a terrible blight on our fine country
Ricky Doyle played up to the competition. [Fuller]
While our attention has, for the most part, turned to football in the offseason, a new KenPom feature has me digging back into hoops. On individual player pages, KenPom now displays split stats for performaces against (1) conference opponents, (2) games against top-100 opponents, adjusted for game location, and (3) games against top-50 opponents, with the same home-court adjustment.
This is a very useful tool for parsing out how well players did against better competiton. Michigan's big man situation continues to fascinate me, so I thought it'd be useful to see how last year's troika performed against the best teams on the schedule, especially since the disparity in big man quality tends to be large between bad teams and good teams. While KenPom hasn't yet separated out stats for non-top-100 opponents (consider this a humble suggestion from a mathematically challenged blogger), we can get a baseline by looking at each player's full stat line from last season.
|Ricky Doyle||43.7||17.9||117.4||10.4||11.9||12.0||2.6||4.0||39-66 (59%)||72-119 (61%)||0-0|
|Max Bielfeldt||34.2||22.3||107.2||12.4||19.5||13.7||1.9||3.8||22-32 (69%)||54-99 (55%)||8-30 (27%)|
|Mark Donnal||22.3||17.0||119.6||10.2||16.1||9.6||3.8||6.4||19-27 (70%)||25-44 (57%)||7-19 (37%)|
And now, each player's stats against only top-50 opponents. This covers 13 games from last season; Ricky Doyle and Max Bielfeldt played in all 13, while Mark Donnal participated in 11 of them.
|Ricky Doyle||51.4||15.5||117.7||8.3||13.5||12.6||2.2||3.8||16-25 (64%)||33-55 (60%)||0-0|
|Max Bielfeldt||32.7||22.9||91.8||8.7||21.9||16.7||2.1||4.6||8-11 (73%)||19-38 (50%)||3-14 (21%)|
|Mark Donnal||17.0||20.6||128.2||13.9||7.3||3.6||4.0||8.4||7-11 (64%)||12-22 (55%)||2-6 (33%)|
The above helps clarify why John Beilein was comfortable letting Bielfeldt go despite having the opportunity to bring him back. A few takeaways:
Doyle held strong. Doyle's offensive numbers stayed almost exactly the same against top-50 competition; his shooting held at 60%, he took care of the ball, and he allowed the offense to run through the guards/wings. While his offensive rebounding dipped, he still did pretty well in that regard. Equally as encouraging was his ability to hold up defensively; Doyle's foul rate stayed level and he took on a larger share of rebounding duties against top teams.
Bielfeldt's shortcomings became apparent. Bielfeldt proved effective against mid- and lower-tier teams in large part because he dominated the offensive glass, providing himself with easy putback opportunties. Against top-tier teams, however, his offensive rebounding fell off dramatically, his turnover rate rose, and he didn't have a post game or reliable outside shot to make up for either.
Bielfeldt also resorted to fouling more on defense. He was clearly overmatched on that end against high-level competition and that took him out of games even when he had it going offensively; for example, he had nine points on 4-6 shooting in the home overtime loss to Wisconsin but picked up three fouls in 14 minutes because he couldn't defend Frank Kaminsky or Nigel Hayes.
Donnal showed promise on one end. Donnal's decreased role as the season wore on means his sample size is smaller than the others—he essentially played two games worth of minutes against top-50 teams, and he did so in short stints. Those short stints weren't always by design. Donnal was foul-prone in the best of times but especially against good teams; yes, that 8.4 fouls/40 minutes figure is real and speaks to some major defensive shortcomings that were apparent to anyone who watched him play.
There's hope in the offensive numbers, however. Donnal was... good? Again, the tiny sample size makes it hard to draw grand conclusions here, but his rebounding rate and shooting numbers are encouraging.
