"My rule was I wouldn't recruit a kid if he had grass in front of his house.
That's not my world. My world was a cracked sidewalk." —Al McGuire

Wednesday, June 01, 2011

The Value Add Sweet 16 for 2011-12

Now that my spreadsheet of the Statistical Values of all returning players is circulating, others are building on the rankings (Chicago Now ran with them to show the top players for Chicago area teams and New York’s Big Apple Buckets started using them to rank teams in several conferences). This begs the question, “can you add up the Value Adds of all returning players to get a good gauge of how good the overall team will be in 2011-12?

Yes and no. The key factor in doing this is that you can expect a player’s Offensive Value Add to roughly double when he comes back for another year due to improved play, while his Defensive Value Subtract will only improve slightly since 78% of that ranking is based on the team as a whole. If you take that into account and run the numbers for every player, you get the following Sweet 16 for next year:

RnkTeamOffense (X2)DefenseNew (X2)Total
2Duke 22.2%-22.8%17.8%62.7%
3Ohio St. 31.6%-21.9%9.1%62.6%
4Kentucky 23.5%-14.0%21.4%58.9%
5Louisville 22.0%-27.7%8.2%57.9%
6Syracuse 25.5%-19.3%9.1%53.8%
8Connecticut 22.1%-18.7%3.7%44.5%
9Pittsburgh 25.3%-10.2%8.5%44.1%
10Purdue 15.6%-18.0%8.8%42.4%
11Florida St.6.4%-32.4%3.4%42.1%
12Vanderbilt 37.5%-0.2%2.7%40.3%
13Belmont 21.7%-18.3%0.0%40.0%
14Michigan St. 21.7%-12.7%5.5%40.0%
15Nevada Las Vegas 19.4%-18.8%0.0%38.2%
16Cincinnati 19.6%-15.7%1.6%36.9%

The updated 3,140 players (yes, I have now added all players who were sitting out a year to redshirt, transfer or injury to the incoming freshman and returning players thanks to discovering a great site with all college transactions for every team) is available via email by request, but it has become too big to post here.

The two surprises on that list are Alabama, who was left out of the tournament for early season woes last year before going all the way to the NIT finals, and Belmont, who was already in the top 20 for Pomeroy and returns their top 7 contributors for their last year in the ASun before jumping to the OVC.

However, overall I believe it is a pretty good basis for starting a discussion on the top teams next year. BUT, if using it to compare other teams, you need to make subjective adjustments for three potential considerations:

• If you see two teams that total a 0.0 on Defensive Value Sub, look at Pomeroy to compare them from last year. Only 87 teams have players who can erase points at the BCS-level, so while several Stanford players just missed that mark to total a 0.0, they have most players back from a team that gave up 97.6 points per trip compared to Chicago State, which gave up 120.5. So ZEROS aren’t always equal when comparing teams.

• If a team has a glaring weakness at one or more positions, then they are worse that the Value Add will calculate by simply giving them a ZERO rating for that position. For example, my guess is Marquette would have probably rated too high during the final year of the Three Amigos and Lazar because their 0.0 Value calculation at center would reflect that they simply never found a Big East-calibre center and therefore Dwight Burke basically had to fill the role instead of being an excellent back-up power forward behind Lazar.

• When comparing non-BCS conferences, you need to make a distinction based on a player’s rank. Only 373 returning players in the country are good enough to add 2% to 13% to a BCS teams scoring margin, while another 1,653 are good enough to add a marginal 0.01% to 2%. However, while the players ranked 1653 and below are not good enough to add any value if they played at the BCS level, there is a huge difference between their value at a lower D1 level. For example, Kevin Cantinol is ranked 1655 because of the tremendous value his 97.7 ORtg and 106.9 DRtg add to Florida Gulf Coast in ASun games, so he adds a lot more value at the non-BCS level than DeVontae Wright who has a 49.7 ORtg and 115.8 DRtg for The Citadel. They both have a zero Value Add because on a BCS team they’d both be on the bench, but in their conferences there is a huge difference.

