News Article: Meet the 'secret weapon' in Alabama's Final Four run: Analytics...

[oskie]

Meet the 'secret weapon' in Alabama's Final Four run: Analytics guru Michael Schwimer has many roles for Tide

Schwimer and his firm, Big League Advantage, have done more than just crunch numbers for Nate Oats' coaching staff

www.cbssports.com

We all know about Nate Oats' predilection for the numbers, but here's a deeper dive into what's going on with the "analytics" part of the equation.

This "third-party analytics company we use" ( a reference from Coach's Clemson post-game remarks) turns out to be way more involved - with everything - than I could even imagine.

... "We do, literally, everything," Schwimer said. "[We assist with] who do we get in the transfer portal? How do you build the team? Then the skillsets around each player. Then I sit down for 90 minutes with each player before the summer even starts. I figure out motivational factors for players."...

... Schwimer claims to have video of "every shot every player has taken" from age 15...

This is a fascinating article and a very good read.

 

[TexasBama]

 

[edwd58]

A friend sent me a link to the article this morning and it is a terrific read. Bama and Duke have been their only college basketball customers, but I'd imagine that might be about to change.

 

[GboyBama]

Wow, that's one heck of a read. I hope we can use this to find Connecticut's kryptonite.

 

[4Q Basket Case]

Fascinating article!

I took enough statistics, and worked in bank risk management long enough, to have a rudimentary grasp of the concepts.

Really, it comes down to the law of large numbers. As Oats is quoted in the article, “reverting to the mean.”

The inconsistency is this: Over the course of a season, you get about 1,400 minutes of information. The law of large numbers definitely kicks in at that level.

But all stats re-set every 40 minutes. IOW, it isn’t really a 1,400 minute season. It’s 35 games of 40 minutes each. And the Law of Large Numbers ignores psychological factors like momentum and ideas like “who has the hot hand,” as well as “who has the cold hand.”

Put in the context of an NCAA tournament: At most, it’s 6 games of 40 minutes each — 240 minutes. And it’s single elimination — if you don’t win TONIGHT, your season’s over.

Your relevant time frame just went from 1,400 minutes, to 240, to 40. When you’re getting gashed by somebody who’s unconscious shooting that night, when do you abandon the idea of reversion to the mean? Or do you?

And that doesn’t account for intangible things like psychology. So how do you incorporate that stuff into your analytics? I don’t know. You just got beyond my mathematical understanding. But I know enough to think it can probably be done.

I would love to sit down with Schwimmer for an hour and just nerd out.

 

[JDCrimson]

I think the bigger tell analytically is on UConn. Meaning you are trying to figure out what they do well and what they don't do as well. Then you narrow it down to how they did against players similar to who you have. The 1400 minutes will be diced by a ginzu knife. There is a way to beat UConn, its just if we can execute it.

We played Purdue earlier this year and had the lead for over half the game. We will be able to take some of that and apply to UConn. We played Creighton who beat UConn and we will study them hard again.

I think this is a game for Estrada to shine. I think we must let him take mid-range jumpers and the baby bank shots over UConn bugs. The mid-range game will open up the 3pt line.

 

[RammerJammer14]

View attachment 41626

I took this a different way at first…😆

 

[CajunCrimson]

I also think it helps you ID what UConn hasn’t seen yet. IOW, if your stretch 5 is out, what does that mean for Clingan, does he go out? Or does he switch and now is on RG? Or Sears? Or Estrada? Or Wrightsell?

and can Grant make him pay from 3? Or dribble around him? Or does that give us open wings because Lurch has to now chase a guard?

where can you neutralize their advantages, etc

what a fun week