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

[LoyalBirdFan]

Who gives a crap about projections? This isn't a first for them when it comes to the Orioles, they spend too much time looking at numbers. I suggest they try watching some games this year.

MLB says the Orioles have to play ALL 162 games next year even though the sabermetric nerds have decided how many games they will win.

We should all know by now...................... Matt gone, Machado gone, Hardy power drought, Davis horrible, etc... and this team wins 96 games.

Wonder what Buck's WAR is? Oh NO, Buck has NO stats, what am I gonna do? :rofl:

Let's do something productive and laugh at the 'optional word goes here' who cause us to has threads like this.

[CA-ORIOLE]

I agree it is not intentionally biased. However, there is no sanity check.

Tillman: 2.93, 3.73, 3.34.....4.35?

Gonzo: 3.25, 3.78, 3.23......4.56?

Those are particularly egregious to me, and suggest that this system blindly follows peripherals against all contrary evidence.

Yes, some guys can out perform their FIP (Gonzo certainly seems to be out there...right now) but lets take a look at Tillman. For his career he has a 4.51 FIP and 4.34 xFIP as compared to a 3.71 ERA. Last year was a particularly interesting case, His fWAR (based on a 4.01 FIP) and his rWAR (based on a 3.34 ERA (adjusted)) were exactly the same at 2.4 WAR. So even with a 0.67 differential between ERA and FIP, the WAR values were exactly the same. Chen came out considerably better last year on FIP based WAR.

[awannabegeek]

I agree it is not intentionally biased. However, there is no sanity check.

Tillman: 2.93, 3.73, 3.34.....4.35?

Gonzo: 3.25, 3.78, 3.23......4.56?

Those are particularly egregious to me, and suggest that this system blindly follows peripherals against all contrary evidence.

You don't change the model because you have outliers. The question becomes why the model is off on the Orioles so much? Is it something the Orioles do or are they just an anomaly? That should be the debate.

As a fan, that loves the statistical side of baseball and absolutely loves Fangraphs, I just don't get why their projections get under people's (not you) skin here so much. Enjoy the game...

[awannabegeek]

Who gives a crap about projections? This isn't a first for them when it comes to the Orioles, they spend too much time looking at numbers. I suggest they try watching some games this year.

MLB says the Orioles have to play ALL 162 games next year even though the sabermetric nerds have decided how many games they will win.

We should all know by now...................... Matt gone, Machado gone, Hardy power drought, Davis horrible, etc... and this team wins 96 games.

Wonder what Buck's WAR is? Oh NO, Buck has NO stats, what am I gonna do? :rofl:

Let's do something productive and laugh at the 'optional word goes here' who cause us to has threads like this.

Because the "eye test" is infalliable and not at all susceptible to bias or error. Never.

Also you do realize they had Hardy hitting for power, Matt, Machado, and Chris Davis playing like an MVP and didn't make the playoffs the year before?

With that said, I love this team and the projections don't take away from my enjoyment not one bit.

[byrdz]

Fangraphs must really love the Orioles if they're predicting last place.

[DrungoHazewood]

Any 2015 projection done now should be analyzed for an agenda. Otherwise, what is he doing with unsigned guys like Scherzer.

Or you could treat it as Rev 0.05. It's an alpha release. As long as the authors are also treating it that way I'm good. If they think it has any real predictive value they're a little nuts. Heck, if you projected all teams to .500 and then applied a standard random number to each team based on average variation of record due to luck you'd get about the same spread in wins as they're showing. This projection really could be derived from [71 wins + (RANDOM * 20)].

[DrungoHazewood]

Exactly! That is the whole point. The projections intentionally are based on incomplete information, totally overlooking defense, and are quite obviously worthless from the get-go. They are projecting FIP, yet claim that it is a projection of ERA. Absurd on its face.

What I don't understand is why they don't account for defense, or don't appear to. It's perfectly legitimate to use FIP as your baseline, as long as you understand that won't be accurate in a number of team cases (specifically teams with very good or very poor defense). And you need to adjust runs allowed from the FIP baseline.

[Ohfan67]

Fangraphs goal is to drive traffic to their website, not accurately predict the outcome of the MLB season (especially in December).

[Frobby]

Tillman has a 4.30 FIP last 3 years, and Gonzo has a 4.6 FIP last 3 years. Thats where they are getting their numbers from, they just based the ERA on the FIP since their metrics put more weight into FIP than ERA.

I realize that, but at some point you have to confront the reality that these two pitchers outperform their FIP every single year, and factor that in to your projections if you want the projections to be as good as they can be. I don't think the creators of Steamer are inclined to make individual exceptions to their methodology. That's fine, and I don't necessarily blame them, but in these two cases the chances are very good that their projections will be far off base.

[Roll Tide]

I don't know about last ...but with the current team I don't have them winning the division.

[Frobby]

I think it's interesting to see the PA allocations that Fangraphs is using to come up with its projections:

Wieters 60% of the games at C, Joseph 35%.

Davis 55% of the games at 1B, Walker 25%, Pearce 15%

Schoop 75% of the games at 2B, Flaherty 15%

Machado 85% of the games at 3B, Flaherty 10%

Hardy 90% of the games at SS

De Aza 70% of the games in LF, Lough 20%

Jones 95% of the games in CF

Pearce 50% of the games in RF, Lough 35%

Davis and Walker each 30% of the games at DH, Pearce 20%, Wieters 15%

They seem to like Walker.

[weams]

I realize that, but at some point you have to confront the reality that these two pitchers outperform their FIP every single year, and factor that in to your projections if you want the projections to be as good as they can be. I don't think the creators of Steamer are inclined to make individual exceptions to their methodology. That's fine, and I don't necessarily blame them, but in these two cases the chances are very good that their projections will be far off base.

Very good points.

[weams]

They seem to like Walker.

I think we should like Walker.

[waroriole]

I think we should like Walker.

He could be a cheaper version of Delmon, with a lot more upside, on this year's roster.

[DrungoHazewood]

I realize that, but at some point you have to confront the reality that these two pitchers outperform their FIP every single year, and factor that in to your projections if you want the projections to be as good as they can be. I don't think the creators of Steamer are inclined to make individual exceptions to their methodology. That's fine, and I don't necessarily blame them, but in these two cases the chances are very good that their projections will be far off base.

Again, I have no problem with FIP-based forecasts for pitchers. But there has to be an additional step that accounts for the difference between FIP and RA based on defense and other factors. I don't think it's unreasonable to think Gonzalez might have a 4.60 ERA if pitching for some generic team with an average defense. But this projection isn't that, it's him in the context of the Orioles.