Fangraphs projects CWS at 81-81, 4GB in ALC

[GGajewski18]

I would agree with 81 wins.

[Bananarchy]

I would be okay with 81 wins.

[Jose Abreu]

The records of teams in these predictions, even teams they are predicting to win the division, are always awful. Does anyone really think all 5 teams will be separated by 7 games in the standings?

This.

[Chicago White Sox]

FWIW KC led baseball in "cluster luck" last year.

 

https://thepowerrank.com/cluster-luck/

 

Their rotation is still awful, they've lost a couple key bullpen pieces, they have a black hole at 2B, and it's very likely we'll see regression from Cain, Morales, Hosmer, and Moustakas (him in particular because he's never come close to hitting that well before), and Escobar and Dyson can't hit. I do think the projections are harsh on them (I don't think WAR accurately represents relievers' value yet, and the defensive metrics for WAR are not particularly reliable either, especially for projections), but I agree with raBBit and the projections that Cleveland is currently the team to beat in the ALC.

How do you measure cluster luck? And wouldn't a team that strikes out infrequently be more likely to string hits together frequently?

[caulfield12]

I think it's because it's very hard to statistically measure "keep the line moving" and momentum shifts and swings in a baseball game...or game-changing plays on the basepaths and especially from a defensive standpoint.

 

One losing team's bad luck is another winning team's pattern of cluctchness or being comfortable/confident in those situations. Success breeds more success.

 

Just the same way in the NFL breaks seem to go to teams like the Steelers, Patriots or Seahawks when they're facing teams without as much experience on the biggest stage.

[QuickJones81]

If you replace a 0 WAR player with a 3 WAR player does that add 3 wins to your projection, or is it more complicated than that?

[shysocks]

How do you measure cluster luck?

Every possible result of a plate appearance has a number of runs it creates based on historical data. This table is dated I'm sure, but it shows that a homer adds an linear weighted average of 1.409 runs, a triple 1.063 runs, etc. Add up all a team's outcomes in a season and you can find how many runs it "should have" scored. If they scored more than that, they had good cluster luck. The same thing can be done on the pitching side.

 

A team could put up a single, two walks, and a homer in an inning, but depending on the sequencing they could end up with between one and four runs. It could have only nine hits in a game, all doubles, and could get anywhere from zero to nine runs from them. Linear weights removes the context and just gives each outcome a number.

 

And wouldn't a team that strikes out infrequently be more likely to string hits together frequently?

I would think a team that strikes out less should have a higher batting average, so yeah.

[OmarComing25]

How do you measure cluster luck? And wouldn't a team that strikes out infrequently be more likely to string hits together frequently?

There doesn't seem to be that strong of a correlation between how often a team strikes out and their cluster luck. The Braves struck out the 2nd least behind the Royals and rated poorly. The Orioles and Padres struck out a lot and rated well. The Tigers were by far the worst and they were middle of the pack in strikeouts. Unfortunately it looks like they only have data for 2015, it would have been nice to see multi-year trends on this.

[Chicago White Sox]

Every possible result of a plate appearance has a number of runs it creates based on historical data. This table is dated I'm sure, but it shows that a homer adds an linear weighted average of 1.409 runs, a triple 1.063 runs, etc. Add up all a team's outcomes in a season and you can find how many runs it "should have" scored. If they scored more than that, they had good cluster luck. The same thing can be done on the pitching side.

 

A team could put up a single, two walks, and a homer in an inning, but depending on the sequencing they could end up with between one and four runs. It could have only nine hits in a game, all doubles, and could get anywhere from zero to nine runs from them. Linear weights removes the context and just gives each outcome a number.

 

 

I would think a team that strikes out less should have a higher batting average, so yeah.

So I guess my beef is calling it luck. Get a team full of high-contact, high AVG hitters and it's no longer luck, as you've basically put a hedge in place to reduce the randomness of sequencing. This is one of the flaws with linear weights, as it completely ignores context when in reality it's incredibly important.

[shysocks]

So I guess my beef is calling it luck. Get a team full of high-contact, high AVG hitters and it's no longer luck, as you've basically put a hedge in place to reduce the randomness of sequencing. This is one of the flaws with linear weights, as it completely ignores context when in reality it's incredibly important.

You can expect this team to score more runs than other teams, but I'll argue they're no less dependent on sequencing. They can have 15 hits a game, but even though they're all good hitters, we still have no idea how they'll string those hits together. And how they do it will not be consistent from game to game and season to season.

 

I don't think it's an effort to ignore context, as obviously context incredibly important. I think it's just an attempt to have a measurement free of it.

 

Realized I didn't address the "calling it luck" part, which I can understand. I agree there is more to a big inning than just luck. The "luck" label is there to address the uncertainty of whether you can repeat it over and over.

Edited January 13, 2016 by shysocks

[Boopa1219]

Its funny because I see Cleveland as the least threatening of all of the teams in the division.

 

If they trade for an OF, like a Jorge Soler (Who I still believe has the tools to be as good as Kris Bryant if he stays healthy) to help hold the fort down till Brantly is healthy, I think they could win 83+

[caulfield12]

You can expect this team to score more runs than other teams, but I'll argue they're no less dependent on sequencing. They can have 15 hits a game, but even though they're all good hitters, we still have no idea how they'll string those hits together. And how they do it will not be consistent from game to game and season to season.

 

I don't think it's an effort to ignore context, as obviously context incredibly important. I think it's just an attempt to have a measurement free of it.

 

Realized I didn't address the "calling it luck" part, which I can understand. I agree there is more to a big inning than just luck. The "luck" label is there to address the uncertainty of whether you can repeat it over and over.

 

That's where overall line-up balance in terms of obp, slg, ops and rhb/lhp/switch hitters also factors in.

 

Better to have 6-7 really good hitters than 2-3 great ones combined with a bunch of average to below average.

 

Do you have the type of team that can manufacture runs in tight games?