OFFSEASON FRANALYSIS: A LOOK AT BIG TEN BASKETBALL RECRUITING STARS - PART 3

Now that we've covered the general expectations for a recruit based on their Rivals star ratings, I want to go back and do something similar to what I did in the first post of this series. I was curious to see which teams had been able to get the most and the least out of their respective recruits, so I calculated a few more sets of numbers in order to try and identify teams that were successful in the last eight seasons and teams that were not. I think we already got a general idea of this when we looked at winning percentages vs. the recruiting makeup of each team in the first post, but I wanted to get a little more specific.

In general, I wasn't expecting anything earth-shattering from this exercise. Rather, I went into this thinking that most teams would come out very close to the Big Ten average when it came to player performance/development. This is because I think we have a large enough sample (although a larger sample would be better) and because if a coach was really that bad at player development or scouting, or whatever it is that makes players overperform or underperform, they would likely not be coaching in the Big Ten. Or not for very long at least.

Armed with those low expectations in hand, and with the obvious anticipation of Bo Ryan continuing to look like a damn wizard, I looked at each team's various rated players in aggregate. I looked at three things when attempting to see who was getting the most and who was getting the least out of their players based on the expectations of their star ratings. First, I looked at the average win share total for the different star ratings for each team. Here are those results:

Avg. WS	Average 0 Star	Average 2 Star	Average 3 Star	Average 4 Star	Average 5 Star
Illinois	1.5	2.3	1.7	2.3	N/A
Indiana	0.4	N/A	1.6	1.5	5.4
Iowa	0.1	0.3	1.9	2.1	N/A
Michigan	0.2	N/A	1.8	3.2	3.7
Michigan State	0.2	1.8	1.7	2.5	3.5
Minnesota	0.9	N/A	2.1	2.3	N/A
Nebraska*	0.3	2.0	1.8	0.4	N/A
Northwestern	0.7	1.2	1.8	1.5	N/A
Ohio State	0.5	N/A	1.9	3.0	5.6
Penn State	0.9	1.4	1.8	N/A	N/A
Purdue	0.4	1.7	1.6	2.9	N/A
Wisconsin	0.2	1.6	2.7	3.3	5.1
Average	0.5	1.4	1.9	2.6	4.6

Before I start getting into the details, allow me to present the other two things I looked at.

Because a team's average win shares could be pulled up or down based on a small sample size (Nebraska and Northwestern have just one season played by a 4 star recruit, for example) or because of a crazy outlier or two, I next decided to calculate each team's average standard deviation from the mean. This would give us another idea of how often the type of recruit is over or underperforming for each team.

Std. Dev. Mean	0 Star	2 Star	3 Star	4 Star	5 Star
Illinois	0.8	0.8	-0.2	-0.2	N/A
Indiana	-0.2	N/A	-0.1	-0.6	0.5
Iowa	-0.6	-0.9	0.1	-0.2	N/A
Michigan	-0.4	N/A	0.0	0.6	-0.6
Michigan State	-0.5	0.1	-0.1	-0.2	-0.8
Minnesota	0.6	N/A	0.1	-0.3	N/A
Nebraska*	-0.3	0.2	-0.1	-2.3	N/A
Northwestern	0.1	0.0	0.0	0.0	N/A
Ohio State	-0.2	N/A	0.0	0.2	0.6
Penn State	0.5	0.1	-0.1	N/A	N/A
Purdue	-0.2	0.0	-0.1	0.3	N/A
Wisconsin	-0.4	0.3	0.4	0.1	0.2
Average	-0.06	0.1	-0.02	-0.3	0

This measure too could also be a little skewed depending on how many outliers a team has had or due to a small sample size, so lastly I looked at what I call the "Success Rate" of each team.

Since I had already calculated the average win shares for the various stars based on which class they were in, I decided to go through and see how often a player had a "successful season" for their team. For instance, the average win share for a 3 star freshman is 1, while the average win share for a 4 star senior is 4.1. Thus, with success rate, I wanted to see how many instances each school had of a player meeting or exceeding the average win share for their star and class.

