APBRmetrics

Analysts...be gentle. First time poster here. Would love to learn from you on a chewy problem.

Just starting to look at analyzing basketball with numbers after coaching U11-U13 youth girls for 4 years. As dad I'm growing as they grow in the game. If this isn't welcome, fair game. I'm a big believer in that which is observed can be improved and I love to watch and learn by analyzing footage with stats to back it up.

I would like your opinions on a method I'm looking at to track aggression and confidence in youth girls driving to the hoop.

My goal is to develop a metric that can be tracked over time to see if there is measurable improvement. There's significant ambiguity in play here but I think I found some interesting markers to monitor while considering some other factors.

Watched 8 U15 girls play their first outdoor 4x4 scrimmage after 6 months of Covid rules that didn't let them play. It was bad basketball with a new team shaking off the rust and finally getting to play together after nothing but drills and skills.

Recorded 60 minutes of play on video.
8 girls each got their hands on the ball about 35-50 times.

Used Dean Oliver's shorthand from Basketball on Paper to record what happened on each possession. It's a great tool for being able to record so many aspects of what happens in a game. Was able to do all the basic stats for FG%, rebounds, steals, turnovers etc. Standard stuff.

Really noticed how hesitant some girls were to drive/or shoot on the hoop, choosing instead to pass the ball over and over until an opening came up. I love the idea of getting the right shot to the right look but I just kept seeing girls who weren't confident driving/shooting and wanted to find a way to score or measure that.

I have 14 year olds and 13 year olds and I find a big difference in size and maturity between this age, especially with the smaller ones who aren't fully through puberty.

Here's what I recorded from the shorthand. If they had the ball in their hand and could either pass or drive on the basket I tracked that decision. Inbounds where you have to pass were not tracked. Could I get an insight from a drive/pass percentage about who was passing too often or giving up their dribble?

Patterns began to emerge. 

Top scorer is a 14 yr old BIG. Size advantage, confident scorer, good ball handler.

21/35 times or 60% she drove/shot on the hoop.  Scored 14 points but ZERO assists. 14 pts total impact on game

Recognize that assists matter as well in creating something for the team.

15/52 for 28%   Scored 8 points but 4 assists (a small gutsy smart 14 year old) 16 PTS impact POINT GUARD
18/53 for 33%.  Scored 8 points but 2 assists   (an experienced 13 year old)  12 PTS impact POINT GUARD

26/44 for 59% Scored 4 pts 2 assists (a small but confident 13 year old who shot poorly)  8 pts impact
12/44 for 28% Scored 4 pts 2 assists (less confident 13 year old)  8 pts impact
24/50 for 48% Scored 4 pts 1 assist (a small but confident 13 year old who shot terribly for the night) 6 pts impact
20/39 for 51%  Scored 4 pts 1 assist (14 year old with size but poor handles)  6 pts impact
11/33 for 33% Scored 0 pts 1 assist (our smallest least experienced player)  2 pts impact on game

So a higher percentage appears to indicate a confidence to drive/shoot the hoop or finish in traffic. To what outcome or impact? Is it worth doing?

I see a need to reward creating for others and getting assists.
Point guards also need a mitigating factor. They're supposed to create for others and facilitate and should be expected to drive less as they should be hooking up their teammates.

I think I'm asking for a way to create a weighted formula.

% of drives * PTS scored * (assists*WEIGHT FOR ASSISTS) * (PG YES OR NO*MITIGATING WEIGHT FOR POINT GUARDS DOING WHAT THEY ARE SUPPOSED TO DO) = CONFIDENCE RATIO?

I recognize this may be a strange one (or hopelessly naive) that I might be pushing too hard to quantify. If anyone could help me with a formula and some weights that could take into account the impact of assists and the duties of a point guard it would be greatly appreciated.

Be kind...

Coach J.

Sounds like you're measuring, or have the ability to measure, everything you want to. And you're off to a great start!

I have two thoughts. One you already have in there somewhat, which is the possible adjustment for point guards. I would make it a little more broad than that and think about what role each player has on the team. If you have someone in the post a lot then they probably just aren't going to drive very much. If you have someone who isn't a good passer, maybe you don't want them driving. Whether you make that a weight in a formula or a qualitative adjustment when looking at the numbers is up to you.

