Home
Artists
Search Recent Random
Posts
Search DMs Popular Hash Lookup Tags Random
Importer
Import FAQ
Register
Partychan Matrix ThePornDude
Home Artists Posts Import Register
The Offical Matrix Groupchat is online! >>CLICK HERE<<

Stats 101, Part 3 (sneak peek) (Patreon)

Published:
2019-09-23 20:52:18
Imported:
2023-07

Content

Here's an early look at the third installment of the Stats 101 series on YouTube. As mentioned on the new Athletic podcast "All Axes" with Seth Partnow, the series will start to grow more advanced as it progresses. (Episode available here for free: https://theathletic.com/podcast/8-back-to-back/?episode=46)

Most of this video will be review for you, although hopefully there are some thought-provoking nuggets in there. I'll also point you to some additional resources if you're interested: 

The last two links are particularly interesting, both suggesting that teams could be more aggressive crashing the glass without a loss in transition defense. I think there's a bit more to it than that -- including fatigue! -- but it's certainly interesting to think about how teams could exploit offensive rebounding edges as the league plays smaller/fluid lineups. 

Enjoy!

Files

(No title)

Comments

Anonymous

Terrific content Ben! I am currently reading the Dean Oliver book, great that you're constantly referring to some of his concepts.

Anonymous

Regarding "the four factors", I find it interesting that they overlap Dick DeVenzio's four simple rules from Stuff Good Players Should Know: 1. Always "MOOOOOVE!" 2. Always Throw the Ball to Your Own Team. 3. Take Only Very Easy Shots. 4. Never Give the Other Team an Easy Shot.

Anonymous

Thank you for another excellent entry to this series! I have a question about your valuation of scoring: it does not seem to capture the value of being able to sustain efficient prolific scoring for more minutes or possessions per game. For example, Michael Jordan (as demonstrated in the second half of 1998 ECF Game 7) was able to score most efficiently by driving hard to the basket and drawing fouls, but it expended too much energy to sustain for an entire game. He could have “gamed” your metric by going all out for 20 minutes a game, but it wouldn’t have made him a more valuable “scorer” for his team. Likewise, KD might be able to average 30+ per 75 on 10+ rTS% by going all out for 20 minutes a game, and it wouldn’t make more valuable, because he would have piled up a smaller advantage for his team. Is there a way to quantify the value of prolific, efficient, AND sustained scoring per game?

Anonymous

I LOVE IT HER!!!

Anonymous
>>27147045

I'm interested to hear Ben's answer. What I tell people at work (analytics) is not to get too hung up on an individual metric if we can get to what you want. We can multiple ppp by possessions per game per player and get to what you are talking about - sustained ppp per actual game. There is probably value in that. At the same time, what if someone, let's call him Al, can play all 48 minutes and average 1.1 ppp, but we have a fast-burner, Bob, who can score 1.2 ppp in 24 min with a teammate, Chuck, who can put in a lower quality 1.04 ppp for the other 24 min. Bob and Chuck give you 1.12 ppp for the game, so you are slightly better off than with high-endurance Al. There is a breakeven point for sure - like would you use a roster slot for a player that could hit 100% of his 3s for 1 possession per game? Probably not.

Ben Taylor
>>27147045

Many good thoughts from Ray. Frank, almost every star player logs similar minutes, so subtle differences in effort/output at a scoring level is immaterial to me. Medium-minute players haven't historically hit a wall in per minute stats when their playing time increases by a reasonable amount. Empirically, it does seem upticking to 45 mpg or playing an entire half will hurt, but again, these are very rare instance.

Ben Taylor
>>27147045

There's also what I think of as "the Manu Ginobili problem" -- some players are just hyper efficient in medium-length periods but can't uptick playing time. That seems to be a more case-by-case basis, but of course looking at huge MPG differences or extrapolating out to per game value (you'll see I do this with impact metrics a lot) can numerically capture those differences to a degree.

Anonymous
>>27147045

Ray, for that specific example, you would in the 4th quarter of every close game. ;) (If only for psychological reasons and because average eFG% drops at the end of games, so that would actually be the most effective time to deploy.) But I'm being facetious. I keep asking these questions about per-possession metrics, because I feel that they slightly undervalue players from older eras, which leads to seemingly artificial era curves that, Ben, you sometimes apply (to players like Oscar, West, and Barry) e.g. to address Rick Barry's complaint. There's something to be said for those players being able to play more minutes and more possessions a game at the time (even if they might not be able to in today's game). I'm really nitpicking here, as I am the biggest fan of all your work and appreciate everything that you've published. Thanks for your answers, Ray and Ben. :)

Anonymous
>>27147045

If Jordan could sustain crashing for fouls throughout the game, he would be a better player. He can't, and the metric rates him accordingly.