Value Add can certainly be a valid tool toward comparing teams for next year, but when comparing lower D1 conference and in some cases even Mid-Majors where most players have ZERO Value figures at the BCS-level, I believe averaging the ranking of the team's top 8 players probably be a good counter balance to a straight reading of the top Value figure in these conferences. To site an example from Big Apple, because Ken Horton could be a dominant player at the BCS level, he could tear it up for Central Connecticut. His Value Add is higher than most NEC teams. However, when you take the average rankings of the top 8 players from each conference team (and add a 3500 for any top 8 players that didn't make the top 3120) to account for the fact that a lot of 0% guys in the NEC are actually adding tremendous value, Long Island actually comes out on top: Long Island 1202; Robert Morris 1356; Wagner 1575; St. Francis PA 1604; Central Connecticut 1767; Sacred Heart 1820; Quinnipiac 1836; St. Francis NY 1870; Monmouth 2010; Mount St. Mary’s 2184; Bryant 2242; Fairleigh Dickinson 2976. The truth is probably somewhere between the two, with Horton's ability to play at another level probably letting them content, but the broader talent of the four teams in front of him being a major counter balance.

Defense will be the key to a return Sweet 16 run

Which leads us back to Marquette. Since I am accused of being too optimistic on Marquette, it was good that I had to crunch the numbers for all the teams to see how I would feel about Marquette’s chances next year if I had never seen them play. When we compare their composite numbers to the Sweet 16 teams above, we get:

TeamOffense (X2)DefenseNew (X2)Total

Offense. When I first ran the 2500 offensive players Value Add’s I was ecstatic. As you can see, Marquette’s 19.7% Value Add is better than several of the “Sweet 16” teams above, Jae Crowder is the 24th best offensive player in the country, DJO is in the top 100 despite the terrible start last year, and with Davante Gardner already in the Top 500 despite limited minutes, MU could be awesome scoring. Vanderbilt actually has the highest Offensive Value Add of any school, so the game at the Bradley Center really looked like a Sweet 16 showdown when I had just calculated offense from returning players.

New Players. The overall Value of players who did not play last year is based on comparably based players, but is not broken down into offense and defense. MUs 1.6% does look that high since Jamil didn’t have big numbers at Oregon and our incoming players weren’t ranked that high in services. However, I am confident the new guys will continue our trend this year of finding and/or developing players from the 3-star Dwyane Wade, to the unknown Jimmy Butler, to the second best guy on his high school team in Lazar Hayward. I am not worried at all that our new players will actually have a much bigger punch than the raw numbers show.

Defense last 4 years: 10th, 51st, 57th, 61st

Which leads us with the big question for next year – will Marquette have enough defense to be ranked and/or repeat Sweet 16 run?

The last time we had a great defense was 2008, when Ousmane Barro swatted shots to backup up our quick guards and make MU the 10th best defense in the country (point per trip adjusted for competition on Pomeroy). Then during two years of having no one to contest the rim we slipped to 51st and 57th.

Finally last year we were able to once again contest at the rim, shockingly blocking more shots in Big East play than we had blocked due to Chris Otule and Jae Crowder erasing points with swats – and we still dropped to 61st in the country – meaning we had to be doing much worse job overall on defense except for the efforts of Otule, Crowder and Blue, who are the only three Marquette players to rate as defensive stoppers on the Value Subtract system.

Our problem is players have to play both ends of the floor, and so far Jae Crowder is the only player who ranks as both a great offensive player and a great defensive player.

Can Blue and Otule improve their offense enough that they can stay on the floor to play their lockdown defense in 2011-12?

Can Darius Johnson-Odom, Junior Cadougan and Davante Gardner improve enough on defense that Buzz can keep them on the court to give MU one of the top offenses in the land?

Can Jamil Wilson, Jamail Jones and Juan Anderson provide the athleticism and length to fill the role Jimmy Butler did when he improved the defense dramatically by taking on Kemba and Tu 40 feet from the basket and shutting them down? It seems with that many fast, long 6-7 guys, we could have the ingredients for the kind of defense we saw at the end of the year. Remember, Jones is the top NBA prospect, Jamil was already an above average shot blocker and rebound as a freshman at Oregon before working out for a full red shirt season and Anderson has apparently retained his incredible speed while growing a good bit from his original 6-foot-6.

What kind of defense will we get at the guard spot from a combination of the brother of an NBA player whose own stock has been rising (Todd Mayo), a guy who was one of the top steals guys in the country before transferring (David Singleton) and a guard tough and athletic enough to be wooed by D1 football programs as a running back (Derrick Wilson)?

Marquette has enough offense to repeat a Sweet 16 run. Marquette has the athleticism and length to potentially have a great defense that could repeat a Sweet 16 run.

The question becomes beyond the one given two-way star in Crowder. Can the defensive stoppers improve their offense enough to contribute on both ends of the court, and can the offensive stars play enough defense to do the same? If Buzz has to substitute offense for defense at every whistle, he has a tough coaching job ahead of him. If not, this season could be special.