Analytics

[sedz]

I want to offer an analogy and hopefully clear the air a little bit. Basketball metrics have a lot in common with temperature metrics. The average person doesn't understand what temperature is really measuring (it's not heat!). But you don't have to understand temperature to know that 70 feels nice or that it might snow under 32. You can set your thermostat or decide what to wear without taking a course in thermodynamics. You can also say that temperature doesn't tell the whole story, that just going outside is better. That's true. Humidity, air pressure, dew point, elevation, wind, cloud cover, sun angle, and solar irradiance can also affect how the air feels. So we have a "heat index" or "feels like" temperature that takes all of that into consideration. Box Plus Minus is like that. It takes a bunch of individual rate stats along with the score to generate a sort of "performance index" for basketball players.

Maybe you think heat index and BPM are mumbo jumbo. That's fine - you don't have to use them. Using rate stats or BPM is like using Celsius instead of Fahrenheit. Anyone can do it, you don't have to be smart. You just have to learn a new system that opens up all kinds of analysis that's impossible under the old system.

[Lobot]

The NCAA has announced that the BartTorvik.com rankings will now be included on official teamsheets during NCAA selection committee proceedings.

[GoBearcats31]

As of this morning, UC is 9th in the country in "2023-24 Power 6 Production" (scoring) on next year's roster
https://twitter.com/CoachAdragna/sta...26573838032932

[sedz]

The Pelicans are putting on a masterclass in how to defend with no bigs. They are switching everything, and the Lakers are getting LeBron matched up against tiny Alvorado, but the Pelicans are doubling to force the ball out of his hands.

[sedz]

What do analytics, especially defense tell us about the one and done players the past five seasons- versus players drafted with 3 or 4 years in college? It seems to me that the NBA doesn't play or cherish defense in the "younger guys with a ton of potential". I'm curious to see if what helps a player help his team win at the college level is what helps him get drafted higher. My thesis is the NBA only cares about defense outside the Lottery picks (except the kid from France whose name I couldn't spell if I did watch more NBA).

The NBA is so different from college basketball it might as well be a different sport. With the shooters they have, screens have to be defended out to 30 ft and the defense has to scramble to cover a huge area. Length, athleticism, and quickness are absolutely critical. If you have those qualities, you'll get drafted and they will teach you how to shoot and defend. I don't think NBA GMs and scouts care much at all about statistical output in college. They just want to see the tools. We've seen that on our teams. Lance Stephenson was our highest draft pick since Kenyon, but he was worse than Rashad Bishop and Deonta Vaughn as a college player. Sean Kilpatrick and Gary Clark went undrafted. They care about a player's potential to defend the more open NBA game. They don't care if you can shut down college players.

[sedz]

If Reed can improve his offense, do the stats show him to be as good a defender asI think he is? He made mistakes in both the past two seasons, but seemed to be the guy relieving Newmann and being a lockdown defender mid way through this year. Maybe if you can break down his stats just to conference games, will we see an improvement over first part of the year? Also, add Rayvon as RS freshman. I believe the year on the bench will let him shine next year,

Yes, you're exactly right. In conference play, Reed was our top defender statistically. If we remove Ody due to low minutes, Reed was tied with Newman as our third best defender over the whole season, behind Aziz and Day Day. His ability to defend all five positions is very valuable. Hopefully he can add something on offense.

[red_n_black_attack]

just quoting to get the mention.

What do analytics, especially defense tell us about the one and done players the past five seasons- versus players drafted with 3 or 4 years in college? It seems to me that the NBA doesn't play or cherish defense in the "younger guys with a ton of potential". I'm curious to see if what helps a player help his team win at the college level is what helps him get drafted higher. My thesis is the NBA only cares about defense outside the Lottery picks (except the kid from France whose name I couldn't spell if I did watch more NBA).