With a bulked-up DJ Wilson—listed at 6'9, 240 on the updated roster—set to bolster depth up front, it makes sense for Beilein to prioritize developing Donnal and Wilson into reliable options instead of giving significant minutes to a redshirt senior whose limitations become very apparent in the most important games. With a year of development under Ricky Doyle's belt and a logjam at the four, Michigan may only need one of those two to play a major role off the bench anyway.
Kenpom. The all-knowing. This year's best team not invited was #45 Florida. The worst team in is #250 Hampton. I had to know, so I looked up the worst team in the history of Kenpom (since 2002) to be invited to the Dance. It's Mississippi Valley State, the 298th team in 2008.
In fact there's a pretty enormous drop-off from most of the 16 seeds and the MEAC and SWAC entrants, who had an average Pyth of 0.296 (i.e. they'd win less than 30% of their games vs. an average opponent). For reference, the worst Big Ten teams in that span (2003-2005 Penn State) were .361, .334, and .341 respectively. In fact I only found six power conference teams—2013 TCU, 2008 Oregon State, 2012 Utah, 2013 Mississippi State, 2012 BC and 2011 Wake Forest—who've ever been worse than the average MEAC or SWAC champion. I get the part about giving the top seeds basically a bye, but the tournament can find more deserving small schools than whoever won a conference tournament whose competition level is below that of many high school leagues.
For what it's worth, Kentucky this year is the best team in the history of Kenpom. The only two in sniffing distance were 2008 Kansas and 2013 Louisville.
Seth's Annual Matchup Maker. This lets you set a chaos factor and match any two opponents, immediately seeing where the game will be played and any relevant injuries. You just input the teams and the round. Front page looks thus:
The Power Rank. Listeners to Brian's weekly roundtable on WTKA know Ed Feng. Ed creates this:
(right is zoomed)
…interactive chart using his win probabilities so you can see how stupid your picks are relative to each other. Run around the circle to make sure you haven't picked a dead in the water first round upset, but I think he's best at end game. This year you can see Kentucky is in a league of its own, then Zona, Duke, Nova, Wisconsin, Gonzaga and Virginia are a clear second tier.
Bracket Science Bracketmaster. Peter Tiernan is getting better at monetizing his comprehensive bracket database, which is unfortunate because I really liked to use the Bracketmaster for patterns, like what kind of team does Wisconsin usually lose to in the tourney, has this coach been to the Dance before, and things that super hardcore NCAA basketball fans know and I don't know offhand because I was off the wagon for a time. If you don't mind paying (there are far worse people you can give money to than Tiernan) you can get all the goodies, but the free stuff is great for narratives, for example if you want to track how Big Ten teams have fared since 2005 vs other Power Confs:
WSJ's Blind Comparison. The Wall Street Journal's blindfold bracket is your bias check, though this year they didn't do, opting instead for a slider-based bracket generator using things like "defense wins championships!"
Disclaimer: You will be wrong.
Site note: As with last year, we'll be having a basketballgasm liveblog for Day 1 of the tournament, shifting to the hockey game at 3, and then going through the Round 1 matchup with Wofford. DraftStreet, whose 40k tourney is still filling up (as of this morning ~1600 of the 2000 spots are filled), is sponsoring, and a few former players will be joining us to promote the Go Blue Bowl.
Speaking of filling things, you're probably filling your brackets right now, so here's my now-annual post and tool for helping with that. Last year was the first since 2000 that I didn't win at least my buy-in back. Things I use:
The Power Rank (friend of the blog Ed Feng)'s interactive bracket. Ed is one of the cutting-edge guys in sports analytics. On his tool if you hover over any team you can see their probabilities to reach each round, or hover over a spot in the circular bracket to see every team's likelihood of getting there. Michigan is 58% to reach the Sweet 16; from there every game is virtually a toss-up.
