Untitled Document
Brown Columbia Cornell Dartmouth Harvard Penn Princeton Yale



 Page 8 of 8 « First<5678
Username Post: next year, the deepest of deep dives        (Topic#21367)
palestra38 
Professor
Posts: 32685

Reg: 11-21-04
Re: next year, the deepest of deep dives
03-21-18 01:43 PM - Post#254010    
    In response to Jeff2sf

Matt Howard was a nice player, but let's get real. He never had a positive assist/turnover ratio, he was a mediocre foul shooter, an average rebounder and an OK 3 point shooter his senior year after being horrible prior to that year. He had trouble guarding quick guards and wasn't big enough to guard bigger forwards. He was a good spot up 2 point shooter in an era where that is not that great a stat, especially when he didn't shoot FTs well. He was the best player on a bad team (really AJ was his last year, but Howard was very good that year). We more than replaced him this year. He deserves respect for playing out 4 years under the circumstances and playing hard. Betley, barring injury, will far surpass his numbers over his career.

Now this is not to put down Howard, but Jeff's view is not realistic. We will replace Foreman in much the same way we replaced Howard.

 
pennsive 
Junior
Posts: 200

Reg: 11-21-04
03-21-18 02:43 PM - Post#254015    
    In response to Silver Maple

The Daily Pennsylvanian did a podcast with Tony Price after the Ivy Tourney. In referring to the 1979 team's upset of UNC in the the tourney that year, he expressed the thought that if you believe what the experts say or, extrapolating, what some of the frequent contributors to the Message Board say, about the relative contributions of a certain player, or players, then why bother to play the game? The intangibles, leadership qualities, work ethic, blending one player's moves into movements of a beautifully choreographed assemblage of five players, love for each other, sharing a common goal, and allowing your team to be led by one who refuses to lose, count a whole lot more than the per game stats of individual players that do not adequately take into account these far more important factors. To me, Matt Howard and Darnell Foreman were two of my favorite players of the last decade because of their presence on the court, and how, on their own, for minutes at a time, with the support of their teammates, they sometimes could take over the entire game. They can be followed by other players or combination of players, who may or may not do for Penn what these two did, by they won't be replaced or forgotten by the Quaker faithful. BTW, The City of Basketball Love did a nice article on our new recruit. Perhaps you will pick up the same vibe I did that there may be some similarities in his dedication and in his recruitment that may lead SD to find some Matt/Darnell qualities in this young man as well. Wouldn't that be sweet?!!

 
palestra38 
Professor
Posts: 32685

Reg: 11-21-04
03-21-18 03:03 PM - Post#254016    
    In response to pennsive

Part of the problem of looking at players as if they are represented by a discrete set of statistics is that a discussion on how we replace players (as we must in college basketball every 4 years) becomes viewed as an attack on what went before. No one can take away what Matt Howard and Darnell Foreman brought to the team and program. And Foreman, in particular, had a large role in one of the most enjoyable years in memory. But Tony is right---it's the way the team comes together that determines the ultimate fate of the team, not any one player nor any set of statistics. And let me add, that argument has long been used in the debate (among us old timers) of whether the 1978 team or 1979 team was better. Remember, the '78 team had 1 great and 2 very very good seniors (McDonald, Greene and Crowley) who were replaced by freshmen who did not get extensive playing time. But that team certainly came together to make a historic run.

 
SomeGuy 
Professor
Posts: 6391

Reg: 11-22-04
Re: next year, the deepest of deep dives
03-21-18 03:57 PM - Post#254017    
    In response to palestra38

A couple of stats on Howard as a senior. He had a PER of 20.1. Nobody had a number that high this year. His WS/40 was .155. Only Caleb Wood beat that. His ORAT was 113. Again, only Caleb beat that. He may have been a mediocre rebounder, but his rebound percentage was 13%, which is higher than anyone but Max this year. So statistically no one person came particularly close to replacing him, and his numbers were good enough that I have trouble with the narrative that losing him wasn’t a big deal. But replace him we did. I just think it was very much a group effort.

