I can't believe people itt care about the merits of what Simmons said at this point. The only thing that really matters relative to this conflict with Mike Golic is whether he was slamming LeBron after four games or not, which he clearly wasn't. Whether his hot take was valid or not couldn't be more irrelevant and is pretty much a derail.

I think your first sentence is wrong. People are reacting to Simmons as they would anyone else talking for 5 minutes on national radio about: 'LeBron looking a different player, lost his explosiveness, has 40k minutes in the league might he be in a different phase of his career??'

re: your second sentence, how did Mike and Mike "deliberately [take] his words out of context"?

I also enjoyed this piece. Much more artistically written than I would have expected on Grantland.

The only way you can possibly ask this if you don't know what the phrase "taken out of context" actually means.

Argh, why does neither a search for "Sumo" or "Sea of Crisis" in the top right corner of Grantland bring up the Sumo article? It seems to only be bringing back blog posts, but seems like people might sometimes search for something other than blogs, no?

FWIW I thought the NFL Loan fantasy draft article by Mays was pretty entertaining read and a pretty entertaining concept that would never ever happen. I also just enjoyed thinking about adding JPP to Detroit's DLine...

What's worse is that this isn't even his original take. He stole it directly from TZ which he was clearly reading moments before he went on the air as Caldarooni made that reference first.

the bigger issue here is that Simmons looks like Ellen DeGeneres

This piece seems poorly-conceived. It doesn't make much sense to use ELO (which incorporates actual draw rates) to generate a points-per-game prediction (which includes draw rates), then hold those two as fixed quantities while manipulating draw rates as your variable.

This is all correct, but I'm not sure he's teasing out the implications. To continue the trend: if draws are expected 80% of the time, Carlsen is expected to win 100% of decisive games. If draws are expected 90% of the time, Carlsen is expected to win 150% of decisive games. The supported conclusion is just that, if you generate result Y based on variables A and B, fixing Y and changing A will mathematically affect B. It doesn't mean that, "Given this analysis, bookmakers may be underestimating the defending champ’s chances."

Here's where he misses a valuable sanity check. Of course the favored player isn't better off playing to draw more frequently. By definition, if Carlsen is a favorite in every individual game, every draw is harmful to his chances of winning. He's "a weaker player overall" when playing to draw more not because it compromises his tactics (which is how I'm reading his implication here) but because .6 > .5.

makes me wanna wander around Tokyo...

It's the first hit if you just google "grandland sumo" (I haven't read it, but it's one of those articles that all the sports writers I follow share with notes like "I hate this because it's so good it makes me feel like a jealous failure")

article's name is "sea of crises", plural. it comes up in the search.

amazing

My understanding is that ELO is a mapping from two players -> expected points per game. However, we cant use that to generate a match win probability due to the game draw probability variable. The author is basically saying "there isnt a plausible game draw probability assumption you can make that leads to any conclusion but Carlsen is undervalued", which seems perfectly reasonable. Put another way: even if ELO cant be used to tell us an exact match win probability, its impossible to use the model to come away with a match win probability that isnt Carlsen ~>80%


That said, I think its overboard to conclude that there is value in a line when far and away most well known model for the bet says there is value.

No time to edit, but just read through the article. The author says that there "may be" value in the line, and then gives caveats about how the article assumes ELO is accurate which but it isnt, so that doesnt seem overboard at all.

Here's my best go at explaining my theoretical objection to this, apologies for length:

Elo's calibration is based on aggregate frequencies of win/draw/loss at various rating points and differentials. If we assume 50% is the true draw rate at this ratings environment, then, over time, a 71-point gap will equate to a 70% chance for the stronger player to win any decisive game. But the background draw rate is built into what this 71-point gap means. If the two players agreed to roll a six-sided die before the game and draw unless the die was 6, we wouldn't think, "Since our expectations are .6 ppg, Carlsen must win >100% of decisive games." We'd have no problem accepting that the 71-point-gap isn't calibrated for this situation.

