How the Bucks Use Analytics
Re: How the Bucks Use Analytics
Not enough about the HOW beyond the who and the most basic but jumping around, vague description of process.
How do they choose lineups, factor priorities, shots?
Continue to have Maker in the rotation despite his horrendous RPM estimate? Plan to pay Parker big or huge despite a mild negative RPM that wouldn't support it? Seth Partnow has critiqued, downplayed RPM in past writing. Reading his stuff in past creates an impression to me- right or wrong?- that RPM gets little if any weight and is not "pushed forward". Were these many candidates reviewed in part based of ability to produce RAPM (and splits) or for their opinion about the metric and how to use it or not?
It would have insightful if the interviewer asked these questions and got a clear answer. But since it didn't happen... hey Seth want to clarify any or all of that? If RAPM isn't important / trusted, should be easy to say that and clear up any doubt. Occasionally a peer will say it in public or private. I'd take either but if you have a firm opinion, maybe say it in public. I probably won't ask again.
How do they choose lineups, factor priorities, shots?
Continue to have Maker in the rotation despite his horrendous RPM estimate? Plan to pay Parker big or huge despite a mild negative RPM that wouldn't support it? Seth Partnow has critiqued, downplayed RPM in past writing. Reading his stuff in past creates an impression to me- right or wrong?- that RPM gets little if any weight and is not "pushed forward". Were these many candidates reviewed in part based of ability to produce RAPM (and splits) or for their opinion about the metric and how to use it or not?
It would have insightful if the interviewer asked these questions and got a clear answer. But since it didn't happen... hey Seth want to clarify any or all of that? If RAPM isn't important / trusted, should be easy to say that and clear up any doubt. Occasionally a peer will say it in public or private. I'd take either but if you have a firm opinion, maybe say it in public. I probably won't ask again.
Re: How the Bucks Use Analytics
What's the right way to use Jabari Parker? Worst raw plus minus on team suggests it might not the chosen way.
Re: How the Bucks Use Analytics
More related discussion in tweet chat here @bballstrategy and here viewtopic.php?f=2&t=9458 and both places in past many times, fwiw to you.
Re: How the Bucks Use Analytics
Still early in the Parker comeback. PER of course says above average (a lot because of high usage). WS/48 says average. BPM says hideous (because of low assists and defense). RPM says bad.
What do Bucks' analysts, Coach, GM and owner use and think? Better know what you believe in & why. And better be right.
What do some of the less visible / used / newer metrics say? What do you think?
What do Bucks' analysts, Coach, GM and owner use and think? Better know what you believe in & why. And better be right.
What do some of the less visible / used / newer metrics say? What do you think?
Re: How the Bucks Use Analytics
4-6 in last 10. Giannis and Bledsoe just tiny fractions of a point above neutral for raw team plus minus on the court. The other guys over 20 minutes per game all negative. Starting unit that was good is now terrible. Complete reversal of shot distributions, results or both offense & defense. Lousy defense. What has Coach tried? Analytics suggested? Leadership vetted & endorsed? Not working right now. Parker -2 on RPM impact. Might be lucky it ended up as 4-6. Half game above 8th seed. Is Kidd's time was unacceptable under-performance, what is this?
Re: How the Bucks Use Analytics
2 team level factor strengths are getting to line and forcing turnovers. But own foul rate completely negates the first strength and extra opponent offensive rebounds completely negates the second.
Shot development? Giannis still below 30% from 3 pt land this season and for career, winding the 5th year. Why is he shooting it so badly? Location, contest, footwork, release, time of clock, reaction to scoreboard conditions or what? What do they know and why isn't it helping? Purely talent issue beyond help or improvable with different research and / or guidance?
Giannis started season with long stretch of neutral to negative team raw plus minus then got better then drifted back to it. What happened? What didn't happen to prevent it? Accept it or make further changes in processes, guidance, personnel?
Giannis' career to date: 66 losses below .500 and no playoff series wins. Any recent "superstars" with less team achievement? Who? Why? Did it change? How? Boogie in Sac... trade. Drummond in DET... trade others... then trade him? These seem like closest similar cases. Failures. Pelicans? May have just escaped total failure but the playoffs need a positive mark for it to mean anything. Forgot the Hornets. Very similar. And recently deemed unacceptable. Pretty late. Probably too late.