His defense isn’t going to be something we can quantify. We were a smart defensive team in 16-17 with Howard, and Brodeur and Howard were the guys who really made that go. But we did a lot of 1-3-1 and a lot of different looks that always kind of gave an impression of smoke and mirrors to cover an inability to match up. This year, we mainly just put out our 5 guys who could guard man-to-man and played straight up. Howard guarded bigs more than Woods did, but could he have done any better against Towns? I don’t know that he could. As a team we were obviously demonstrably better defensively this year.

 
SomeGuy 
Professor
Posts: 6391

Reg: 11-22-04
03-21-18 04:05 PM - Post#254019    
    In response to mrjames

Little surprised at how high Columbia is. Dartmouth has 3 guys who got 3 stars on VC, while Columbia only has 2. And you said that you now have Wang top 5, which makes it surprising that Penn’s 2 man class would trail so significantly (unless Ellis is top 5 as well, I guess).

In regard to Yale being overrated, is that Kelly and Cotton being overrated? Or something else?

 
mrjames 
Professor
Posts: 6062

Loc: Montclair, NJ
Reg: 11-21-04
next year, the deepest of deep dives
03-21-18 04:46 PM - Post#254027    
    In response to SomeGuy

So the one thing I should clarify is that the recruiting win shares model is based on NOTHING other than the recruiting rankings. I get a fair deal of anecdotal color that helps (sometimes...) get a sense of whether a prospect is overrated (Yale's 2018 class) or underrated (Desmond Cambridge, et al), but I put none of that into this model. For Maka, he's three star at Rivals, ESPN, 247, 247 Composite, and 4 star at Future150. That's an average of 3.2, which makes him the 13th highest rated consensus recruit in my records. That level triggers a model feature, reserved for the highest-ranked recruit, that leads to higher win share predictions.

Dartmouth's trio aren't rated at Future150 and Taurus isn't rated at Rivals (what an oversight...). All three are still Top 100 of the 635 recruits I have on record, but none triggers that feature that determines the highest win-share boost.

Wang is mentioned but not rated at Future150 and 3-star everywhere else. Thus he falls just short of the highest win-share boost feature.

Here are the Top 10 recruits by team with their consensus average rating and overall rank in the model (fun to notice some of the guys who the buzz was on correctly, but weren't in their school's Top 10 based on recruiting system rankings):

Team Player Year Avg Rank
1 Brown Tucker Halpern 2009 1.75 81
2 Brown Adrian Williams 2007 1.50 96
3 Brown Sam Manhanga 2003 1.50 97
4 Brown Joe Sharkey 2011 1.38 108
5 Brown Travis Fuller 2015 1.33 117
6 Brown Steve Gruber 2006 1.17 134
7 Brown Andrew McCarthy 2009 1.13 153
8 Brown Leland King 2013 1.00 160
9 Brown Christian Gore 2011 1.00 165
10 Brown Marques Coleman 2009 1.00 168

Team Player Year Avg Rank
1 Columbia Maka Ellis 2018 3.20 13
2 Columbia Ben Nwachukwu 2004 3.00 19
3 Columbia Wesley Matthews 2006 2.50 32
4 Columbia Jaron Faulds 2017 2.10 56
5 Columbia Zach Crimmins 2007 2.00 60
6 Columbia Matt Johnson 2008 1.88 70
7 Columbia Arnel Scott 2002 1.67 83
8 Columbia K.J. Matsui 2005 1.50 102
9 Columbia Jason Miller 2005 1.33 118
10 Columbia Justin Armstrong 2004 1.33 120

Team Player Year Avg Rank
1 Cornell Khaliq Gant 2004 2.33 44
2 Cornell Will Scott 2004 2.17 51
3 Cornell Shonn Miller 2011 1.38 110
4 Cornell Galal Cancer 2011 1.13 148
5 Cornell Chris Wroblewski 2008 1.13 150
6 Cornell Jordan Abdur-Ra'oof 2014 1.00 161
7 Cornell Conor Mullen 2005 1.00 178
8 Cornell David Onuorah 2013 0.93 180
9 Cornell Holt Harmon 2012 0.88 182
10 Cornell Errick Peck 2009 0.88 187

