Something that led to our demise

[Riverbrian]

10 hours ago, TheLeviathan said:

Well, the beefy data set is the platoon split advantage.  Given that well-established evidence, you're going to need a lot of contextual data to dispute it.  

I'm not saying you're wrong, there may even be really strong evidence that some teams are executing the strategy far better than we are too.  I'd be curious to see how it plays out, I just think there is a lot of angles you'll have to compile to get there given your inability (darn space time continum) to run a counterfactual.

How about this....we only play switch hitters.  Problem solved!

I recognize the beefy data set as real... However that beefy data set occurred over time.

Time is what is being stolen from the developing. We don't give them time. We throw them directly into this beefy data set grinder.  

Therefore is it hurting more than its helping?  

The question then becomes: What are you gaining on a per AB or per game basis to justify the time theft?

If we are to believe AI. The Per AB and Per Game difference between a .600 OPS and a .700 OPS is: .03 runs per AB - .13 per game. That'll add up to about 1 run per 7 games. That's the gain. Then consider that the .600 and .700 that I am using for this example is not static it is extremely fluid. 

It takes time to reach the numbers that we hang players with and time is what we take away and we are taking that time away fully completely.  

What do you lose by stealing time from the developing. What is the cost? 

Trade Value because you capped it. You have made them unplayable against one side. 

10 million a year for Margot or Grichuk types to play 28% of the time just to shield them. Or worse yet... the sacrificing of right handed prospects like Martin or Gonzales because someone has to do the shielding. 

Plus you have to roster lesser players. When the club calls up Gonzales instead of Erod. It is calling up lesser players just to feed this beefy data set grinder. Any system that requires the selection of lesser players and can justify the selection of lesser players deserves our suspicion as inherently wrong. 

If we have 10 players to jam into the opposite betters box against left handed starters and we do. We should be more equipped to face a left handed starter. The beefy data set says so. But yet my data is showing that it is not.  

The data that @MN Brian provided from 2016 to 2025 at first blush seems to indicate the same as the 2026 data that I compiled by hand. 

So... the numbers are not matching. The obvious reason why would be talent.   

Here we are calling up Ggon because he is right handed. Ggon wasn't chosen because he is more talented than Erod. We have a beefy data set in St. Paul that strongly suggests that he is not more talented than Erod.

Yet it is because we can't let Larnach face left handed pitching and that train pulled out of the station 4 years ago.

If we continue doing this and we seem to be continuing. 4 years from now. Jenkins will need a Grichuk. Erod will need a Grichuk and then you get to the point where you are capping the number of left handed hitters that you can roster and the percentage of arms is still 72% to 28%. It seems to me that the correct way to play the beefy data set would be more players who can attack the 72% but very few consider the disparity betwen 72% and 28%. The 28% alone will mitagate the .600 OPS that with time could become .700 OPS.      

It's possible that our methods are doing more long term damage than the short term benefit gained. 

[ashbury]

17 hours ago, Riverbrian said:

Enjoy yourself. Yosemite is one of the most beautiful spots in the U.S. You should focus on that.

Okay, since you insist.

 

[Riverbrian]

7 minutes ago, ashbury said:

Okay, since you insist.

 

I don't see a hotel or convenience store anywhere.