actual use of such projections if they rely on what uberfans already acknowledge as common-placed assumptions.

Edited by PseudoSABR, 27 January 2014 - 09:54 PM.


#42 KirbyHawk75

KirbyHawk75

    Member

  • Members
  • 96 posts

Posted 27 January 2014 - 10:04 PM

Our pitching is better.. It can not be much worse. Our staff will not be word beaters, there are no Johan Santana's obviously. Hopefully we have a Brad Radke or two? Could our staff be like the one when we had Radke, Joe Mays, Eric Milton, etc.? That would be a good staff if we had a good offense. I just do not have faith in our offense unless some kids can produce.

#43 drivlikejehu

drivlikejehu

    Senior Member

  • Members
  • 885 posts

Posted 27 January 2014 - 10:21 PM

The whole point of computer projections is objectivity. They use historical precedent to establish a baseline of expectations. In any individual case, the model may fit poorly, but criticizing a projection on that basis makes no sense at all.

Regardless, there is no method of analysis that shows the Twins in a positive light for 2014. They aren't even close to an average club on paper. Pretty much everything has to go right for a run at .500, which certainly would be nice but isn't exactly a world-beating accomplishment.

#44 TheLeviathan

TheLeviathan

    Twins News Team

  • Twins News Team
  • 6,790 posts

Posted 27 January 2014 - 10:22 PM

I just question the actual use of such projections if they rely on what uberfans already acknowledge as common-placed assumptions.


Most future projections are full of easy things to criticize. Afterall, we're trying to guess the future.

To me, the use of things like this is to diffuse unrealistic pessimism or optimism. I imagine this won't be the last projection that sobers us even after a successful offseason. My guess for Vegas is our over/under will be 73.5.

#45 old nurse

old nurse

    Member

  • Members
  • 1,904 posts

Posted 27 January 2014 - 10:35 PM

Looking at the 2012 and 2013 projections for the teams versus what happened on the field would lead me to believe these projections should be taken for entertainment value only. The issue becomes what players will elevate their games. Will being away from Yankee stadium boost Hughes? Is Pelfrey capable of becoming an above average pitcher? Will the players that played a bit in their first season in the majors last year become above average players? Will a rookie surprise? The projection would say no because a computer has no way of knowing this. The teams that won far more than their projections would have had these things happen. If established players go to the mean of their careers and no one develops, it will be a long season.

#46 jokin

jokin

    Twins News Team

  • Twins News Team
  • 9,386 posts

Posted 27 January 2014 - 10:41 PM

The whole point of computer projections is objectivity. They use historical precedent to establish a baseline of expectations. In any individual case, the model may fit poorly, but criticizing a projection on that basis makes no sense at all.

Regardless, there is no method of analysis that shows the Twins in a positive light for 2014. They aren't even close to an average club on paper. Pretty much everything has to go right for a run at .500, which certainly would be nice but isn't exactly a world-beating accomplishment.







Looking at the 2012 and 2013 projections for the teams versus what happened on the field would lead me to believe these projections should be taken for entertainment value only. The issue becomes what players will elevate their games. Will being away from Yankee stadium boost Hughes? Is Pelfrey capable of becoming an above average pitcher? Will the players that played a bit in their first season in the majors last year become above average players? Will a rookie surprise? The projection would say no because a computer has no way of knowing this. The teams that won far more than their projections would have had these things happen. If established players go to the mean of their careers and no one develops, it will be a long season.


The projections think so highly of the Twins current projected opening day roster, that they have 4 guys slated for minor league duty, namely Pinto, Sano and Buxton as the second, third and fourth-best position players and Alex Meyer as the second best pitcher.

Edited by jokin, 27 January 2014 - 10:44 PM.


#47 Twins Fan From Afar

Twins Fan From Afar

    Senior Member

  • Members
  • 660 posts

Posted 27 January 2014 - 10:57 PM

Compared to last season, the Twins' roster definitely has improved. There's little dispute there. And I'll throw to the side the crazy Mauer prediction.

My question, though, is this: relative to the teams they play, have the Twins improved more, less, or roughly the same as have those teams?

If the answer is that the Twins generally kept pace with their competition (even though, to us, they look considerably improved), that's probably a large reason why they are expected to fare largely the same.

Just thinking aloud....
Andrew Walter
Twins Fan From Afar
Providing Twins and Rock Cats Coverage
http://twinsfanfroma...gspot.com<br />Follow me on Twitter: @MNfanfromafar

#48 jokin

jokin

    Twins News Team

  • Twins News Team
  • 9,386 posts

Posted 27 January 2014 - 11:19 PM

Compared to last season, the Twins' roster definitely has improved. There's little dispute there.


