Colorado Rockies baseball - how bad will it get?

[Munch]

@John_DiFool

I am only interested in finding optimal ways of measuring talent

@John_DiFool

well you can’t get any more fundamental than wins

So a player who hits 50 home runs on a team that wins 0 games doesn’t have any talent? I don’t think you or James would suggest that. That’s why he weasels out of the question by saying that Win Shares measures value, and not talent.

I’m certainly willing to admit that a player who hits 50 HR batting in front of Aaron Judge is better than a player who hits 50 HR batting in front of TJ Friedl Jr. - I just don’t know how much better. That’s tough to do, and that’s why analysts (James included) have formulated seemingly countless measurements to do so (one should come up with a metric measuring analysts’ new metrics…). Lately I’ve enjoyed digging into xwOBA for a hitting metric - it seems to utilize the mountains of data that Statcast and other tracking systems are able to collect, and presents them in a pretty clear way to understand it.

[RickJay]

@Munch

It’s not a flaw, because WAR is meant as a measure of individual talent and contribution regardless of team context. 6.1 WAR player Hector Consuelo of the 20-92 Albany Crapholes is just as good as 6.1 WAR Johnny Bonneville of the 103-14 Metroville Cosmos.

Yes, it should be, as long as it’s reflective of how many games the team won. A player can absolutely put up a lot of WAR on a shitty team. Look at Cal Ripken in 1991. If he’s not on that team, they don’t win 67 games. They win more like 57. I guarantee you, would bet a lot of money, that he piled up a ton of Win Shares, too.

But if your claim is that the player on a team with 88 wins are collectively more valuable than the player on a team with 98 wins, well, sorry, but they can’t have been. That doesn’t make any sense at all. Your analysis might reasonable conclude that based on their individual stats, they would be expected to probably win more games going forward. I could totally believe that. But if in the past they did not win more games, then that’s an absolute fact that cannot be ignored. Actual value in the past is winning games.

The entire point to sabermetrics is establishing the relationship between what baseball players do on the field and winning games.

@Munch

So a player who hits 50 home runs on a team that wins 0 games doesn’t have any talent?

You are confusing talent and value. TALENT is the ability to apply skill to accomplish an objective. A player has TALENT before they even step on the field. 50 home runs is evidence of talent, but it is not itself talent, it is a product of talent.

VALUE is the outcome of the application of talent, measured against the accomplishment of objectives. In baseball, value is winning games.

A player who hits 50 home runs for a bad team will still cause his team to win more games than they otherwise would have, even if the team wins few games. A player who hits 50 home runs a year can be expected to continue in the future to help his team win games. No team wins zero games.

James’s point that WS measres value, not talent, isn’t a weasel at all, it’s an honest and totally correct statement, and incredibly important to understand.

@Munch

I’m certainly willing to admit that a player who hits 50 HR batting in front of Aaron Judge is better than a player who hits 50 HR batting in front of TJ Friedl Jr.

Absent some other evidence, no, he is not.

[Munch]

@RickJay

James’s point that WS measres value, not talent, isn’t a weasel at all, it’s an honest and totally correct statement, and incredibly important to understand.

I think you’re misunderstanding my point - in a discussion about talent, James came in and said the metrics (and MVP awards) should be about value, and he’s spent a good portion of his career trying to avoid criticism by living in that forced shift of the discussion. His base metric is the win - a stat that is inherently derived from a team context. WAR’s base metric is the run (despite its name), and focuses on how a player’s individual performance scores runs. I personally don’t really like Win Shares, but I’m not offering a criticism of it, let alone a dismissal of it.

@RickJay

Absent some other evidence, no, he is not.

We can look at the types of pitches players who bat in front of mediocre hitters see versus the types of pitches players who bat in front of elite hitters see, and we can measure the results of how those players perform. This is the essence of xwOBA. My example is obviously impossible to demonstrate, absent finding players traded either to or from a lineup that drastically changes their situation.

[RickJay]

@Munch

I think you’re misunderstanding my point - in a discussion about talent, James came in and said the metrics (and MVP awards) should be about value, and he’s spent a good portion of his career trying to avoid criticism by living in that forced shift of the discussion.

I really do not understand you here. James’s point here has always been about assessing a player’s value when looking at what he has actually done in the past, and he’s correct. That’s not a “forced shift of the discussion,” it’s literally him discussing a specific concept. How is he shifting anything?

@Munch

His base metric is the win - a stat that is inherently derived from a team context. WAR’s base metric is the run (despite its name), and focuses on how a player’s individual performance scores runs. I personally don’t really like Win Shares, but I’m not offering a criticism of it, let alone a dismissal of it.

James is right and WAR is wrong. WAR is disconnected from the very thing it purports to measure: the WINS that a player adds above replacement level.

[Munch]

@RickJay

James’s point here has always been about assessing a player’s value when looking at what he has actually done in the past, and he’s correct.

That’s incomplete - James’ Win Share only assesses a player’s value in games they’ve won. I am of the opinion that it’s silly to say that any accomplishment performed in a losing effort is unworthy of measurement or assessment. If you’re fine with that, so be it.

@RickJay

WAR is disconnected from the very thing it purports to measure: the WINS that a player adds above replacement level.

Why, because it is primarily derived from runs? You refer to Pythagorean records all the time - that’s derived directly from runs.

[TroutMan]

I think the concept of adding the WAR for individual players on a team is flawed and a misuse of the term. Some of those wins are for the same game.

Take an extreme example of nine players each with a WAR of 20. I don’t think even that team would win 180 games.

[RickJay]

@Munch

That’s incomplete - James’ Win Share only assesses a player’s value in games they’ve won.

Nope, wrong. That is absolutely not how it’s calculated at all, not in any way is that true, man. I can understand why you don’t get the purpose of it now.

James (granted, it takes dozens of pages) has the entire process in the book.

[justdoit]

Rockies imrproved in second half of season and are 41-109 with .273 winning percentage with about 12 games to go. Would still need to lose all remaining games to tie most losses in season	121 (2024 Chicago White sox). Very unlikey but they could come close with 115 losses or more. They will not reach lowest winning percantage in modern era of .235 (1916 Phil Athletcis) but I don’t think below .300 happens often. Will be a very bad season but not as awful as it was looking.

Rockies imrproved in second half of season and are 41-109 with .273 winning percentage with about 12 games to go. Would still need to lose all remaining games to tie most losses in season, 121 (2024 Chicago White sox). Very unlikey but they could come close with 115 losses or more. They will not reach lowest winning percantage in modern era of .235 (1916 Phil Athletcis) but I don’t think below .300 happens often. Will be a very bad season but not as awful as it was looking.

[justdoit]

I have no idea why my post above is repeated with weird formatting. Wish I could fix or delete it.

[Railer13]

@justdoit

Rockies imrproved in second half of season and are 41-109 with .273 winning percentage with about 12 games to go. Would still need to lose all remaining games to tie most losses in season, 121 (2024 Chicago White sox). Very unlikey but they could come close with 115 losses or more. They will not reach lowest winning percantage in modern era of .235 (1916 Phil Athletcis) but I don’t think below .300 happens often. Will be a very bad season but not as awful as it was looking.

Well, I’ll quote it for you. Looks fine. Except for the misspellings, of course.