Shot development? Giannis still below 30% from 3 pt land this season and for career, winding the 5th year. Why is he shooting it so badly? Location, contest, footwork, release, time of clock, reaction to scoreboard conditions or what? What do they know and why isn't it helping? Purely talent issue beyond help or improvable with different research and / or guidance?
Giannis started season with long stretch of neutral to negative team raw plus minus then got better then drifted back to it. What happened? What didn't happen to prevent it? Accept it or make further changes in processes, guidance, personnel?
Giannis' career to date: 66 losses below .500 and no playoff series wins. Any recent "superstars" with less team achievement? Who? Why? Did it change? How? Boogie in Sac... trade. Drummond in DET... trade others... then trade him? These seem like closest similar cases. Failures. Pelicans? May have just escaped total failure but the playoffs need a positive mark for it to mean anything. Forgot the Hornets. Very similar. And recently deemed unacceptable. Pretty late. Probably too late.
Re: How the Bucks Use Analytics
Bucks began Giannis era as a lousy team with no major strengths. Kidd then came in and they were strong and / or lucky at shot defense (strong with Sanders and Pachulia inside and lucky on the 3?) and good at opponent turnovers. Next year good and/ or lucky at own 3s, bad and / or unlucky at 3pt defense. This year they have 2 strengths negated as previously noted. Maybe "better" but not enough.
Re: How the Bucks Use Analytics
There was plenty of meh to negative for Bucks with Kidd but he did have 5 very strong top 5 most used lineups. Since he left? Top lineup negative. 11 of top 20 negative. 6 strong lineups in that mix but not playing near enough minutes. Typical mish mash that looks way distant from potential optimization. A step back probably. And, as almost always, too much use of dink lineups.
Re: How the Bucks Use Analytics
What analytics supported signing Shabazz Muhammad and Brandon Jennings? Might not play much. Probably shouldn't.
Muhammad terrible to extremely terrible on RPM estimates for at least 3 years and probably beyond. But Bucks assumed to not be using RPM.
Jennings was typical backup modest negative in recent years then out of league. Him over somebody else, maybe new? Eh.
Sure, they need more 3 point shooting but they need rebounding as much and a center who can defend without fouling and without getting ignored by the defense.
Muhammad average scoring efficiency, terrible from 3. Jennings average from 3, terrible overall scoring efficiency.
Teletovic was an unwise pricey signing that didn't make sense to me. Guys in the mid price range are usually meh to poor values. All the analytic guys were there. What did they recommend? Didn't carry the day for no...
Skill based evaluations and decision making (especially for 1-2 strong or supposedly decent skill players) often misses the big picture. RPM ignored at your own peril.
Teletovic with bump in individual box score stats in 15-16 but thoroughly average overall for career and with DRPM that more than cancelled out his ORPM. Almost 31 at time. Prime case for overpay. Thanks Bucks / non-RPM staff. Dellevedova in spotlight but average on RPM. Overpay, thanks Bucks and excited eyes. Complete dud. Horrible RPM this season. Not playing because he shouldn't.
Muhammad terrible to extremely terrible on RPM estimates for at least 3 years and probably beyond. But Bucks assumed to not be using RPM.
Jennings was typical backup modest negative in recent years then out of league. Him over somebody else, maybe new? Eh.
Sure, they need more 3 point shooting but they need rebounding as much and a center who can defend without fouling and without getting ignored by the defense.
Muhammad average scoring efficiency, terrible from 3. Jennings average from 3, terrible overall scoring efficiency.
Teletovic was an unwise pricey signing that didn't make sense to me. Guys in the mid price range are usually meh to poor values. All the analytic guys were there. What did they recommend? Didn't carry the day for no...
Skill based evaluations and decision making (especially for 1-2 strong or supposedly decent skill players) often misses the big picture. RPM ignored at your own peril.
Teletovic with bump in individual box score stats in 15-16 but thoroughly average overall for career and with DRPM that more than cancelled out his ORPM. Almost 31 at time. Prime case for overpay. Thanks Bucks / non-RPM staff. Dellevedova in spotlight but average on RPM. Overpay, thanks Bucks and excited eyes. Complete dud. Horrible RPM this season. Not playing because he shouldn't.
Re: How the Bucks Use Analytics
Snell playing bad. Middleton not fitting, not being used correctly or something. Maker finally cut back almost half in minutes. Kinda slow about that.