Success Rate	0 Star	2 Star	3 Star	4 Star	5 Star
Illinois	58.3%	66.7%	33.3%	37.1%	N/A
Indiana	21.1%	N/A	41.7%	30.4%	55.6%
Iowa	0.0%	0.0%	46.4%	25.0%	N/A
Michigan	17.6%	N/A	43.9%	66.7%	25.0%
Michigan State	0.0%	50.0%	36.1%	41.0%	16.7%
Minnesota	72.2%	N/A	53.1%	36.4%	N/A
Nebraska*	31.3%	66.7%	37.5%	0.0%	N/A
Northwestern	41.7%	50.0%	48.1%	100.0%	N/A
Ohio State	50.0%	N/A	30.3%	54.3%	75.0%
Penn State	52.0%	42.9%	36.2%	N/A	N/A
Purdue	25.0%	50.0%	35.1%	56.4%	N/A
Wisconsin	10.0%	50.0%	53.2%	66.7%	66.7%
Average	32.2%	42.9%	42.7%	47.6%	44.9%

One issue with success rate is that if a player's win share total was even 0.1 short of the average win share for the season, I counted it as an unsuccessful season. Normally I wouldn't consider that an unsuccessful season, as win shares aren't accurate to the point of 0.1. However, in order to avoid having to decide whether 0.2 below or 0.3 below or whatever was a decent cutoff for a successful season I just made the Big Ten average for the last eight seasons the hard and fast cut off. It's not perfect, but it is what it is.

Additionally, you will notice that the Big Ten average success rate for each category never reaches 50%. That's because there are extreme seasons in every category that pull the average up from where it would likely be if we had a larger sample. Essentially, what I'm saying is seasons like Frank Kaminsky's senior year or Trey Burke's sophomore year set the bar ridiculously high for 3 star players. I'm also reiterating that I would ideally like a bigger sample than just eight seasons.

Now that I've laid out the tables and the limitations, here are some observations.

Bo Ryan...

Let's just get this out of the way right now, Bo Ryan is not human. I hate heaping praise on the man, but he really is an evil basketball genius. Over the last eight seasons he has seen every player category except 0 star recruits come out above the Big Ten average. His most impressive category is probably the 3 star recruits, in which they average 2.7 win shares per season compared to the 1.9 Big Ten average. That's an average standard deviation of 0.4 above the mean, which is the highest among the 3 star category and one of the highest out of all the categories. His 3 star players have also tallied a successful season 53% of the time in the last eight years.

And if Bo Ryan's coaching abilities with 3 star guys wasn't enough, he's pretty damn good with the 4 star players he lands too. While the Big Ten average is 2.6 win shares per season, Wisconsin's 4 star guys tend to average 3.3 per season and 14 out of their 21 seasons played by 4 star guys have been successful. That's a 67% success rate.

Finally, he's also done just fine with the 5 star guys that he has been able to get to come to Wisconsin. Out of the six seasons Wisconsin has had a 5 star player tally minutes for them in the last eight years, four of them have been successful and the Badgers have gotten an average of 5.1 win shares out of them, compared to the 4.6 Big Ten average.

Scouting, player development, and coaching. Bo Ryan has it all. And that is the last time I'm going to say anything nice about him in this post. I feel so dirty. (/immediately takes a boiling hot shower)

Indiana's 5 Stars Have Been Good and That's About it

Tom Crean's recruiting abilities are not in doubt. It's the whole player development and coaching thing that is. The 5 star recruits that Crean has gotten to suit up in the candy stripes have been great for him and the school. With 5.4 win shares per season, the 5 star players are not the problem.

It's the other guys, Indiana's 3 and 4 star players, who have not lived up to the expectations of their given stars. The average Big Ten 4 star recruit has averaged 2.6 win shares over the last eight seasons, but Indiana's have only tallied 1.5. Out of the list of 4 star players for the Hoosiers in the last eight years, only Christian Watford consistently produced successful seasons at Indiana. But, for every Christian Watford, there are a handful of Hanner Mosquera-Pereas and Maurice Creeks. And there are also a decent amount of three star recruits who have also underperformed.

The good news, however, is that 4 star recruit Troy Williams easily outplayed the sophomore average last season and he's back for his junior year surrounded by a bunch of other highly-recruited players. Maybe this is the year?

