Three Point Musings

novak-3 douglass-3 manny-3

Villanova by the Numbers has been investigating the effect that moving the three point line back a foot had on last year’s college basketball season. Nova by the Numbers focused on how Luke Winn’s predictions from a little over a year ago look in retrospect. Winn’s hypothesis was that the biggest winners from the move would be teams that don’t rely on the three point shot or defend the three point line well. The losers would be perimeter oriented teams that shoot a lot of threes and teams that make a low percentage of their three point shots in general.

The hypothesis makes so much sense that it almost seems rudimentary. Judging by Michigan’s 2007-2008 statistics, they should have been negatively affected by the move (and that’s probably an understatement). The 2007-2008 Michigan team was one of the worst in the country at making (31.2%) and defending (38.1%) three point shots yet they still shot 40.7% of their field goal attempts from behind the arc.

So how did Michigan deal with the move? They shot even more three point shots at a slightly higher percentage. (They also made a dramatic improvement in three point defense — their opponents shot 31.1% compared to 38.1% in 2008.)

Michigan kept on chucking and somehow found success, going from 10 to 21 wins in only a season.  The three point shot is obviously a staple of the Beilein offense.  Therefore it’s not surprising that Michigan was more reliant on the three point shot than any other NCAA tournament team.

Top 10 3FGA/FGA for NCAA Tournament Teams (National Rank to the left, NCAA seed in brackets)

  7 Michigan[10] (B10)               47.9
 11 Arizona St.[6] (P10)             45.3
 13 Butler[9] (Horz)                 45.2
 15 Portland St.[13] (BSky)          44.8
 26 Oklahoma St.[8] (B12)            41.3
 29 Chattanooga[16] (SC)             41.2
 38 Akron[13] (MAC)                  40.0
 40 Mississippi St.[13] (SEC)        39.7
 54 Utah[5] (MWC)                    38.4
 58 American[14] (Pat)               38.1

That list is an eye opener. There are very few high major teams in that group, only Arizona St., Oklahoma St., and Mississippi St, and none of the top ten made it passed the second round.

This clearly isn’t a strategy that is replicated often with success at the high major level. Of those four high major teams, Michigan’s 33.4% three point shooting percentage was the worst among the group by a pretty significant margin. It’s clear that perimeter oriented teams were adversely affected by the three point move.

beileinNova By the Numbers points out that three of the ten interior oriented offenses that Winn identified before the season ended up in the Final Four (UNC, UConn, and Michigan State). Perimeter oriented offenses didn’t perform so well but I’m not sure John Beilein is too concerned as his tournament success is well documented. Despite the success of interior oriented teams, John Beilein’s teams still perform better against their seed than any other college coach.

Michigan’s three point shooting numbers are even more shocking when you realize that Manny Harris and DeShawn Sims combined to take 64% of Michigan’s 2 point field goals. Stu Douglass (80.3%) and Zack Novak (77.8%) took the highest percentage of their shots from long range on the team with more than 3 out of every 4 field goal attempts coming from behind the arc. Seven players on the Michigan roster took 65% or more of their field goal attempts from three point range. Only Jevohn Shepherd, Zack Gibson, DeShawn Sims, and Manny Harris took more than half of their shots inside the arc.

I’ve mentioned high variance strategies in the past (explained here) and there is no doubt that relying on the three point shot qualifies as a high variance strategy. To most in the college basketball world, three point shooting is typically thought of as an underdog  strategy marred with desperation. The common thought is that truly great teams will pound the ball inside rather than shoot away.  However, John Beilein’s best teams actually appear to shoot more three point shots than his teams that struggle.

image

 Michigan      2009 21-14 (7)      47.9
 Michigan      2008 10-22 (42)     40.7
 West Virginia 2007 27-9  (5)      49.0
 West Virginia 2006 22-11 (2)      51.4
 West Virginia 2005 24-11 (7)      45.9
 West Virginia 2004 17-14 (25)     40.1
 West Virginia 2003 14-15          37.6

John Beilein coached teams play their best when they favor a strategy that is generally frowned upon in the basketball world. The numbers tell the whole story last year: Michigan succeeded by shooting a lot of threes (7th) even while making a low percentage (195th).

The question is what happens going forward — does Michigan continue to shoot so many three point shots or do they revert toward a more traditional style of play. Looking at John Beilein’s history, it is hard to imagine Michigan relying on the three ball any less than they did this year.

dmo driveThe first place potential change would come from would be the incoming freshmen class. Of the four scholarship players and two walk-ons, three players would be described as shooters first and foremost (Vogrich, McLimans, and Bartelstein). Jordan Morgan gives Michigan a big guy that they were lacking last year, but after knee surgery his impact is questionable. Eso Akunne is without a position, but he possesses a well rounded versatile game.

