FWIW, Fangraphs Projects O's in Last Place in 2015

[MurphDogg]

Their projection systems always expect pitchers to regress back to their FIP, ignoring that they pitch well because of our defense.

Yeah, and I don't understand it. Why can't the defensive statistics, flawed as they may be, be integrated into their projection system?

[AlbionHero]

Yeah, and I don't understand it. Why can't the defensive statistics, flawed as they may be, be integrated into their projection system?

Thats why pitcher sabermetrics are still a work in progress. For hitters, you don't really need to depend on others to play your game, so you can just look at what that one player does and see his value. But for a pitcher, you need to depend on 8 other players too, and the systems still haven't figured out a way to accurately project how a pitcher will do with defense backing them up... so all they can really do is look at things they can control, like strike outs, walks, home runs etc. It makes it tough to really get a good idea about how they'll pitch if you don't take into account defense.

[Phantom]

Yeah, and I don't understand it. Why can't the defensive statistics, flawed as they may be, be integrated into their projection system?

Quite simply because ignoring them leads to more accurate results. It's unclear if the stats are just bad, or more likely, defense and luck tend to vary a lot year to year, so stripping them out of predictions is a good idea.

[Guest]

Didn't this same bull**** outfit project us for 78 wins last year?? And they were consistently wrong on free agent projections.

I think so.

[Phantom]

Didn't this same bull**** outfit project us for 78 wins last year?? And they were consistently wrong on free agent projections.

You're right. We should go with the projection systems that are 100% accurate.

[Number5]

LOL. So what else is new? IIRC they had us last last year, as well. Pretty much every year, in fact. Last year, I believe they pretty much had their projected standings end up being the inverse of the actual AL East results. Anyone have that handy? It is hard to be that consistently poor in projecting. I will begin to worry if the ever have us projected to do well. :laughlol:

[MurphDogg]

Quite simply because ignoring them leads to more accurate results. It's unclear if the stats are just bad, or more likely, defense and luck tend to vary a lot year to year, so stripping them out of predictions is a good idea.

I don't think that is true. While there is more statistical noise in defensive statistics than there is in offensive statistics, it doesn't mean they should be thrown out entirely.

Projecting every team to win 81 games and lose 81 games would also be more accurate, but I don't see anyone pining for that.

[waroriole]

I tweeted at Dave a couple weeks ago when they first posted it.

They have the O's ERA increasing by I think it was .7 runs.

They seemingly refuse to adjust for defense. They are irrecoverably wedded to FIP.

That's why he's a joke. It must be some outlier, huh Dave.

[LoyalBirdFan]

Why does anyone care what they think?

[Dr. Unk]

Weren't we supposed to have a losing record in 2014 also?

[Can_of_corn]

That's why he's a joke. It must be some outlier, huh Dave.

It isn't as if Dave is the only person involved.

He is just the unblinking face of the web site.

[Spy Fox]

You can just throw this away. The top team in all of MLB is projected to win 89 games?

Can someone tell me the last time in a 162 games season where the #1 team in baseball didn't have 90 wins?

This is the difference between a prediction and a projection. A prediction is what you actually think will happen in one run-through of the season. A projection is an average of what a computer thinks will happen over a countless number of run-throughs of a season. So, in mathematical projections you will almost never see a team with more than about 90 wins. Because a lot needs to go right for a team to be that good, so even a top-tier roster will only "average" about 90 wins when you run enough possible trials. So the lack of a dominant team doesn't devalue the projection.

But still, 79 wins is way too low for the O's and I fully expect to outperform these projections for the fourth year in a row!

[LoyalBirdFan]

Didn't this same bull**** outfit project us for 78 wins last year?? And they were consistently wrong on free agent projections.

They don't actually watch any games, they stare at numbers all day/night, bunch of geeks. They are so hosed up..... don't get me started about those lost souls.

[Phantom]

Projecting every team to win 81 games and lose 81 games would also be more accurate, but I don't see anyone pining for that.

This is just wrong. Everything about it is wrong. I am having trouble expressing how wrong it is. Really, really, wrong.

Fangraphs and the other advanced stats projections systems are a lot more accurate than predicting every team at .500. They are also more accurate than any other prediction system. They are quite broad, and no one should be upset if it has your team improving or regressing. They are not making some value judgement. They are saying that by the numbers we can collect, history suggests that you will finish ____.

[AlbionHero]

LOL. So what else is new? IIRC they had us last last year, as well. Pretty much every year, in fact. Last year, I believe they pretty much had their projected standings end up being the inverse of the actual AL East results. Anyone have that handy? It is hard to be that consistently poor in projecting. I will begin to worry if the ever have us projected to do well. :laughlol:

This is the problem with Defense teams. Royals are also projected to be a .500 team since they take the defense out of the picture so it doesn't help their pitcher projections. These team projections are pretty worthless for teams with a lot of focus on defense.