Thad Matta's a Good Coach

Outside of last year -- with the zone defense debacle and fact that Ohio State had 91% of their minutes come from 4 and 5 star guys last year and failed to surpass the 70% win mark for the first time in forever -- Thad Matta's numbers seem to back the reputation that he's a good coach. And, in all fairness, last year's squad had a short bench, still won 69% of their games, and made the round of 32.

But over the last eight seasons, Matta has been one of the best at getting the most out of his 4 and 5 stars in the conference. He's even been right around the conference norm when it comes to the production of his 3 star guys. And while a constant flow of highly-rated guys like Jared Sullinger, D'Angelo Russell, Deshaun Thomas, Evan Turner, and Jon Diebler into the program are why Ohio State is consistently one of the best in the league, Matta's coaching abilities should not be discounted either. Because the three seasons prior to last year, his team minutes played by 4 and 5 stars consisted of 64% and 57% and they still won over 70% of their games and almost hit 80% twice. Those highly recruited guys are always important, but Matta's ability to scout and develop guys like Aaron Craft and Lenzelle Smith Jr. is also important.

Iowa has Turned the Program Around With 3 Star Guys

As Hawkeye fans, we were already aware of this. But I thought I would still point it out and also point out the difference between McCaffery and Lickliter. Even with a couple of Lickliter's seasons mixed into this data, Iowa's 3 star average, standard deviation and success rate has been above average thanks to his predecessor. But if we look at McCaffery alone in comparison to Lickliter's last two seasons at the helm, the differences are truly stark.

Coach	Avg. STD from Mean	Average 3 Star	Success Yes	Success No	Success%
Lickliter	-0.2	1.5	10	16	38.5%
McCaffery	0.2	2.2	22	21	51.2%
Big Ten Average	0.0	1.9	237	318	42.7%

Of course, the differences aren't that surprising, considering a team's total win shares should add up to approximately their season win total, give or take some. But it's still interesting to note that 3 star players under McCaffery have averaged 2.2 win shares per year compared to just 1.5 under Lickliter in his last two years. And their success rate has gone from 39% to 51%. Whether it's scouting and development-related or just because McCaffery is actually able to keep his best players from transferring every season, it doesn't matter. Whatever it is, McCaffery has clearly been an upgrade for Iowa.

Purdue and Illinois are Similar yet Different

Purdue and Illinois are two programs that are able to consistently get 4 star talent on their roster, but that hasn't necessarily translated into a lot of success in recent years. The first thing you would think would be the problem is that their 4 star recruits just aren't panning out, right? Well, not necessarily.

The 4 star talent that the Boilermakers have had over the last eight years have been great. Robbie Hummel, E'Twaun Moore, and JaJuan Johnson were a deadly trio that made Purdue one of the better teams in the nation. Once they graduated, though, Purdue was in a rebuilding mode and they haven't been the same since. They restocked the cabinet with more 4 star guys, but like we saw in the second post of this series, 4 star guys still generally take a bit of time to develop. And the above mentioned trio were actually very, very good college basketball players. For example, two of JaJuan Johnson's four seasons were the top two seasons by a 4 star player in the last eight years within the conference. And when you add a guy like that to Robbie Hummel, who was a consistent 5 win player throughout his entire career, and toss in E'Twaun Moore, who wasn't far from Hummel's win share totals, you can easily see that Purdue had three truly special players.

Guys like Rapheal Davis and A.J. Hammons have been good over their first three years on campus, but they have not had quite the production that guys like Johnson, Moore, or Hummel had. That leaves Purdue as an interesting example of just how important 3 star recruits are when your 4 stars aren't instant impact type of players early on in their careers. Unfortunately, for Purdue, their 3 star recruits haven't really done much lately.

That brings me to Illinois, who in addition to getting, on average, below standard production from their 3 star players, haven't been getting much out of their 4 star players lately either. The Illini seem to be on a run where they are having a hard time getting all their 4 stars to develop and be on the team at the same time. Demetri McCamey was very good, but he really had little help. Brandon Paul had his best year as a senior, but that season sunk before it started when Meyers Leonard jumped ship to the NBA after his sophomore season, leaving a 7 foot hole on the Illini roster.