And I'll throw to the side the crazy Mauer prediction.

My question, though, is this: relative to the teams they play, have the Twins improved more, less, or roughly the same as have those teams?

If the answer is that the Twins generally kept pace with their competition (even though, to us, they look considerably improved), that's probably a large reason why they are expected to fare largely the same.

Just thinking aloud....


I think most fans think the rotation has considerably improved. By contrast, the Twins lost 2 major league hitters and they, themselves admit the offense is still full of holes.

#49 Twins Fan From Afar

Twins Fan From Afar

    Senior Member

  • Members
  • 660 posts

Posted 27 January 2014 - 11:25 PM

I think most fans think the rotation has considerably improved. By contrast, the Twins lost 2 major league hitters and they, themselves admit the offense is still full of holes.


For sure, the offense still has holes. No dispute there. If I had to bet on a "surprise player" or 2 coming from the minors and doing well with the big club, I'd bet on a hitter, not a pitcher. In other words, I guess I'm suggesting that hitting holes could be more easily remedied than pitching holes. We'll see.

That wasn't my larger point, though.
Andrew Walter
Twins Fan From Afar
Providing Twins and Rock Cats Coverage
http://twinsfanfroma...gspot.com<br />Follow me on Twitter: @MNfanfromafar

#50 Oxtung

Oxtung

    I don't skinny dip. I chunky dunk.

  • Members
  • 1,558 posts

Posted 27 January 2014 - 11:56 PM

Mauer's 2014.

.331 15 HR 85 RBI .420 OBP .851 OPS 152g


In the last five years only 11 of the thousands of players in baseball have had an OBP of >.420. Mauer's K% has been rising steadily for 4 years straight and even with an absurd BABIP of .383 last season he was unable to even approach .420. I'm curious what you think will be different this year that allows Joe to achieve such a difficult task?

#51 Oxtung

Oxtung

    I don't skinny dip. I chunky dunk.

  • Members
  • 1,558 posts

Posted 28 January 2014 - 12:07 AM

I get that. But really, if you go look at each player on the Twins roster, even for the atrocious pitchers last year, their projections drift towards the mean. Which is likely in the aggregate, but when I see it applied to every individual on an entire roster, well that looks formulaic if not a bit lazy.

I just question the actual use of such projections if they rely on what uberfans already acknowledge as common-placed assumptions.


Considering that their predictions are ultimately derived from an underlying equation does it surprise you that it looks formulaic?

Many people, and I'm not just talking to you here Pseudo, seem to be having problems with these predictions. IMO you are all looking at these too literally. Predictions like these aren't saying, "Hey this is exactly what is going to happen!" Rather they are saying, "Given the data we input, this has the highest probability of occurring." However, like any data set when your n=1, as is the case with Joe Mauer, their confidence level is going to be pretty darn small and their error bars are going to be very large. Now, when you apply statistical modeling to a team, and your accounting for not just 1 player but 25, now you have something that can be given a little bit more credibility.

#52 PseudoSABR

PseudoSABR

    Twins News Team

  • Twins News Team
  • 2,115 posts

Posted 28 January 2014 - 12:29 AM

While we know we're always working with a probabilistic model when dealing with projection, there's a lot more at work than a simple mathematical formula--and hopefully, if that were simply the case we'd see more sophistication and variation than such vanilla regression towards the mean. That a metric of pure data could deduce something like plate appearances based on such abstractions as injuries and minor league development and assumed playing time is damningly tenuous. There's biases (inevitably) at work both in terms of what data gets input-ed and what bases the mechanics of the formula are built. That we pretend that such prognosis are somehow mathematically innocent seems silly to me. The thing about good prognosis models is that they should seem organic in the macro.

I don't mind that projections are based on formula; but I want that formula to do better than some of the commonly held base assumptions (we already have that). Again, I'd like to see a projection formula that produced nuanced results player by player, and don't simply carry the assumptions of the aggregate to the particulars of the each specific player.

In my mind, there's difference between playing the odds and earnestly prognosticating specific player production. If current statistical modeling doesn't grant much confidence in specific player production (as with Mauer), then why make publicly available such prognosis? Again, what value do they have (as suggested with the population of 1(n=1))?

If the point is that such endeavors are better when the lens is pulled back, that the projections are more useful for projecting team wins than individual production, I think we can agree.

Edited by PseudoSABR, 28 January 2014 - 01:06 AM.


#53 Reider

Reider

    Senior Member

  • Members
  • 399 posts

Posted 28 January 2014 - 03:05 AM

I'm a bit concerned with the offense, but Joe Mauer is not one of my concerns. I can see another 90 loss season, but if all goes right, I can see the team taking a small step forward as well. I guess we'll just have to see how things play out. Either way, I don't expect a large step forward or backwards this year.