The Wall Street Journal's blind comparison. They show you two profiles and say a little about the team, and you make your pick presumably without bias, though you can often figure out exactly who they're talking about:
Bracket Science's Bracketmaster tool. Peter Tiernan's blog is a standard for following bubble teams and gets things right that others don't (like Louisville as a 4 seed). The Bracketmaster+ tool lets you get into data going back to 1985. If you're a member it gets deeper but non-members can use it to do things like show Beilein's Michigan teams in the tournament:
Poologic Tool. This helps you decide how many upsets to pick based on the size of your office pool (in a large pool it's best to be the only one with a certain champ). Also you can calculate ROI on various picks.
My tool (download the excel sheet) Which uses straight-up Kenpom scores and provides a weak confidence score based on the premise that 16 seeds never beat 1 seeds. I also added injuries for each team. Looks like this:
What I do is normalize the closest 16-1 matchup (Wichita St vs. Cal Poly) as 100% for the 1 seed to win, set that as the "chaos factor," and use the KenPom ratings to percentile everyone else's games into a confidence number. Then I roll through anything under 70% and decide if my knowledge of those teams might justify taking the under.
If you're in a big pool, run multiple brackets, each with carefully selected upsets.There's no such thing as an NCAA tournament without lots of big upsets and at least one surprising run. The 1 seeds all made it to the Final Four just once. If you submit one milksop bracket you're up against every other milksop bracket and will get beat by the one crazy guy who had LSU going to the Elite 8 or something. Hitting on a carefully selected upset that rearranges a bracket and lets you ride a different high seed to the Final Four is the most typical route to a win.
If you're in a small pool, play conservative. One or two points won't usually make a difference in a small pool, but the likelihood of something crazy like that one guy's wife who picks based on the cuteness factor of mascots winning is cut down so you don't need to take risks to get ahead.
Pick the upsets the most carefully. I love picking 6-11 upsets because if you get it wrong they're bound to get wiped out by the 3 anyway. If you roll the dice on a 3-seed or lower losing early though, you'll feel like an idiot as the rest of your pool collects the easy points. A tournament without upsets never happens, but neither does a tournament with all the upsets. You can totally undo a great pick with a terrible one elsewhere.
Get value for your upsets. Know who's in your pool and the inefficiencies. This year, those of you in Michigan are facing the mother of all inefficiencies in that Spartan fans are bound to submit extra brackets just to have one that has State going all the way. Fans will generally take their favorite team to go two rounds later than they really belong and conference teams to go a round further. This is an inefficiency (even if MSU looked like they could dominate the tourney on Sunday).
Be really really lucky. This is really the only rule.
Also dear yente if you could saddle Wisconsin in an impossible bracket that'd be great.
In two days it shall be March. We're already familiar with the Big Ten Tourney participants, so let us look beyond to this NCAA tournament thing.
What, my dear yentes, makes a good or bad matchup for this Michigan team, what are some of the teams out there we might hope to avoid, and who among expected high seeds would Michigan match up well against?
Brian: Bracketology consensus has us a three or four right now; I'll go under the assumption they're a three just to simplify things. That means Michigan is looking at the top eight teams on S-curves trying to suss out a good matchup. Wisconsin is in that group for some lucky 7-seed but won't end up in Michigan's region. The others: Wichita State, Florida, Syracuse, Arizona, Kansas, and some combination of Cincinnati/Villanova/Creighton.
No one in that group seems hugely appealing, but I like the Syracuse matchup best of the current one-seeds. They've only got one shooter, they're 11th on Kenpom, they've had a lot of close calls against not particularly good teams, Michigan played their zone last year, and they've got shooting from everywhere. 'Cuse's current backcourt is much smaller and less athletic than last year's version and Michigan's shot generation is a lot bigger, so going over the zone is much more of an option. Also, undefeated or not, Wichita State is short, largely untested, and not laden with NBA superstars future. I will take either of those one seeds.
Conversely, I want nothing to do with Arizona. Michigan damn near beat them earlier in the year, yeah, but that was thanks in large part to an avalanche of missed putbacks. Teams that can just implode Michigan on the boards are my biggest fear. Kansas and Florida are also teams I'd like to avoid.