If there is any hope, however, they do keep getting more highly-ranked guys to sign letters of intent to play for them. 4 star guys like Malcolm Hill, Kendrick Nunn, and Leron Black are solid guys to build a program around. And they've got two more 4 star guards in Jalen Coleman-Lands and Aaron Jordan entering the fold this year. If they can get everyone to stay on the roster and not leave early for any reason, this team could have a bright future.

Michigan's Found Success with Recruiting and Scouting

On the opposite end of Purdue and Illinois, Michigan has succeeded with 4 star recruits lately.

In Ann Arbor, John Beilein has had a lot of success in recent years loading the program with 4 star players like Nik Stauskas and Mitch McGary, a 5 star guy in Glenn Robinson III, while also striking gold on a 3 star recruits like Trey Burke, Tim Hardaway Jr., and Caris Levert.

The bad 5 star ratings for Beilein are a small sample issue, so they should be taken with a grain of salt. Basically, the four seasons between Robinson III and Zak Irvin are the only 5 star seasons Michigan has seen over the last eight years. However, while Robinson III was a great player, Zak Irvin has not quite lived up to the instant impact expected of a 5 star guy out of high school. He's been a very good player and he looks primed for a big junior season, but so far he has been below the average Big Ten 5 star recruit over his career.

Instead it's been 4 star guys like Stauskas and McGary who lived up to their billings as major college basketball players and it's also been John Beilein's eye for talent and knack for development that has turned the Michigan basketball program into a potential powerhouse. His ability to get big time recruits onto Michigan's campus and his ability to take less-heralded guys like Trey Burke, Tim Hardaway Jr., and Caris Levert and make them into top Big Ten players has been invaluable.

Michigan under Beilein is my hope for what Iowa can eventually get to under McCaffery. I'm not saying McCaffery is as good at the X's and O's as Beilein is, but his scouting and player development track record thus far is comparable. If they can get a couple of 4 stars who can come in and make an impact, and then develop a couple of 3 star players into something special, Iowa too could be a potential challenger for a Big Ten title most years.

Michigan State is Weird

Tom Izzo's 4 and 5 star recruits have been a little underwhelming lately. Over the last eight years, only Gary Harris as a sophomore and Adreian Payne as a junior were the only 5 star recruits to hit the 5 win share mark for a season. Their success rate for 5 star players has only been 17% since 2007-2008 and their 4 star players have had a successful season only 41% of the time compared to the Big Ten average of 48%.

Now, it's not like Michigan State's highly-rated recruits have been terrible basketball players. Even if a 5 star player isn't quite living up to the expectations of a blue chip recruit, they are still providing 3-4 win shares on a regular basis, and Izzo's 4 stars are still giving him 2-3 win shares on top of that. That is big production for a team. My theory for their underperforming numbers is that it is more a case of Izzo's squads actually practicing team basketball and less of a case of their players just not being as good as advertised. When you have a team full of players who can give you anywhere from 2-4 win shares, on average, and have a few guys break out for a 5-6 win season, your team is going to be good.

It also doesn't hurt to find a few 3 stars who can explode for 6-7 win shares as seniors, like Draymond Green or Travis Trice.

Nebraska, Northwestern, and Penn State Show that Recruiting is Important

I won't go into depth here, but you can see that the bottom three teams of the conference really aren't that far below average when it comes to developing their players. In some cases, they are above average, like Northwestern's success rate for every type of recruit they put on the court. Of course, when your roster is made up lightly-starred guys, their ceilings are going to be much lower. So while they may reach or even surpass those ceilings, on average, unless they are doing so by a Frank Kaminsky-like margin on a regular basis, the team's potential for winning is limited.

So that's that. Hopefully you've learned something interesting over this three part series. I don't think it taught us anything that we didn't already necessarily know. We knew recruiting mattered, but we also knew that it could be overcome with good scouting, player development, and coaching. I do, however, hope this series has been able lend some concrete examples to help give us more accurate expectations for players on the current roster and for incoming and future recruits.