#54 Brock Beauchamp

Brock Beauchamp

    Owner

  • Administrators
  • 10,821 posts

Posted 28 January 2014 - 06:21 AM

In the last five years only 11 of the thousands of players in baseball have had an OBP of >.420. Mauer's K% has been rising steadily for 4 years straight and even with an absurd BABIP of .383 last season he was unable to even approach .420. I'm curious what you think will be different this year that allows Joe to achieve such a difficult task?


In the past five years, Mauer has posted an OBP of .444 and .416.

Is it likely he'll post .420 or better in 2014? No. Is it impossible? Definitely not.

I'm not even sure his increasing K rate is a concern right now, though that BABIP sure is.

#55 diehardtwinsfan

diehardtwinsfan

    Twins Moderator

  • Twins Mods
  • 6,021 posts

Posted 28 January 2014 - 07:16 AM

I definitely think there's good reason to question the projection. I tend to think this is a 70-80 win team right now, and if things break you may see that swing up if guys like Meyer and Sano hit the ground running when they get called up. That said, I think there's good reason to find faults with a few things. Mauer will regress at one point, but there's no real reason to think it will be this season, and going from an .880 OBP to a .750 OPS is quite odd. Likewise, Hughes and Nolasco should be pretty massive upgrades over the guys they trotted out last season and there's reason to think Pelfrey will move a bit closer to career norms as well. Willingham will likely bounce back too. There really wasn't anyone who had a season that screams "regress to the means" on the other side of it, so prediciting 90+ again seems odd, especially with reinforcements arriving.

The biggest difference I see is that this year's Twins team should be much more watchable.

#56 old nurse

old nurse

    Member

  • Members
  • 1,904 posts

Posted 28 January 2014 - 07:28 AM

Considering that their predictions are ultimately derived from an underlying equation does it surprise you that it looks formulaic?

Many people, and I'm not just talking to you here Pseudo, seem to be having problems with these predictions. IMO you are all looking at these too literally. Predictions like these aren't saying, "Hey this is exactly what is going to happen!" Rather they are saying, "Given the data we input, this has the highest probability of occurring." However, like any data set when your n=1, as is the case with Joe Mauer, their confidence level is going to be pretty darn small and their error bars are going to be very large. Now, when you apply statistical modeling to a team, and your accounting for not just 1 player but 25, now you have something that can be given a little bit more credibility.


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.

Edited by old nurse, 28 January 2014 - 08:13 AM.


#57 CRArko

CRArko

    Don't need no stinkin badges

  • Members
  • 2,741 posts
  • LocationNagoya
  • Twitter: crarko

Posted 28 January 2014 - 07:40 AM

They are a 90 loss team until they prove they're not.

The proof begins in a few weeks. Here's hoping all the lemmas work out.

#58 jay

jay

    Senior Member

  • Members
  • 1,429 posts

Posted 28 January 2014 - 08:52 AM

I pointed out a number of issues at the n=1 level with the models that FG is using (ZIPS and Steamer) in another thread. Mauer and Willingham don't get much love and our positions that were bad last year generally regress in a positive direction towards some semblance of reasonable production. These models, for as smart as they are, aren't smart enough to apply individualized real-world factors.

I'd peg us for 76-78 wins this year, personally.

#59 jay

jay

    Senior Member

  • Members
  • 1,429 posts

Posted 28 January 2014 - 09:06 AM

Win differential from 2013 actual to this 2014 projection...

Bottom 5:
1 (t) ATL, BOS, PIT -11
4 KC -9
5 (t) BAL, OAK -8

Top 5:
1 HOU +19
2 CWS +16
3 MIA +13
4 SEA +12
5 SF +9
7 (t) MIN, LAA +6

Notice any trends here? Teams that were bad or worse than expected last year get better (regress towards the mean). Teams that were good or better than expected last year get worse (regress towards the mean). This highlights what PseudoSABR was referring to. Not exactly an earth-shattering concept.

#60 jay

jay

    Senior Member

  • Members
  • 1,429 posts

Posted 28 January 2014 - 09:12 AM

One more illustration...

Teams with actual results in 2013 that were plus or minus 10 (or more) wins from his pre-season projections:
STL (+14), PIT (+13), CLE (+13), BOS (+12), OAK (+12), ATL (+11), BAL (+10), CHC (-10), TOR (-12), CWS (-13), LAA (-13), SFG (-16), HOU (-21).

Yes, you read that right. 13 teams were more than 10 wins off from their projection. Don't let this projection sap your hope for a... .500 season?