Fangraphs article on why their projections hate us

[Oriole1940]

These examples of Fangraph go to show how poor they are at their job, and not worthy to be read, or paid any attention to. The best treatment for people like that is to ignore them and read articles that are worth the time, unless a person enjoys wasting their life away on something like that.  

[nvpacchi]

Fangraphs is becoming a relic of a previous era of the internet. The hey day of guys like Dave Cameron, Carson Cistulli, and Jeff Sullivan is long gone.  The site has gone downhill since it axed NotGraphs, at least then Fangraphs had some fun to balance out the heavy saber tone of guys like Cameron.

Honestly the only thing that makes the site useful is its stat pages,  which tend to be my goto over bref, especially for the game log features.

None of the current crop of writers, maybe besides Danny Szym, are must reads, which isnt surprising now considering how easy it is to open a substack and do your own baseball analysis. Twitter and substack in some tandem can get you better analysis than cherry picked content from Fangraphs that is at the mercy of an editing process and maybe some shareholder initiatives.

This whole power rankings debacle, independent of the quality of the ranking, embodies the downhill spiral of a site that just feels like its going through the motions. I won't cry when the inevitable day comes and Fangraphs can't pay for server and cloud space and writers thus becoming purely a direct competitor to bbref.

[DrungoHazewood]

14 hours ago, NCRaven said:

Fangraphs playoff odds were updated throughout the season, including as recently as this morning.  They are based on the team and individual player performances from what I understand.  They use the same methodology for every team in baseball during the season and every team still alive in the post season.  I think that one of the reasons that they are missing on the O's this year is that our young players lack a long performance history and their system is slow to adjust to more recent performance vs. past performance.  It's also true that teams with our run differential do not historically do as well as this year's Orioles team has done.

And they do things like assume that Ryan O' Hearn's last five years of performance are meaningful, and that he's likely not a .850 OPS guy going forward. Orioles fans look at O'Hearn playing awesome for Orioles, and pretty much throw away whatever he did with the Royals. Fangraphs almost certainly weights more recent performances more, but they don't throw away his last four years of a .633 OPS.

If you're a fan and you want to assume the 29-year-old O'Hearn had a transfiguration and you want to throw away everything prior to 1 April 2023 in your personal projections, good on you. If a widely read site with a projection system did that I'd assume they're not serious.

Similarly we see Tyler Wells as a potential playoff relief beast because he's 6' 8" 260 and looked real good the last week of the season. Fangraphs sees a guy with a 3.64 ERA and a 4.98 FIP, and a last two years of a 4.20 ERA/4.20 FIP. Dean Kramer out-performed his FIP by .4 runs/game. I don't know how they treat Grayson Rodriguez, but I doubt they're just assuming he's a postseason ace based on the information available. They see him as a 3.93 FIP.

When the Orioles have an unexpectedly good season part of that is players out-performing reasonable expectations. You can't expect a projection system to throw out reasonable expectations sometime in the middle of the year and say "Screw it! We're on the Orioles bandwagon now! They're winning it all baby!" You make the best model you can, and whatever happens happens. Every year a fanbase or three will have a 338-page thread about how you hate them because you don't have them as odds-on favorites. That's just part of the deal.

[DrungoHazewood]

14 hours ago, Oriole1940 said:

These examples of Fangraph go to show how poor they are at their job, and not worthy to be read, or paid any attention to. The best treatment for people like that is to ignore them and read articles that are worth the time, unless a person enjoys wasting their life away on something like that.  

I'm guessing that in 1968 when the Sporting News pre-season edition said the Orioles were going to finish 4th you wrote a letter to the editor telling them that their projection wasn't worth the paper it was printed on and you don't understand why anyone would waste their time with such a BS publication.

[Sports Guy]

15 hours ago, dystopia said:

Why does that matter? FG is looking at a 100 team and saying “we don’t believe you’re anywhere close to being that good”. Doesn’t matter if it’s personal or not. It would be one thing if it could be statistically defensible but it’s not. The Orioles and their fans have every right to feel disrespected by that, and there’s 29 other teams and fan bases who would feel the same way. If you think otherwise, then you’re either lying or not a real fan. 

Part of why I dislike some of these projection systems is that they run simulations and they put a lot of emphasis in games that are never played. It’s all made up.

That being said, in any given season, a team’s success is random, which is the point about the. You could play this season again and the Os win 80 games. I personally don’t believe that this is a 100 win team in terms of talent but I believe a lot of things came together at the right time and you won 100 games.

FG is essentially saying that. Now, with that said, I think these fangraphs playoff projections are complete bs. To say the Dbacks and Twins have better odds to win the WS than the Os is one of the dumbest things I have ever heard, especially Arizona.

They have the Astros with the second best odds yet the Twins, who play them next, have better odds than the Os?  Logically that doesn’t make sense.