Team Player Year Avg Rank
1 Dartmouth Evan Boudreaux 2015 2.83 22
2 Dartmouth Zac Watson 2018 2.40 42
3 Dartmouth Taurus Samuels 2018 1.80 75
4 Dartmouth Elgin Fitzgerald 2006 1.67 85
5 Dartmouth Wes Slajchert 2018 1.50 93
6 Dartmouth Alex Mitola 2012 1.50 94
7 Dartmouth Chuck Flynn 2004 1.50 104
8 Dartmouth David Rufful 2008 1.25 129
9 Dartmouth Ian Carter 2016 1.17 131
10 Dartmouth DeVon Mosley 2005 1.17 135

Team Player Year Avg Rank
1 Harvard Zena Edosomwan 2013 4.14 1
2 Harvard Bryce Aiken 2016 3.83 3
3 Harvard Noah Kirkwood 2018 3.80 4
4 Harvard Spencer Freedman 2018 3.80 5
5 Harvard Chris Lewis 2016 3.67 7
6 Harvard Seth Towns 2016 3.50 8
7 Harvard Wesley Saunders 2011 3.50 10
8 Harvard Robert Baker 2016 3.33 11
9 Harvard Kenyatta Smith 2011 3.25 12
10 Harvard Siyani Chambers 2012 3.00 14

Team Player Year Avg Rank
1 Penn Justin Reilly 2006 3.00 15
2 Penn Ryan Pettinella 2003 3.00 16
3 Penn Steve Danley 2003 3.00 20
4 Penn Jarrod Simmons 2017 2.80 24
5 Penn Patrick Lucas-Perry 2011 2.75 25
6 Penn Matt Howard 2013 2.71 26
7 Penn Cameron Lewis 2005 2.50 33
8 Penn Michael Wang 2018 2.50 34
9 Penn Jamal Lewis 2012 2.50 38
10 Penn Henry Brooks 2011 2.50 39

Team Player Year Avg Rank
1 Princeton Jaelin Llewellyn 2018 4.00 2
2 Princeton Harrison Schaen 2003 3.67 6
3 Princeton Ian Hummer 2009 2.50 41
4 Princeton Alec Brennan 2014 2.33 43
5 Princeton Patrick Saunders 2008 2.25 48
6 Princeton Max Schafer 2003 2.00 59
7 Princeton Devin Cannady 2015 1.92 64
8 Princeton Sebastian Much 2017 1.90 65
9 Princeton Will Barrett 2009 1.88 68
10 Princeton Scott Greenman 2002 1.83 74

Team Player Year Avg Rank
1 Yale Jordan Bruner 2016 3.50 9
2 Yale Sam Kaplan 2003 3.00 17
3 Yale Matt Cotton 2018 2.70 27
4 Yale Ross Morin 2005 2.67 29
5 Yale Isaiah Kelly 2018 2.60 30
6 Yale Austin Morgan 2009 1.88 69
7 Yale Wyatt Yess 2017 1.80 76
8 Yale Jesse Pritchard 2010 1.50 92
9 Yale Alex Zampier 2006 1.50 101
10 Yale Ed White 2005 1.50 103

 
SomeGuy 
Professor
Posts: 6391

Reg: 11-22-04
Re: next year, the deepest of deep dives
03-21-18 06:11 PM - Post#254032    
    In response to mrjames

Interesting. I was kind of down on mcglaughlin’s first class or two,but kind of neat to see they have 3 of their historical top 5 coming this year. Kind of scary that 3 of the top 5 overall are in this class, and that Harvard has 2 of their top 5.

Though that does raise a question— are you sure there isn’t noise here based on changes in how the inputs are calculated. In other words, are there changes to how they rate players now that give recent players deceptively more weight?

 
OldBig5 
Masters Student
Posts: 639

Age: 66
Reg: 02-18-18
03-21-18 07:17 PM - Post#254033    
    In response to SomeGuy

I said it earlier and I will say it again, the ranking of recruits does not really factor in their defensive skills and desire. It's mostly built on offensive abilities/stats and athleticism.

If it were up to recruit rankings then UNC and Duke would have finished way ahead of Virginia this year in the ACC.