The one guy who has to potential to redefine the offense is Darius Morris. Darius has the ability to create for himself and others in a way that none of the point guards on last year’s roster could. If Darius can provide a threat to drive and score it would go a long ways toward making Michigan’s offense multi-dimensional and giving Michigan three guys who are a threat to score inside the arc rather than just two.

However, relying on a freshman point guard to reshape an offense is probably too much to ask for. I wouldn’t be surprised to see Michigan shoot over 50% of their shots from long range next year but they need to improve their shooting percentage. When you shoot more three point shots than just about every other high major team in the country, you’re going to want to rank higher than 195th in percentage made.

The three point shooting issue is one that has confronted Beilein teams for years and there is no firm answer about what is right and what is wrong. For example, the three point shot was the reason Michigan trailed by 20 points versus Indiana at Assembly Hall (3-18 first half) but it was also what enabled the comeback (7-17 second half). The success that Michigan had last year despite not shooting a very high percentage says one of two things: the strategy really works or Michigan got very lucky.

The topics for discussion on this subject are endless, so let’s hear your thoughts in the comments and then maybe we can revisit this subject in a couple weeks with some follow up.

  • http://umhoops.com Mattski

    Really interesting post. Two quick initial observations: first, the effects of moving back the line should be strongest in the first year and wane as time goes on, as teams adjust. So retrospectively, Michigan weathered the first stages of that storm nicely.

    Second, Beilein’s basic strategy stays the same–his career is built around it–but as he gets better players, inside and outside percentages should rise, and lots more high-percentage close-range shots become available as players weave faster and more adroitly, back-door, etc. Add a better mid-range game for players like Manny, as I noted before, and Beilein might really be able to see his philosophy in action fully fleshed out as never before.

    After that, I assume, he continues adjusting for the new circumstances; that too will be interesting to watch.

  • Tom Too

    Mattski – I agree totally. I was in shock most of the season at the number of threes they shot. Normally JB’s teams take a little more pull-up jumpers, tear-drops, and back doors than the 2008-2009 team. You can literally count on both hands the number of successful backdoor plays last year and those backdoors are why people put Beilein’s offense in the same bucket as Princeton’s…don’t remember Princeton shooting this many threes.

    I think as this system develops, and you get quality big men here, you are going to see a more balanced attack.

  • Avery Queen

    Agree with the previous two posts.

    Also, while I can’t argue that 3 pt shooting isn’t inherently high variance, the fact that you have a high number of players capable of hitting the 3 helps hedge against that variance a bit. If Stu is cold then LLP (in theory) could step up or Novak or Manny. I think we saw this a bit from JB often going with the hot hand. Of course there will be some games where everyone is off but at least it’s not like relying on a single shooter.

    That being said, I would also think the 1-3-1 is relatively high variance because it is high risk/reward. We got burned on quite a few easy dunks and wide open threes but also forced a lot of turnovers.

  • benley

    I agree Tom Too. If you look at the progression of the Michigan team during the season, the back door cut specifically with the entry pass from the top of the key constantly caught teams off guard in the start of the season. (see UCLA upset and some of the Duke game) But, as the season wore on, that first option was shut down and it put more pressure on our three point shooting and Manny’s ability to create. This adjustment accounted for the hardship found in the middle of the big ten grind. I mean, you can only do so much when you are starting a freshmen shooting guard at the four spot. A few big men and some high level recruits, we might see Belein’s offense in full force. If he had realized all of the success and depth out of his system in WVU, it doesnt necessarily make sense to jump ship to UM. I believe that we have some exciting and competitive years to come from UM hoops

  • GregGoBlue

    Avery Queen, I’d be interested to see the individual stats when the three point shooting is cold. I felt at times last year (specifically Indiana) that when one person couldn’t buy a bucket, the whole team had a cold spell.

    I suppose in theory that having more players capable of hitting the three is better in that you have more options, however there is an argument to be made that one only has a limited number of offensive possessions. We saw many a possession fade away last year as Manny (among others) would pull up and clang a three for a few possessions in a row. Having more threats I don’t think can ever be a bad thing, though. Having more options when the three isn’t working is something that we are acquiring with more size and more slashing creators.

