Summer Q&A: Is '08 KU better than '12 Kentucky? And who had the best season in the Bill Self era?
Earlier today, I asked Twitter followers if they'd be interested in a question-and-answer blog on KUsports.com about advanced stats, and the response was great.
The following are some of the questions I received. Let's get nerdy.
Kevin Baker @deutschmarine
why is 2012 UK seen as some all time team (especially defensively) when '08 KU is KenPom's number 1 team of the decade?
It's all a matter of perception and our eyes sometimes getting in the way of what might not be true.
To be fair, the 2012 Kentucky team probably had as much talent in terms of NBA prospects as any team in the last decade. Anthony Davis and Michael Kidd-Gilchrist went 1-2 in the draft, and a total of six Wildcats were drafted that year, which tied an all-time high.
But your point stands ... talent doesn't necessarily mean UK was one of the best teams.
The stat you're citing above is KenPom's "pythagorean winning percentage," which Pomeroy uses as the primary way to rank each team. In his own words, Pomeroy defines the stat as "a fancy way of computing a team’s expected winning percentage against an average D-I team."
According to pythagorean win percentage, the college basketball team since 2003 was 2007-08 Kansas, which had a mark of .9859.
2011-12 Kentucky, meanwhile, had a pyth of .9679, which is ranked 20th in the last 11 years. The Wildcats' season actually falls behind two other Jayhawk teams that didn't win it all (2006-07 KU, .9755; 2009-10 KU, .9683).
This also leads me to another point: Oftentimes, we overrate records when trying to evaluate teams.
Though Kentucky tied an NCAA record for wins in 2012 with 38, a 38-2 record might not actually be better than KU's 37-3 record in 2008 when you consider all the circumstances.
According to KenPom, KU's 2008 strength of schedule (.8202) was much tougher than that of UK's in 2012 (.7061). Though we'll never know what would have happened if UK 2012 faced KU's 2008 schedule, the safe bet would be that the Wildcats would have had at least one or two additional losses, which would have knocked them down a few notches with national perception.
And let's be honest: As sports fans, we love watching superstars. I'll easily argue KU 2007-08 was a better team than UK 2011-12, but KU didn't have an Anthony Davis or a Michael Kidd-Gilchrist. Heck, KU didn't even put a player on the All-Big 12 team that year.
That doesn't mean KU wasn't a great team. It just means that the perception of what makes a great team might be different from reality.
Brian Haase @bhaase86
what do u feel is the best advanced stat that accurately portrays how an individual is performing on offense and defense?
On offense, I'll have to go with two hand-in-hand: offensive rating and usage percentage.
Part of offensive rating's problem is that it's a complicated formula, and trying to explain it is difficult. There's a reason many people are drawn to stats like batting average and simple shooting percentages: because they're easy to tally and understandable to non-statistics majors like us.
If you can accept the fact that offensive rating is widely considered the best individual basketball statistic, though (it was developed by one of the leaders in advanced statistics in Dean Oliver), and it can be extremely useful.
Basically, with our eyes, it's hard for us to take in the entirety of a player's game, even if we feel like we know them well. The stats showed last year that Perry Ellis was one of the best players in the Big 12 at avoiding turnovers. I watched him all season, and I never would have been able to come up with that observation on my own without the numbers.
So basically, offensive rating compiles all of a player's offensive statistics (made and missed field goals, made and missed free throws, assists, offensive rebounds and turnovers) and tidies it up into one stat that shows a player's individual offensive efficiency — or how many points he scores per 100 possessions.
It's useful in that you can compare that to the team's offensive efficiency to easily figure out who should be doing more for an offense and who should be doing less. KU's schedule-adjusted offensive efficiency last year, according to KenPom.com, was 111.8. Both Elijah Johnson and Naadir Tharpe's offensive ratings were under 100. Obviously, KU would have benefitted by those guys taking a back seat to other more efficient offensive players.
Offensive rating also faces an obstacle in that it has to be used with usage percentage to be kept in context. Travis Releford led KU with a 125.8 offensive rating last year, but I don't need to tell you that he wasn't a better offensive player than Ben McLemore, whose offensive rating was 118.7.
The difference in the two players is usage (or possession) percentage — the percentage of possessions a player ends either by making a shot, missing a shot that isn’t rebounded by the offense, or committing a turnover. You can think of this as, "How much offensive load does a player take on?" The average for a player is 20 percent.
Last year, Releford's usage percentage was 15.7 percent, while McLemore's was 22.1 percent.
Basically, on offense, you're looking for players with high offensive ratings that also end a lot of possessions. These players are rare, but they're also extremely valuable, as they're able to keep their efficiency up while also taking pressure off teammates by taking on large offensive responsibilities.
Wayne Simien's 2004-05 season (118.5 offensive rating, 26.9 usage percentage) and Marcus Morris' 2010-11 season (121.9 offensive rating, 26.1 usage percentage) stand out as two of the best offensive seasons in the Self era, and those high numbers combined with each other are rare.
As far as defensive stats go ... it's tough. I mostly stick to defensive rebounding percentage, block percentage and steal percentage, just because the standard box score does not (yet) keep many defensive stats.
If you're looking to get a little more advanced, Basketball-Reference lists both individual defensive ratings and defensive win shares, which attempt to give us a glimpse of a player's entire defensive value.
