data golf

statistics and data visualization, applied to golf.

[No Laying Up @NoLayingUp]

How the top 10 in true strokes gained has evolved over the last five years via @DataGolf.

What sticks out?

[data golf @DataGolf]

@jones_zamboni No. Did not realize he was hurt (but would not have affected model anyways).

[data golf @DataGolf]

@KlpgaEnglish Our rankings use SG, i.e. performance relative to your peers on the same course/day. And then adjusted for the strength of your peers. So the difficulty of the course is not relevant.

[data golf @DataGolf]

@slicedpod @BenColeyGolf Thank you for the kind words.

I do think having SG at the cup events has been enlightening for understanding how a team won.

But it can be taken too far, especially at the player level. As good as Clark was for most of the event, he wasn't great in some big spots.

[data golf @DataGolf]

@ProjectRSL @GolfingTruther @NoLayingUp Not sure what you are saying exactly, but yes we estimate the lineup decisions to be only a 0.4-0.8% difference in win probability. Scottie isn't literally being "wasted" on Nico, but it did hurt the US's chances very slightly to not have him playing someone better.

[data golf @DataGolf]

We also have an unnecessarily long write-up on why allowing Scheffler vs. Nico was, in fact, bad for the US (even if it's a pretty tiny effect): https://t.co/IA4o7IcfC9

[data golf @DataGolf]

@kylehvmt Nothing really changed though. These stretches are mostly just randomness.

[data golf @DataGolf]

Not defending the US leadership, but the 8-26 foursomes record since 2023 is really cherry-picked:

[data golf @DataGolf]

@ProjectRSL @GolfingTruther @NoLayingUp no i think the math confirmed everyone's intuition that Scottie was being "wasted" against Nico. Not going to decide the cup, but suboptimal.

[data golf @DataGolf]

@GolfingTruther also, if you believe in momentum or other things like that, then there could be some downside to purely maximizing lopsidedness. today, for example, seems like the INTs lineup is stronger at the bottom.

[data golf @DataGolf]

@GolfingTruther well, 0.4-0.8% isn't nothing. But ya it's not like you would ever "notice" this unless 100s or 1000s of cups were played with these same strategies deployed.

[data golf @DataGolf]

Full table is here: https://t.co/z7FInPYtOB

[data golf @DataGolf]

@KlpgaEnglish in the rankings, all that matters is performance relative to your peers. If you play mountain courses well, and you mostly play mountain courses, you'll be ranked better than otherwise. This is a form of course fit (which rankings do not account for, by design).

[data golf @DataGolf]

@TalkingSP0RT the full names are there on desktop... but ya should probably add an initial to mobile.

[Kyle Porter @KylePorterNS]

An hour ago, @datagolf had the U.S. at nearly 60 percent to win the event. Now, it's back down to 35 percent.

[data golf @DataGolf]

@rob_dimino @outside_the_cup Yes, the tweet technically should have read: Best American *records*...

[data golf @DataGolf]

Best American players across both Ryder and Presidents Cups:

Points Above Replacement is the number of points better than earning a tie in every match.

[data golf @DataGolf]

@outside_the_cup @rob_dimino I'm not saying it's an advanced stat or the optimal way to evaluate players. It's just how many points you've earned above tying every match. To me it's useful for comparing records of players with different numbers of matches played.

[data golf @DataGolf]

@rob_dimino @outside_the_cup You called the table disingenuous. It's literally just a list of the best records. Nobody is claiming it's the optimal way to evaluate players.

[data golf @DataGolf]

@xprezJerry Clicking on names here it can be: https://t.co/punEn5YvWf

[data golf @DataGolf]

@outside_the_cup @rob_dimino 'twas a joke. Can't seriously reply to a person like this.