[GoBearcats31]

KenPom #s through the Wes Miller era:

2022 - #101 overall (145th adjusted offensive efficiency, 75th adjusted defensive efficiency)
2023 - #50 overall (56th, 57th)
2024 - #39 overall (79th, 19th)

[sedz]

Ody's BPM each season: https://barttorvik.com/playerstat.ph...a&t=Cincinnati

2020: -0.3 (-1.9 offense, 1.6 defense)
2021: -1.6 (-0.8 offense, -0.8 defense)
2022: 0.0 (-0.8 offense, 0.8 defense)
2023: -0.3 (-1.2 offense, 0.9 defense)
2024: 2.0 (-1.3 offense, 3.2 defense)

He played significant minutes until this year. He became a good defender in his final season, posting career bests in block rate, steal rate, and foul rate. There were several games in conference I wanted Ody to play instead of Reynolds due to defensive matchups, most notably against OK St when they ran 5 out offense. Jamille had a lower BPM than Ody did this season, partly because he isn't quick enough to defend the perimeter (also because he had the highest turnover rate on the team).

I'll move the Ody discussion over here. Note that the "fancy stats" show that Ody was a good defender this year. 3.2 is third on our team, below Aziz and Day Day but just above Newman and Reed.

[sedz]

On the individual player side, Box Plus Minus or BPM is a player's impact on the scoring margin per 100 possessions. Zach Edey led all players with 15.5 BPM this year, meaning he gave his team a 0.155 point per possession scoring margin when he was on the court. That's almost 11 points over a typical 70 possession game. If he played all 40 minutes in that hypothetical game with 15.5 BPM, he would have 11 / 2 = 5.5 net points (the other half is assigned to the other team as a negative value).

Some of these numbers seem like voodoo at first, but they are simple calculations that make sense when you get into it. The point is to boil contributions down to single numbers that can be compared apples to apples so better decisions can be made.

[sedz]

Forgive my learning curve, but I do actually like statistics. I'm guessing effective field goal percentage has a multiplier for 3-pt attempts? Do any players shoot 54% or better from 3-pt range?

I shouldn't throw terms around like this so much without explaining them. Maybe I should do crash course or glossary of terms at some point this offseason.

Effective field goal percentage converts everything to a two point shot. So yes, there's a 1.5 multiplier applied to three pointers. If you shoot 30% from three, that's 30 x 1.5 = 45% effective field goals.

Doubling the effective field goal percentage gives you points per shot, so 50% effective field goals is a point per shot. True shooting percentage adds free throws. Then you add offensive rebounds and subtract turnovers to figure out how many shots per possession you had, and you end up with points per possession. Those are the basic building blocks of basketball analytics.

[red_n_black_attack]

Top 50 shooting teams make 54% effective field goals, which should be our goal. Only four players in the entire country made that percentage from midrange on at least 100 attempts for the season.

Forgive my learning curve, but I do actually like statistics. I'm guessing effective field goal percentage has a multiplier for 3-pt attempts? Do any players shoot 54% or better from 3-pt range?

[sedz]

You're probably one of those guys that talks about WAR WHIP and other new fangled metrics in baseball. jk

In this week's Reds games the Nationals had wOBA and RC+ on the scoreboard. Even college softball broadcasts use OPS now!

The nerds are taking over sports.

[sedz]

here's a question. it seems that we all love jizzle james. his "signature" (and i'm sure that there's "analytics" that i'm wrong) shot seems to be the midrange, two pointer. since that is now apparently a "bad" shot, should uc make him stop shooting it? **** no - but, my question is somewhat rhetorical - at some point, you have to rely on more than just analytics - different guys have different abiliites and that has to come into play.

If it becomes an efficient shot you give him the green light. Midrange in general is a bad shot. There are only about 25 college players in the country who make a higher percentage from midrange than the national average effective field goal percentage (I don't think there's a single player in the NBA who does). Our effective field goal percentage this year was 50.3%, so a good shot for us is above that. Jizzle made 47% from midrange, so he's really close and he shot it really well late in the season. Vik made 53% with his baby hook. Jamille made 41% and everyone else was under 40%.

Top 50 shooting teams make 54% effective field goals, which should be our goal. Only four players in the entire country made that percentage from midrange on at least 100 attempts for the season.

[red_n_black_attack]

Nate Oats looks at Torvik. Swoon.

"We have a third-party analytics company that does a great job. We look at KenPom a lot, Bart Torvik."
https://www.si.com/college/alabama/c...ess-conference

You're probably one of those guys that talks about WAR WHIP and other new fangled metrics in baseball. jk