The other thought is to think about what you want to reward, exactly, and make sure your measure is as focused on that as possible. If you want to encourage the girls to drive more and just build that confidence, I would take out any points and assists. Obviously you would like points to be the outcome of the drives, but if the current goal is just to get them to drive in the first place then look only at drive percentage or something similar. Maybe after you instill that confidence, or for the girls that don't need the encouragement, then you add in the outcome of the drive and encourage them to shoot better or read the situation and make better passes. You're the coach, so you know what you want them to learn, so you have to decide what the emphasis is and for which players.

Some quick comments:

You give data for touches and drives. What about catch n shoot outside shots? Do they exist? Can they be made? What is the average fg% from "deep" (what distance is that) vs. "mid-range" or "close"?

Can you judge each touch as "well-covered" or "good opportunity to drive"? Itemize distance to rim?

These extra details would enrich your opportunity for analysis and strategy improvement.

Average drive / touch ratio is about 42%. That not bad. 2 of 3 14 yr olds above that, 2 of 5 13 yr olds. Ask for more but have relative patience.

If all baskets are 2s that would make about 55% assisted fgs. Less if their are foul shots or 3s. That is pretty high, suggesting drives are not the only thing of importance.

Could set goals for each player: 5% more drives / touch than last game or last week?

Points per drive? Again depends on catch n shoot opportunities and ftas. But until that is known, maybe assume .32 pts per drive.

What is the points per shot on potentially assisted baskets vs. unassisted? That is pretty important for choosing strategy.

Could ask every player to try for at least one basket and one assist especially early in game and then have a goal for more of the contribution they do best now.

For points impact, 2 pts for basket and 2 pts for assist would be over-counting in the abstract, though it may be alright to reward assist behavior.

Objectively and at a higher level of play, unassisted baskets might be worth 2 pts, assisted baskets 1 - 1.5 and assists 0.5 - 1 to split that credit.

If avg. Is .32pts / drive now, then the next goal might be .4pts / drive. Which is 20% fg% per possession used. So driver should take shots she can hit 20% or dish to someone in a position where they can hit 20%. Make a list by the basic 3 shot zones (or more detailed) of what is a good or acceptable shot for each player. Or pass to someone else to look for those opportunities.

Maybe have a chart showing good shots by name and zone. If a player doesn't have a good zone yet, identify the priority learning / testing shot zone. Focus on success in one then expand. Get a star, get another star... If don't want too much pressure for success, instead of good shot / not, could say better shot / ok shot.

Your top 3 scorers should probably shoot first / most of time or pass to one another for efficiency. Everybody else probably should look for one of the top 3 first or shot if they feel good about it 2nd for efficiency. If efficiency is the goal. If you can stagger the lineups to keep 2 of the best 3 shooters on court at all times, I would. Maybe track assist partners and player pair plus minus to guide lineup flow.

Probably should track driving & passing turnovers.



Is there a shot clock? If so, shoot before a crunch-time decline in efficiency.

Don't worry about a fancy analysis formula. Just compute simple / major ratios.

For girls hesitant to drive, maybe encourage "take 2 dribbles at least every other time and then decide what to do next". Ease into driving. Then more shots (and possibly more dribbles if worth it) with time.

Encourage pass and / or shot fakes before and after dribbling.

Does team have film review sessions or do you offer players comments from film? For makes, a couple things were done right. Congrats / reminder of the key items you did well. Miss? I / you can have a better chance next time if focused on this important step.

If you want to diversify habits, drill the 9 possibilities: Pass to them into an immediate shot, drive or pass, same after 1-2 seconds of looking / thinking / faking or same after or same after perhaps 3-4 seconds max. Every player does every variation, one variation at a time (watching each other for technique) then call out specific coaching requests and / or freestyle some rounds. Know what you have to do for each and just do it, know what it feels like, know that you have done it before. 9 "moves". Maybe limit it to 3-5 for some. Don't feel you have to invent the game of basketball every time. Just do one of 9 things.
Take the drill to different spots on court you want players to get to. String one decision in one spot to other spots. Have players providing passing / shot options and eventually defensive pressure.


Eventually you could track each player by each move by court zone in games. Ultimate points off move, or did it improve position, turnovers, stress...

Take what you want from these thoughts.