 

[Frobby]

1 minute ago, Sports Guy said:

 

They have the Astros with the second best odds yet the Twins, who play them next, have better odds than the Os?  Logically that doesn’t make sense.

 

This is my biggest point.  Especially considering how highly they rate the Astros’ chances.  

[DrungoHazewood]

To summarize this thread:

Fangraphs says there's a 94.5% chance the O's don't win the World Series, and a 93.9% chance the D'backs don't win it, and that's TOTAL FRIGGIN' BS BECAUSE EVERYONE KNOWS THE ORIOLES REALLY HAVE A 91.8% SHOT OF GOING HOME DISAPPOINTED!!!!!  IDIOTS!!11!

[dystopia]

8 minutes ago, Sports Guy said:

Part of why I dislike some of these projection systems is that they run simulations and they put a lot of emphasis in games that are never played. It’s all made up.

That being said, in any given season, a team’s success is random, which is the point about the. You could play this season again and the Os win 80 games. I personally don’t believe that this is a 100 win team in terms of talent but I believe a lot of things came together at the right time and you won 100 games.

FG is essentially saying that. Now, with that said, I think these fangraphs playoff projections are complete bs. To say the Dbacks and Twins have better odds to win the WS than the Os is one of the dumbest things I have ever heard, especially Arizona.

They have the Astros with the second best odds yet the Twins, who play them next, have better odds than the Os?  Logically that doesn’t make sense.

 

Okay, let’s say they are a in reality a 94 win team. That was their Pythagorean record. The question is still how Arizona has better odds than them, or Minnesota, or Philadelphia. All with worse records and worse Pythagorean records. And none of those teams played in as tough of a division as the O’s. 

[dystopia]

2 minutes ago, dystopia said:

Okay, let’s say they are a in reality a 94 win team. That was their Pythagorean record. The question is still how Arizona has better odds than them, or Minnesota, or Philadelphia. All with worse records and worse Pythagorean records. And none of those teams played in as tough of a division as the O’s. 

Ok to be fair, you did basically agree with this, but this is essentially why I’m calling BS on FG and accusing them of bias. If they’re not biased and this is just outright incompetence, then they should just shut the site down. 

[Sports Guy]

15 minutes ago, dystopia said:

Ok to be fair, you did basically agree with this, but this is essentially why I’m calling BS on FG and accusing them of bias. If they’re not biased and this is just outright incompetence, then they should just shut the site down. 

It’s not bias. Why people think this is some personal hate towards the Os is beyond me. The Baltimore Inferiority complex is a real thing.

It’s a system. It’s just math and projections.

You can say the system is flawed and that’s fine but it’s not biased.

[dystopia]

21 minutes ago, DrungoHazewood said:

To summarize this thread:

Fangraphs says there's a 94.5% chance the O's don't win the World Series, and a 93.9% chance the D'backs don't win it, and that's TOTAL FRIGGIN' BS BECAUSE EVERYONE KNOWS THE ORIOLES REALLY HAVE A 91.8% SHOT OF GOING HOME DISAPPOINTED!!!!!  IDIOTS!!11!

The percentage isn’t the issue. The ranking is the issue. Which teams are being given better odds than us is the issue. 

[dystopia]

3 minutes ago, Sports Guy said:

It’s not bias. Why people think this is some personal hate towards the Os is beyond me. The Baltimore Inferiority complex is a real thing.

It’s a system. It’s just math and projections.

You can say the system is flawed and that’s fine but it’s not biased.

Because the evidence available points to bias. You have a system giving a 100 win team with a greater than +100 run differential the worst odds to win the WS with no coherent reasoning behind it. That’s bias, whether it’s personal or not (and I never claimed FG had anything “personal” against the O’s). 
 

They created a system that was inherently biased, whether intentional or not. 
 

Now, maybe I wouldn’t be beating this drum so much if they hadn’t been doing the same thing between 2012 and 2016 too. 

[Can_of_corn]

1 minute ago, dystopia said:

Because the evidence available points to bias. You have a system giving a 100 win team with a greater than +100 run differential the worst odds to win the WS with no coherent reasoning behind it. That’s bias, whether it’s personal or not (and I never claimed FG had anything “personal” against the O’s). 
 

They created a system that was inherently biased, whether intentional or not. 
 

Now, maybe I wouldn’t be beating this drum so much if they hadn’t been doing the same thing between 2012 and 2016 too. 

And 2018?  Were they biased in favor of the O's?

[DrungoHazewood]

8 minutes ago, dystopia said:

The percentage isn’t the issue. The ranking is the issue. Which teams are being given better odds than us is the issue. 

I mean, they could just use metadata and/or your IP address to geolocate you and redirect to a page that bumps the team in that area up 4-5 spots in the rankings. Certainly would make them more popular, if also ethically suspect.

[DrungoHazewood]

1 minute ago, Can_of_corn said:

And 2018?  Were they biased in favor of the O's?

You never know the cryptic motivations of baseball's deep state.