    It’s interesting to see how Beilein’s shooting philosophy, coupled with Michigan’s ability to attract higher-level players, will play out in recruiting in the future. Will we go with a bunch of huge guys with a stroke? Will Beilein’s philosophy change with his ability to recruit higher level talent? Let me know what y’all think.

  • UMDC

    It’s not clear to me how having “Top 10 3FGA/FGA for NCAA Tournament Teams” from last year helps determine how teams were affected by the change in 3pt line.

    In order to determine this one would at a minimum need the “Top 10 3FGA/FGA for NCAA Tournament Teams” from the year prior to the change too. Then we could at least compare before and after. Just having after leads to the potentially spurious conclusion that teams were adversely affected by the change when it is possible that the prior values are similar the post values. Which, of course, would suggest that teams were not affected by the change at all.

  • http://www.umhoops.com Dylan

    UMDC: Last year’s #s are listed in the original linked post.
    http://vbtn.blogspot.com/2009/06/3-point-line-how-winn-called-offense.html

  • UMDC

    yeah, I checked out the link and I still don’t see anything more than a mediocre case being made that teams were adversely affected by the change. The one concrete stat is that 3pt FG made and attempted went down by about 1% each (.9 and 1.3 respectively).

    It seems to me that, as they point out, the reason the teams that did worse last year was as attributable to other factors (e.g., losing players) as the change in the 3pt line.

  • http://www.umhoops.com Dylan

    The bottom line is that teams that shot a lot of three pointers didn’t do well in the NCAA tourney. I kind of went away from the three point line move as the post goes on and more or less looked at how well three point shooting teams do in general.

  • Jivas

    Nice job on the article – definitely an interesting thing to think about during the off-season.

    Couple of technical points:

    (1) If we’re looking for a metric that shows reliance on the three-point shot, I’d recommend divided by . This factors in team’s reliance on free throws, and in my opinion is a better metric to explain what you’re looking for. (Also, as mentioned by UMDC above, looking at prior years is key – a control group is needed).

    (2) Minor point – I recommend 3PA/FGA on the x-axis and winning percentage on the y-axis, to properly illustrate what you’re indicating to be (potentially) the dependent variable and the independent variable. More than anything, the graph would be more intuitive that way.

    But again – great work! Thanks for the quality offseason read.

  • Jivas

    Hmm…issue with the formatting on my last comment. The metric I’m suggesting in point (1) above is:

    Total points on 3-point shots / Total points

    ——-

    Pardon the inconvenience.

  • http://www.umhoops.com Dylan

    Yes… Three point point percentage is good too and the numbers are very similar. Check out Michigan’s Kenpom profile for some more #s. http://kenpom.com/team.php?team=Michigan

    I love the feedback and criticism too guys, I am by no means a full fledged statistician and love to hear your ideas.

  • Ken in Vegas

    I think if we’re going to depend on independent variables then we need the control group to be inverted on the y-axis with the statistical mean…ok I have no idea what I’m talking about.

  • Ken in Vegas

    I’ll just stick to holding up my foam finger and yelling things.

  • 90 M Varsity

    This is what people need to look at when they get all giddy about these perim players. I think JB and his situation can all work out. I am just surprised that people don’t look deeper than the surface.

  • ToBlav

    The stats and graphs may be interesting but if team A’s stats are built mostly around playing traditional style teams, which makes sense because most teams playing that way is why it is traditional, then those stats will not give a valid picture of their play vs. team B a non-traditional team. If team B’s stats are built around playing teams who’s stat vs. them are not valid for that team the conclusions are going to be skewered. Valid independant varibles are hard to come by when teams have hot and cold and up and down games. Comparing teams with like numbers and percentages of three taken and made is not valid because the outcome of the misses varies with the rebounding styles and defensive reaction of the shooting team. In summation, stats and charts can only be indactors and should not be overly relied on in logical thinking.

  • http://www.umhoops.com Dylan

    Also, you guys are right, the graph should be flipped. My bad.

    **Should be better now.

  • old fan

    My thought was that it would be better for Beilein if the three point mark was moved back. When an average forward from any team can move back to hit the trey, this devalues this as a mechanism to separate yourself from the competition. If you need to have to value and work at getting the three point shot in order to use it effectively, this enhances Michigan’s strengths.

    Ideally a lower percentage , but still a percentage that makes this a definitive weapon, with a wider spread between teams would allow Michigan to have a comparative advantage over other teams. Doesn’t look like the numbers helped validate my assumption, at least not yet. We don’t have the shooters yet to make this hypothesis a reality.

Previous post:

Next post: