|
Post by bringbackcaro on Dec 23, 2017 12:33:05 GMT -5
The discussion between Sader1970 and KY in the Siena thread regarding forcing turnovers inspired me to take a closer look at the numbers. Forcing turnovers has arguable been the lone bright spot in the Carmody defense thus far at HC. Despite taking a step back from the #3 overall ranking last year, we are still ranked 89th in the country right now (finished at #80 two years ago). While those numbers appear impressive, to the naked eye (AGENDA! AGENDA! AGENDA!), it appears that we are only able to force turnovers because we very often have 2-3 guys who are fully focused on the ball, which then leaves the defense completely exposed if we are not able to actually force a turnover. But I wanted to take a closer look at the numbers to see if that might actually be the case. I am not sure how to post charts here, but I took a look at the following five KenPom defensive stats to measure the correlation between each and (a) Adj Defense Ranking and (b) Overall Team Ranking: 1) eFG% 2) TO% 3) OReb% 4) 2FG% 5) 3FG% As a starting point to measure the importance of the Adj Defense metric, this has a very strong .734 correlation with Overall Team RankingNow, looking at the correlation between each of those five defensive categories and Adj Defense:
| AdjDef | eFG% | .639 | 2FG% | .545 | 3FG% | .304 | OReb% | .135 | TO% | .055 |
And then looking at the correlation between the same five categories and Overall Team Ranking:
| Team Rank | eFG% | .408 | 2FG% | .355 | 3FG% | .185 | OReb% | ,109 | TO% | .006 |
As these numbers clearly show, there is virtually no direct correlation between forcing turnovers and (a) having a good defense, or (b) having a good team. The other defensive categories (areas where we have been bad-to-awful) are a much better indicator of whether or not a team will be successful. Back to the main question of the thread: What is our formula for success? Do we even have one?
|
|
|
Post by crossbball13 on Dec 24, 2017 19:29:59 GMT -5
“ there is virtually no direct correlation between forcing turnovers and (a) having a good defense, or (b) having a good team”
You’re joking, right. IMHO this could not be more of a perfect example of using statistics so granularly that you lose your logic... missing the forest for the trees or whatever that saying is. Do you really not think turnovers help defenses be good (STRAWMAN STRAWMAN STRAWMAN)? It’s an element to successful defenses, not a one stop shop. Just like making shots is an element of an offense.
I’ll eagerly await your snarky, know-it-all comments
|
|
|
Post by bison137 on Dec 24, 2017 21:06:00 GMT -5
. Do you really not think turnovers help defenses be good (STRAWMAN STRAWMAN STRAWMAN)? It’s an element to successful defenses, not a one stop shop. Just like making shots is an element of an offense. I think it depends on how defenses create turnovers. If it means gambling or often being out of position, then they may not be helpful at all to a successful defense. That is why there is such a low correlation between turnover creation and overall defense. If they happen because of players working hard, establishing good position, etc, then they may help a defense be good. Teams that create a lot of turnovers may also allow a high shooting percentage or a high percentage of offensive rebounds - or commit a lot of fouls. It all depends.
|
|
|
Post by KY Crusader 75 on Dec 24, 2017 21:30:40 GMT -5
Turnovers, turnover percentage , defensive efficiency--guess these are all different items but inter-related in a way, but if it helps the debate, here are the top 5 teams in "turnover margin" and their W-L records to date:
1---West Va 11-1 2- Portland State 10-3 3-- Steve Austin 11-2 4---Saint John's 10-2 5--So Miss 7-6
|
|
|
Post by timholycross on Dec 24, 2017 21:43:44 GMT -5
Didn't Ralph Willard say he wanted a certain number of deflections ...what's the difference between wanting deflections and wanting turnovers? Or maybe some here would disagree that Ralph had a good defensive team.
|
|
|
Post by Sons of Vaval on Dec 24, 2017 21:43:59 GMT -5
Interesting, Kentucky.
Caro, you may want to rethink this thread.
Merry Christmas.
|
|
|
Post by hchoops on Dec 24, 2017 21:52:55 GMT -5
What would you bet on that ?
|
|
|
Post by WorcesterGray on Dec 24, 2017 23:00:54 GMT -5
Didn't Ralph Willard say he wanted a certain number of deflections ...what's the difference between wanting deflections and wanting turnovers? Or maybe some here would disagree that Ralph had a good defensive team. One might argue that Willard's teams were very good defensive teams primarily because they were efficient, kept opponents from making a high percentage of their shots, and rebounded the hell out of the ball to deny opponents extra scoring chances. Their ability to turn opponents over was helpful, but not foundational, to their defensive success. The Bucknell teams from 2010-11 through 2013-14, for example, were excellent defensive teams that were among the worst teams in the country at turning opponents over - didn't matter, because they were efficient, held opponents to low shooting % and rebounded the hell out of the ball. Only four of the top ten teams in "opponent turnovers per possession" last year had winning records. www.teamrankings.com/ncaa-basketball/stat/opponent-turnovers-per-possession?date=2017-04-04
|
|
|
Post by bringbackcaro on Dec 24, 2017 23:27:25 GMT -5
“ there is virtually no direct correlation between forcing turnovers and (a) having a good defense, or (b) having a good team” You’re joking, right. IMHO this could not be more of a perfect example of using statistics so granularly that you lose your logic... missing the forest for the trees or whatever that saying is. Do you really not think turnovers help defenses be good (STRAWMAN STRAWMAN STRAWMAN)? It’s an element to successful defenses, not a one stop shop. Just like making shots is an element of an offense. I’ll eagerly await your snarky, know-it-all comments As the numbers clearly show, it is not a necessary element of successful defenses. The numbers suggest that a defense that consistently creates lower percentage scoring opportunities will perform better than one that gambles to create turnovers and then allows teams to score at will (in other words, exactly what Carmody’s defense does). Forcing turnovers is a great thing. But if that is the only thing your defense can do, your defense is not designed to be successful.
|
|
|
Post by bringbackcaro on Dec 24, 2017 23:31:14 GMT -5
Interesting, Kentucky. Caro, you may want to rethink this thread. Merry Christmas. Merry Christmas to you too. I’m not exactly sure what KY’s point is. Bill Carmody has had a defense in the top 100 a total of 3 times out of the 15 years that KenPom has been tracking data, and the last instance was 11 years ago. We are miles away from have an efficient defensive system.
|
|
|
Post by bison137 on Dec 25, 2017 0:21:55 GMT -5
Turnovers, turnover percentage , defensive efficiency--guess these are all different items but inter-related in a way, but if it helps the debate, here are the top 5 teams in "turnover margin" and their W-L records to date: 1---West Va 11-1 2- Portland State 10-3 3-- Steve Austin 11-2 4---Saint John's 10-2 5--So Miss 7-6 I think you can safely ignore three of the five. SF Austin has played four D3's and overall has one of the five worst schedules in the nation. Portland State overall is below average defensively, allowing a high FG pct and being one of the worst in the nation in defensive rebounding. Southern Mississippi overall ranks #249 in the nation and has played a horribly weak schedule, including four D3's. And St. John's defense and schedule are at best mixed. West Virginia is indeed a very good defensive team and their turnover margin - reflective of their tenacious pressure - is a big factor. When Paulsen was at Bucknell, he had a very firm rule about never trying for a steal. If a player moved out of ideal M2M defensive position to try to get a steal, he was likely to find himself on the bench immediately. In fact, over his full tenure I believe Bucknell was in the bottom five in the nation in steals. And that was despite having some very good, quick defenders. Tyree DPOY, Cohen 3-time DPOY, Muscala DPOY, Kaspar very good defender, etc. Despite the complete absence of steals, they had a very good defensive record - fueled by an overall top 50 in FG pct and an overall top 25 in defensive rebounding. I think he overdid it with the no-steal "rule", but the defense was quite good. Much better than his offense.
|
|
|
Post by hchoops on Dec 25, 2017 8:28:10 GMT -5
What are Paulsen's stats at GUM?
|
|
|
Post by WorcesterGray on Dec 25, 2017 9:06:05 GMT -5
What are Paulsen's stats at GUM? Don't know the more advanced metrics, but here are some relevant GMU rankings, 2016-17. Defensive Rebounding % 19th Defensive EFG % 87th Defensive Efficiency 135th Opp. TO/Poss . 348th Steals/Poss 300th Paulsen implemented his "no steal" rule quickly when he took over. The previous season, GMU had a steal % of 7.1% - in Paulsen's first year (2015-16), that shrank to 3.1%, dead last in D1. Not coincidentally, the defensive rebounding went from 72.8% (92nd) to 79.4% (3rd). Stay at home, stop shooters, clean glass.
|
|
|
Post by KY Crusader 75 on Dec 25, 2017 9:43:23 GMT -5
This has been one of the more interesting discussions we've had--despite the paucity of posts--it really illuminates the tradeoffs you make when you employ different tactics. One other thought--does going for steal cause you to commit more personal fouls???
|
|
|
Post by WorcesterGray on Dec 25, 2017 9:59:13 GMT -5
One other thought--does going for steal cause you to commit more personal fouls??? GMU, starting year before Paulsen PF/Defensive Play. 14-15. 323rd (27.2%) 15-16. 118th 16-17. 37th (20.9%) 17-18. 82nd (21.0%)
|
|
|
Post by WorcesterGray on Dec 25, 2017 10:18:14 GMT -5
Holy Cross under Carmody.