Also, some guys just don't get much better and some make big leaps in college.

 
mrjames 
Professor
Posts: 6062

Loc: Montclair, NJ
Reg: 11-21-04
Re: next year, the deepest of deep dives
03-21-18 07:33 PM - Post#254034    
    In response to SomeGuy

I've looked at this before, and where I think you see the most effect is in the number of recruits that are rated now versus in the early years, but not necessarily the max rating a recruit can receive.

In 2003, we had four 3 star average recruits or better enter the league. That's 20% of all recruits in the model. At the same time, there were 16 recruits in that class that got absolutely zero rating out of the 115 overall. That's about 14%.

The recent classes have had 5 zeroes in 2018 (could decrease), 2 in 2017, 1 in 2016 and 6 in 2015. Those four classes have contributed 9 of the 20 3+ star average recruits.

But the median recruit ranking hasn't really changed. Here's the median by year:

Group.1 x
1 2002 0.1666667
2 2003 0.1666667
3 2004 0.3333334
4 2005 0.5000001
5 2006 0.3333334
6 2007 0.5000001
7 2008 0.3750001
8 2009 0.5000001
9 2010 0.5000001
10 2011 0.6250001
11 2012 0.6250001
12 2013 0.5714288
13 2014 0.1666667
14 2015 0.2500001
15 2016 0.4166668
16 2017 0.4000001
17 2018 0.4000001

So, while there seem to be a lot fewer recruits getting zero, the shift seems to be that now those recruits are getting a star or two here and there.

The other thing to note is that while we have a LOT more recruits at the top now, the league is also a lot, lot better now. If the recruiting rankings improved a ton at the top and the league didn't improve, that would lead me to believe something happened to the rankings themselves.

It's really hard to parse out what is ranking inflation and what is just signal of an improving league, but I tend to believe it's really the latter effect we're observing.

 
penn nation 
Professor
Posts: 21086

Reg: 12-02-04
Re: next year, the deepest of deep dives
03-21-18 07:41 PM - Post#254035    
    In response to mrjames

Out of curiosity, what has been the average probability of zero star recruits panning out?

And separate but obviously related--are the zero star folks in there in the first place because, and only because, of the AI?

 
mrjames 
Professor
Posts: 6062

Loc: Montclair, NJ
Reg: 11-21-04
Re: next year, the deepest of deep dives
03-21-18 08:04 PM - Post#254036    
    In response to penn nation

I generally have used 5 win shares as a solid career marker. Given that here are the odds by average rating:

Rating of zero:
94 under 5
8 over

Rating above zero and 1 or less:
252 under 5
63 over 5

Rating over 1 and 2 or less:
73 under 5
38 over 5

Rating over 2 and 3 or less:
24 under 5
12 over 5

Rating over 3:
8 under 5
7 over 5


 
penn nation 
Professor
Posts: 21086

Reg: 12-02-04
Re: next year, the deepest of deep dives
03-21-18 08:13 PM - Post#254038    
    In response to mrjames

The most fascinating part about that breakout is the seeming irrelevance of rankings ranging from more than 1 star through 3 stars.

  • mrjames Said:
I generally have used 5 win shares as a solid career marker. Given that here are the odds by average rating:

Rating of zero:
94 under 5
8 over

Rating above zero and 1 or less:
252 under 5
63 over 5

Rating over 1 and 2 or less:
73 under 5
38 over 5

Rating over 2 and 3 or less:
24 under 5
12 over 5

Rating over 3:
8 under 5
7 over 5





 
mrjames 
Professor
Posts: 6062

Loc: Montclair, NJ
Reg: 11-21-04
Re: next year, the deepest of deep dives
03-21-18 08:29 PM - Post#254039    
    In response to penn nation

Yep - when I first wrote about this many years back now, that 1-to-3 star phenomenon was one of the more interesting things that stood out. There are essentially four classes of recruits:

1) completely unmentioned (not good)
2) at least mentioned somewhere, but not rated highly anywhere (some chance, but not great odds)
3) solid D1 recruits with a couple stars multiple places, maybe even a three star here and there (decent shot of being productive)
4) consensus 3-star or better (really good odds of being a highly productive player in our league)

Most of our recruits historically have been in bunch number two.

