decoding BBref

[roger_that]

Maybe some of yuz have figured this one out? BBref has a category I just noticed (probably up there for ages) they call "cLI" which measures "the importance of a team's winning the World Series," 1.0 denoting "average" and above that "greater than average importance," etc.



The Mets just went above 1.50 for the first time all year, in the Braves' series. and then went below 1.0 in the Cubs series also for the first time all season (except for one June game vs. the Astros, which came in at .98).



Did you see it like that, subjectively, that the Braves series was the most important one by far and the Cubs series the least important? Not really sure what exactly cLI is measuring here.

[Ceetar]

championship leverage index. Beating a team like the Cubs that aren't competing against you for a spot, versus beating the Braves who are fighting for multiple spots with you is HUGE.

[roger_that]

Is the cLI going to go up as the season progresses against NL East rivals and way down as we face interleague competition? How come it didn't dip below 1.0 when we faced the Angels? Seems to me that series didn't mean as much as facing an NL team, but yet...

[Ceetar]

It's less now against the Cubs because we've banked a ton more wins since the Angels series. The probability that the Met make the playoffs is that much higher, so the consequences of losing to the Cubs in mid-july is just less. Likewise, it'll mostly drop going forward because the chances that the Mets miss the playoffs grows less likely. Seeding just doesn't matter enough to push the needle this far out.

[roger_that]

So let me get this straight: the Mets' cLI is going to rise as the season progresses against its NL East rivals, but drop vs. AL teams, and drop less vs. NL West and Central teams? Or is it going to rise but more slowly against the NL West and Central? Is it going to hit 2.0 against the Braves (assuming they're close behind) in August, and 3.0 in September?



Looking backward, most of the Mets' high leverage games have been vs. the Braves and Phils, which only makes sense, except for the four games facing the Cards in mid-May, which rank #,4, 5, 9, and 10 on the cLI list. Did someone have a Whitey Herzog flashback?

[Ceetar]

honestly it'll probably drop for everyone, except maybe key matchups against Atlanta, depending on how close the standings stay. Fangraphs has the Mets at 99.5% to make the playoffs, so the only thing that's gonna significantly push the WS odds is potentially losing the division. But a lead is more and more stable the fewer games there are, so that works against them. The next Atlanta series might still be high, but if the Mets come out of that better than they are now, with that many more games in the bank, it probably won't change. It's not a particularly interesting number in mid-July.

[roger_that]

Any ideas on why that Cards' series would rank that high? I'm trying to make this metric make some sense to me.

[Ceetar]

my guess is just that it's another team in the hunt for a WC spot and it was early enough that simply winning games against the competition is key.



Mostly I think you're overvaluing the difference between like 1.3 and 1.2.



for instance, in game of of the NLDS last year, LA starting pitcher Walker Buehler averaged a 9.71 cLI in his 6.1 IP. I think this means basically one batter in that game was worth about as much as a random May game.



The first scoring play in game 5 of that series, top six, a double by Corey Seager that scored a run, was worth 77.92.

[Ceetar]

https://www.sports-reference.com/blog/2020/09/__trashed-2/https://www.sports-reference.com/blog/2020/09/__trashed-2/

[Ceetar]

the wild pitch, btw, was worth 635.99.



Topped only by the 643.33 of Keith's 2-run single in the bottom of the 6th of game 7.