Beware of the Phog @Pay_Heed
I have a basic question. What's the football equivalent of a basketball PER?
Football still sits a little behind the curve as far as advanced stats go, and it's easy to see why.
Baseball is great for advanced statistics because it's easy to assign responsibility to one player. If you're in the batter's box, and you strike out, it's hard to blame that on anyone else.
Basketball is tougher, but there still are areas where we can say that certain actions are nearly independent of teammates. When you're shooting a free throw, it's hard to say anyone else contributed to you making or missing it. Grabbing a rebound is a stat we can assign to a certain player.
Football is much harder because almost every play is dependent on other people. Do you give James Sims credit for a 15-yard run, or should the credit go to the offensive line? Was that sack of Dayne Crist because he held it to long, or because a blocker came free on a blitz?
There still have been plenty of advanced stat breakthroughs in the past few years, especially at sites like FootballOutsiders.com, where smart folks are starting to pin down the best way to measure individual players in football.
So the short answer is: there is no PER for football yet.
If you're looking to go a step deeper with college football analysis, though, I'd highly recommend Bill Connelly's advanced stat season preview of KU on SB Nation. He also provides a glossary at the top that helps explain what some of the stats are and why they're important.
Give it some time, and I think we'll see some of these advanced football stats become more mainstream, just as on-base percentage and effective field-goal percentage have in other sports.
Chris Teegarden @firket2000
I would love to get a better understanding of how coach Selfs system works with advanced statistics. What works & doesn't
Had a couple questions about this. A good start, if you haven't read it yet, is the two-part series we had last week talking about how the KU men's basketball team uses new video technology.
Synergy Sports Technology is just one tool that is used by the basketball staff (along with KenPom.com, I'm told) to evaluate its own players and also opponents. From the articles, you can tell the coaches have embraced these new technological advancements in a short period of time.
Also, if you listen closely enough to Self's press conferences, you can tell he uses KenPom. He's referenced KenPom's "Experience" ranker before and a few times (saying something like an opposing team is the 10th-youngest in the country) and has made it a point to explain to media members that if KU allows a high number of offensive rebounds in a game, that doesn't mean the Jayhawks necessarily had a bad rebounding game (and he's right).
A good example is KU's round of 32 win last year against North Carolina. Though the Tar Heels posted the seventh-highest offensive rebounding total of the season against KU (16), that number was artificially high because the Tar Heels missed a whopping 51 field goals.
The better way to look at how well KU rebounded is by looking at its defensive rebounding percentage, which was 68.6 percent. That number was just below KU's season average of 70.7 percent, and in a completely acceptable range considering the quality of the opponent.
Self often talks about how he's a numbers guy, and he brings up field-goal percentage defense more than any other stat (effective field-goal percentage is a better stat, of course, but I digress).
This topic is probably worthy of exploring further, but based on the conversation I had with video coordinator Jeff Forbes, I can tell you that KU is at least accepting of the new statistics and technology out there to try to gain an edge.
Which is more than we can say for the Kansas City Royals.
using advanced stats, who’s had the best season under Bill Self? T-Rob in 2011/12? Sherron in 2008/09? Simien in 04/05?
Great question, and one I've thought a lot about lately, especially when making picks for the best players during our KUsports.com summer series.
After looking it over, though, I think one season stands above all others: Cole Aldrich's sophomore year in 2008-09.
I referenced this in the ranking of KU's centers under Self, but Aldrich's sophomore season stands by itself in terms of Basketball-Reference's all-encompassing Win Share statistic, which is "an estimate of the number of wins contributed by a player due to his offense and defense."
Here are the top five from the Self era:
To me, no one looks out of place on this list, and Aldrich being that far in front of everyone else only strengthens his case.
Of course, his advanced stats at KenPom.com also back up the argument that he's Self's best player in a single season.
Remember, this team earned a share of the Big 12 title and advanced to the Sweet 16 basically with Aldrich, Sherron Collins, a freshman Tyshawn Taylor and a bunch of other guys. The Morris twins were not good statistically their freshman year (or playing much), and while Brady Morningstar added some defensive value, he shot on just 12.4 percent of his possessions, which limited his offensive role.
Basically, Aldrich had to be dominant on both ends for KU. And he was.
On a team that had an adjusted offensive efficiency of 1.14, Aldrich posted 1.24 points per possession — and that was while ending an above-average offensive load for KU (he ended 21.5 percent of the possessions he was in and shot it on 22.8 percent of his possessions).
Aldrich was a great shooter that year. His effective field-goal percentage of 59.8 percent was 63rd-best in the nation, and his 79.2-percent free-throw percentage still stands as the seventh-best in the Self era (minimum 100 attempts).
Big men also never seem to get credit for avoiding turnovers, but Aldrich was exceptional at this as well, as his 14.3-percent turnover rate was best on the team.
Aldrich was rare in that not only did he block shots, but he also was a dominant offensive and defensive rebounder.
Jeff Withey received a lot of fan love, and rightfully so, but he didn't approach the offensive numbers or defensive rebounding numbers that Aldrich had his sophomore year.
In a season where KU desperately needed Aldrich to score, defend and rebound, he did all three at an elite level without a whole lot of help around him.
Looking back, the center probably deserves the most credit of any Self player for keeping the nine-consecutive-conference-titles streak alive.