 
Jeff2sf 
Postdoc
Posts: 4466

Reg: 11-22-04
Re: next year, the deepest of deep dives
03-23-18 09:06 AM - Post#254200    
    In response to SomeGuy

  • SomeGuy Said:
Well, I think playing 1 big or 2 is often driven by whether the other team is doing it. That’s what I mean by matchups. Out of conference we see more teams with 2 bigs, so you see more Max (and Jarrod) and more games where we seldom if ever play 4 guards.

In conference, a lot of teams either play 4 guards or at least play stretch 4s or undersized 4s. So we played our small lineup more. Kansas did the same, so again we played small.

This dynamic wasn’t just Penn’s big freshman. Yess and DeWolf seemed to get less time as well as the Ivy season went on and their teams went to smaller lineups.



Someguy it took me a while to think about this (ok I got distracted on something else)...

...the implication here is that we always let teams dictate the matchup.

Out of conference, Penn's an underdog, we play two bigs to try to keep up. In conference we're the favorites and we try to play 4 out in order to keep up too?

And Yale benches its bigs in order to keep up with us who benches our bigs in order to keep up with Yale? seems circular.

Shouldn't we be the ones dictating how we wish to play in some situations? If we think two bigs is the way to go, let's play that way against everybody. If we think 4 guards is the way to go, let's let that be our default setting. Either way though, there has to be our default setting, our version of playing "straight up".

Maybe you could say that our preferred way of playing is 4 out but we get hammered on the boards too much against non-conf teams so we play Max more there?


 
SomeGuy 
Professor
Posts: 6391

Reg: 11-22-04
Re: next year, the deepest of deep dives
03-23-18 03:31 PM - Post#254223    
    In response to Jeff2sf

Every year is different. My theory is that in a lot of cases (but not all) the better you are, the less concerned you are with controlling matchups. I think our 16-17 team wanted (maybe needed) to control matchups by forcing teams to figure out how to match up with a 6’4 power forward. We played a 1-3-1 to force people to deal with the matchup issue.

I think this year with the core 6 (and frankly a weak schedule) we were confident we could matchup with teams straight up. Columbia played small at the guards because it had to. But despite having other options, we actually could play small better than they could.

What made Princeton so tough in 16-17 was their versatility— they had 4 guys who could arguably guard 4 different positions. We weren’t on that same level this year, but we were the closest to that. We didn’t generally need to control matchups because of the defensive versatility of the core group.

 
weinhauers_ghost 
Postdoc
Posts: 2125

Age: 64
Loc: New York City
Reg: 12-14-09
Re: next year, the deepest of deep dives
03-23-18 03:52 PM - Post#254225    
    In response to SomeGuy

Big Five note: LaSalle and Giannini are parting ways.

 
palestra38 
Professor
Posts: 32685

Reg: 11-21-04
Re: next year, the deepest of deep dives
03-23-18 04:11 PM - Post#254228    
    In response to weinhauers_ghost


Not surprised, as LaSalle, since the Sweet 16, has consistently underachieved its talent level. But LaSalle is a tough place to compete and win. In my view, they never should have left the MAAC---which consists of schools very much like LaSalle.

 
Jeff2sf 
Postdoc
Posts: 4466

Reg: 11-22-04
03-23-18 04:47 PM - Post#254231    
    In response to palestra38

wait so what is straight up in your mind, Someguy?

 
 Page 8 of 8 « First<5678
Icon Legend Permissions Topic Options
Report Post

Quote Post

Quick Reply

Print Topic

Email Topic

12092 Views





Copyright © 2004-2012 Basketball U. Terms of Use for our Site and Privacy Policy are applicable to you. All rights reserved.
Basketball U. and its subsidiaries are not affiliated in any way with any NCAA athletic conference or member institution.
FusionBB™ Version 2.1 | ©2003-2007 InteractivePHP, Inc.
Execution time: 0.182 seconds.   Total Queries: 16   Zlib Compression is on.
All times are (GMT -0500) Eastern. Current time is 05:23 PM
Top