PF/Defensive Play
15-16. 66th (22.2%) 16-17. 113th 17-18. 222nd (23.8%)
|
|
|
Post by WorcesterGray on Dec 25, 2017 10:24:38 GMT -5
. . . here are the top 5 teams in "turnover margin" and their W-L records to date: 1---West Va 11-1 2- Portland State 10-3 3-- Steve Austin 11-2 4---Saint John's 10-2 5--So Miss 7-6 PF/Defensive Play W. Virginia 261st (24.6%) Portland St 327th (26.6%) SF Austin 349th (28.8%) St. John's 234th (24.0) S. Miss. 33rd (19.8%)
|
|
|
Post by Sons of Vaval on Dec 25, 2017 10:25:18 GMT -5
Forcing turnovers or staying back and playing more conservatively on defense isn't really the problem for Holy Cross basketball.
The most alarming statistic is offensive and defensive rebounding, where we rank in the bottom ten in both categories. I don't buy the "we are young" excuse for this shortcoming, as this is simply a flawed strategy from Carmody.
|
|
|
Post by WorcesterGray on Dec 25, 2017 10:31:48 GMT -5
Wouldn't a more conservative defense that stays home around the paint improve the defensive rebounding, reduce personal fouls, and among other things - to beat my favorite dead horse - allow Matt Faw (17.7p and 5.5r per 40) to play significantly more than 11 minutes a game?
|
|
|
Post by WorcesterGray on Dec 25, 2017 10:47:05 GMT -5
Another, maybe counter-intuitive stat . . . all this flying around defensively on the perimeter to try and create turnovers not only weakens the interior defense, but actually seems to encourage teams to shoot threes at very high rates (since there are often wide-open looks)
HC, Opponent three-point rate.
15-16. 327th (40.8%) 16-17. 349th (46.7%) 17-18. 331st (44.8%)
|
|
|
Post by hchoops on Dec 25, 2017 10:58:59 GMT -5
Wouldn't a more conservative defense that stays home around the paint improve the defensive rebounding, reduce personal fouls, and among other things - to beat my favorite dead horse - allow Matt Faw (17.7p and 5.5r per 40) to play significantly more than 11 minutes a game? Matt's limited minutes, alternating at the 5 with Jehyve,may be more due to the offensive end than the defensive side though in the last game we were playing the 4 at the back side of the 1-3-1. The coaches are not going to change defenses now
|
|
|
Post by Sons of Vaval on Dec 25, 2017 11:05:05 GMT -5
All D1 coaches use Ken Pomeroy as a resource -- dvqnced metrics have become such a big part analyzing an opponent. I really do wonder what our staff thinks when they see we are ranked #289 on offense and #296 on defense. How are we going to correct this?
While some of our issues may be attributed to youth, not all of them can be.
We don't rebound. Period.
We don't get to the free throw line. And when we do get there, we are among the worst teams in the country at free throw shooting.
We don't make it difficult for opponents when they are on offense, exhibited by our 2PT and 3PT percentage defense.
We may be looking at a single-digit win season.
|
|
|
Post by lou on Dec 25, 2017 11:14:29 GMT -5
I think rebounding is going to improve (can't get worse?) Jehyve will be among the strongest in the PL. Connor is beginning to assert himself. And Faw appears to me to be more willing to mix it up among the bigs he's competing with. Grandy and KC will be a bonus. And of course Stevens is a "big" unknown
|
|
|
Post by sader1970 on Dec 25, 2017 11:50:26 GMT -5
Being outrebounded 10 a game.
Here's our 2nd (1st?) biggest problem: Our opponents are scoring almost 6 more points a game at the line on average. Considering there is an 8 point total scoring differential per game, that's 3/4s of the problem. For FTs, as always, it's who is taking those FTs. You can have 10 guys on your team shooting 90% but if the guy(s) who take the actual shots are only making 40%, you aren't going to do well.
While KC has taken the most FTs (41) is making a pretty decent 76%, JF has taken the next most (31) and only shooting 42% and PB has also taken 31 FTs and making 68%. Almost as important and I have no stats on it but my observation is that we miss way too many front ends of 1&1s.
As a side note, PB shooting 24% from beyond the arc should have a strob-redlight, CL at 22% is even worse.
As for steals impacting defense, we have a standoff with opponents at 64 total steals for us vs the other guys.
|
|
|
Post by WorcesterGray on Dec 25, 2017 12:06:47 GMT -5
Matt's limited minutes, alternating at the 5 with Jehyve,may be more due to the offensive end than the defensive side Faw's #s and team rank - Pts/40 17.7 (1st) Reb/40 5.5 (2nd) 2FG% 57.7 (2nd) 3FG% 37.5 (2nd) EFG% 57.1 (2nd) FT% 87.5 (1st)
|
|