63 replies to this topic

[dude]

Ok so dollars per WAR is recalculated at the end of every year so I bet that 9.1 million per WAR is not going to stand up by the end of the season. In fact, the actual dollars per WAR they're using on their site is 7.8 for 2020 and 8 for 2021. And I suspect after the revenue issues of 2019 and depressed FA market that number will drop further.

 

You said earlier WAR is a backwards looking stat.  Hey, we can agree on something.  That's not how most actually use it though.  They are using it to 'project' a number of things, including contract projections.

 

Why exactly would we need to write article after article describing what already happened?  Again, my issue is largely with the blind presentation of information in a way which leads - just about everyone - to do horribly misleading "analysis".  That's the exact language I started this with when you said "[these pitchers] are worth 60-80M" .

 

When Chris tells me in the other thread that he thinks Mancini is getting 4-5 years and 85-100M at the AS break, how is he not trying to using "$/WAR" as the projection tool.  ***Chris, you certainly speak for yourself, but that's why you're projecting that, right?  He's a 3WAR player, regress for age and a hometown discount and that's basically the contract you are suggesting.

 

...and wrt to that, the Orioles can do anything they want. I don't think they will and time will tell, but if they do, good for Trey.  (and Trey, if they give you something like that to put in my face, I'll take a million and you're welcome) 

[dude]

Also, at least for ZIPS, it's a median projection. However, for a player like Simien, who is in his 30s, the 10th percentile projection is going to be pretty close to 0, while the 90th percentile projection is certainly not going to be 8 WAR. So the mean outcome is almost certainly going to be lower than the median. You have to discount the WAR projection to account for this.

 

....so here's what you just said.  "My range of potential answers encompasses the entire set of possible answers so whatever the answer is, my projection is right."  I asked which projection system you wanted to use, you can pick anything you want.

 

Here's another answer, it's a horribly misleading presentation of flawed information.

 

The Trade Simulator site uses a bunch of things that are all generally wrong.  They are a person's opinion of things they don't actually know and apply them in often incorrect ways.  When a trade happens, they go back in and 'update their proprietary formula' so the trade values match. Ugh.  It's actually a fun tool.  Generally wrong, but fun.

 

The next part of this is the notion of "value" and "worth".  

[BSLSteveBirrer]

ANY stat or metric you want to use is a forecasting tool based on back looking data. Period.

[hallas]

....so here's what you just said.  "My range of potential answers encompasses the entire set of possible answers so whatever the answer is, my projection is right."  I asked which projection system you wanted to use, you can pick anything you want.

 

Here's another answer, it's a horribly misleading presentation of flawed information.

 

The Trade Simulator site uses a bunch of things that are all generally wrong.  They are a person's opinion of things they don't actually know and apply them in often incorrect ways.  When a trade happens, they go back in and 'update their proprietary formula' so the trade values match. Ugh.  It's actually a fun tool.  Generally wrong, but fun.

 

The next part of this is the notion of "value" and "worth".  

 

I think it's a bit much to say that every stochastic projection is automatically invalid because they provide a range of outcomes.  You can look at a larger sample size to see if you were in the ballpark more often than not.

 

Secondly, I think it's fair to acknowledge that there are a lot of moving parts here, maybe too many to have a super accurate projection for both salary and performance, but they're meant to be more of a ballpark anyway.  It's worth noting that all the contracts were wrong in the same direction.  This tells me that the contract projections based on ZIPS/WAR/value are at least semi accurate relative to each other, but that there may have been other systemic forces suppressing salaries.  It's not really hard to figure out what that might have been in the 2020/2021 offseason.