Similarly, if we estimate that two players are drawing 70% instead of 50% of the time, the conclusion should be "This isn't a typical 71-point-gap grandmaster matchup," not "Well, we now expect Carlsen to win 83.3% of decisive matches instead of 70%." If we had infinite matches between the two at 70% and the gap stayed at 71 points, then we'd make that change. Over one match, Elo doesn't imply a causal link like this. It's that conflation of long-term calibration and short-term causality that's where the author goes wrong in paragraphs like:

Carlsen is unequivocally not better off playing to draw more often, as he is the favorite, and additional draws do not make him more likely to win decisive games.

The author throws in the necessary caveats, but I think he's still making this fundamental error.

I enjoy the concepts but I have not much agreed with the decisions in those redrafts. Taking Eli over Cam seems comical to me, as does not taking DeAndre Hopkins in the top 20 of a 2013 redraft.

This is fair, but I think the accurate treatment of the ratings gap reduces to "Elo is higher on Carlsen than the market."

"He reminds me a tiny bit of Pujols and the Angels, and I hope that's not how this plays out, but..."
Golic says this is ridiculous to say 4 games into a season.

There's no further context necessary for listeners to understand that Simmons is comparing LeBron to Pujols after 4 NBA games, and Golic objects to such characterization.
i.e. Simmons himself acknowledged he's making the comparison based on a short sample size, that is irrelevant to Golic's position that it's too early to make such a comparison regardless of whether you acknowledge the small sample size or not.

Except that he's not. As has been demonstrated multiple times in this thread, be providing the ADDITIONAL CONTEXT surrounding what Simmons said.

i liked the part where he quoted the post about taking things out of context and then wrote a post based off a 21-word chopped quote

jesus make up your mind

I think that Golic and some of his defenders don't really understand what people mean when they talk about sample size in sports. If Simmons were talking about LeBron getting off to a cold shooting start, then absolutely sample size would be a defense. But making physical observations of an aging player after he publicly announced he dropped significant amounts of weight during this offseason is something completely different. There's a very good reason to worry about whether he's athletically the same guy he was last season, and the early returns seem to be that he isn't. It's not out of bounds at all to discuss that even if it's only been four games.

Footnote from simmons picks column explaining his record

I linked your response in the chess sub's '14 WCC thread. It's gonna be bumping come game day(s).

I hadn't thought about whether the piece's statistical underpinnings were sound or not, but good points. What I liked: I felt the chess-specific language was refreshingly accurate. Most mainstream chess articles are rife with niggling misuses of jargon and terminology, combinations of words few reasonably serious chess players would consider using (q.v. Nate Silver's sloppy use of "threaten" upthread).

This patch is a little bumpy:

The first sentence (post em dash) is the wrong stat -- the draw rate in 2750+ vs. 2750+ games is even higher than 50% afaik. The half figure includes games played against sub-elite competition. The graph he linked* cuts off at 2600 anyway. Additionally, "breakthrough" has a specific meaning in chess; you don't want to use it so casually. Not all wins are achieved through "breakthroughs".

But apart from that it was pleasant.

I can't imagine this sort of thing bothering most of a mainstream chess article's target audience. But if it's not obvious why these errors or a sentence like this can be annoying, compare it to this howler.


* Incidentally, it made two mistakes of the sort I've been discussing. If you have to use the word "expert", the ratings on your graph had better fall within 1900–2100 FIDE or 2000–2200 USCF. And we all know that "match" isn't a synonym for "game".

Both would be true; playing to draw is generally bad, for both psychological and on-the-board reasons.

I'm not even sure how Carlsen could play to draw. His "trademark", if you can call it that, is winning dry, "drawn" endgames, almost as if through sheer endurance. It has even been the subject of scorn of the YOU JELLY? variety: example #1, example #2. Two of the three decisive games in last year's Carlsen–Anand match were of that sort.