Dayton/Duquesne Preview (or, A Test Drive for the "KenPom-to-English" Dictionary)

Hey, everybody: this is Your Favorite Internet Semi-Star "The" Rick Scaia, stepping up and ready to prove I'm every bit as ingenious and insightful when it comes to Flyer Hoops as I am when it comes to well-oiled men rolling around with each other on national TV every Monday night. Which is to say: very ingenious and insightful.

If you heard my first appearances on the UDFN Podcast (here and here), you know I'm gonna get pegged as something of the "stats guy" for UD Flyer Nation, and it's a role I'll fill gladly, if for no other reason than because -- in any debate or discussion -- I feel it's disrespectful and downright retarded to not be able to back up your punditry with facts and truths, and present nothing but bluster and BS in their place. Unlike much of what passes for sportstalk radio or political commentary or almost any form of discourse in this day and age of the intarwebs and "democratized intelligence," I don't believe arguments are won by shouting your "opinion" the loudest or most number of times in a given hour.

Repetition and decibel levels don't make you right. Being RIGHT makes you right. Anything else is a load of crap to me. If you're objectively and factually wrong, but still feel the need to express yourself with the defense that "Hey, it's my opinion, and you can't prove me wrong. Respect my opinion!", I assure you, I will be wanting to punch you in the throat more than I care to debate you. "The sky is green," "Obama is not an American citizen," "the Earth is 6000 years old," and "Kurt Huelsman is the worst player ever to don a Flyer uniform" are not Valid Opinions That Must Be Respected: they are demonstrable falsehoods that must be ignored, marginalized, and perhaps even mocked.

Ergo: if I'm "Stat Boy" around here, so be it. I'll tell you a number, or inform of you an actual factual event, which cannot be disputed, and THEN maybe I'll tell you what I think it means. I will delineate between fact and opinion, the former being bulletproof, and the latter being open for discussion.

"UD shoots almost 40% of its FGA from three" is a fact; "That's wayyyyy too many" is my opinion. In this case, we probably all agree on that little fact/opinion synthesis... it's the sort of rational thing you'll be seeing out of me. I'll leave the "UD lost to UNM" (fact) followed by "BG is horrible coach and must be fired" (opinion) caliber of syntheses to others. And boy howdy, are there enough of those others out there if that's the sort of insight you're in the market for... so go ahead, by all means ditch out now and ignore all my future postings if any of this rubs you the wrong way. Trust me: I won't be offended if you leave UD Flyer Nation, never to return again. Secaur will be, since this is his baby and he wants to be famous. But I'm already famous enough, baby, and can afford to antagonize the audience!

I digress.

So, "What are we gonna be doing here today, Rick?" I hear you asking... well, we're gonna look at tomorrow's Duquesne/Dayton game, using a specific set of numbers and data as our baseline, and see if we can't come up with some predictions and "Keys to the Game" that are a little better than Derrick Dukes stumbling over the concept that "For Dayton to win, they must score 10 points more than the opposition." Or whatever Harry-Carey-esque gem it was along those lines that he uttered during the GMU game.

That set of numbers/data? Ken Pomeroy's, from his profile pages for both teams, and the source for his own Pomeroy Rankings (which currently place Dayton as the 58th best team in the nation). I figure this is a worthwhile exercise because KenPom really does include some pretty cool figures that go FAR beyond a player's points per game or team rebounding or other things like that which most fans are most used to hearing about. I compare the stats and factors used by Pomeroy to the increasingly common use of stats like OPS in baseball circles. If you really think batting average fully encapsulates what makes a player productive in baseball, you're probably the sort of baseball dolt who thinks Tim F. McCarver brings a lot to a TV broadcast; there are now a few other stats which (when applied properly) cut more to the core of How You Win Baseball Games. Similarly, KenPom puts some numbers out there that are NOT the ones that show up on ESPN or in box scores, but which are very illustrative in terms of what causes teams to score points (or prevent points on defense).

Pomeroy takes all those numbers, and just spits out his ratings and a predicted outcome for every game. For him, that's where it ends. He has his numbers, and for tomorrow's game, he spits out "Dayton 66, Duquesne 56," and that's the extent of it. We can do better. Or at least: I can TRY to do better. And thus, I present my embryonic "PomPom-to-English" translation filter! Take what information Pomeroy has collected, synthesize it, and see if we can't figure out what's gonna go down at the D-Rena tomorrow afternoon... it's worth a try, at least, no?

Let's start by making sure we're all looking at the same numbers. KenPom's profile pages for Dayton and Duquesne are here:

http://kenpom.com/team.php?team=Dayton
http://kenpom.com/team.php?team=Duquesne

Glance through them. What do you see (besides a lot of gibberish, if this is the first time you've ever visited KenPom)?

In a nutshell, I'll tell you what I see: what I see here is a potential -- or even a need -- for Dayton to win a game "BG Style." With defense and rebounding.

Less of a nutshell, and a bit more in-depth:

• Duquesne's defense profiles as being quite formidable, and a poor match-up for us: it looks like they play a tough, mostly-zone style which causes (a) an above average number of steals, and (b) an even MORE above average number of poor shots by the opposition. Duquesne ranks in the top 20% in the nation in Opponent's EffFG%, seemingly driven by an incredibly high percentage of blocked shots (poor shots forced up in the paint) and attempted 3pts (or other long jumpers forced late in the shot clock). They do all this without sending other teams to the line (one of the lowest FTA-against rates in the nation). This is not an easy team to score against in the half-court game.
  

• The only weakness in Duquesne's defense is that they allow WAY more offensive rebounds by opponent's than they should (likely a result of being one of the most-undersized teams in the nation, 300th or so in KenPom's "Effective Height" metric).
  
  
• KenPom's paints a picture of our offense as a team that shoots a lot of 3's (but not particularly accurately) and that doesn't get to the FT line nearly enough. Obviously: these two things are related to each other... but when we do shoot 2pt shots, we're one of the best shooting teams in the nation, which means we'd be best served to try to emphasize that while shying away from the treys early in the shotclock. We need to play AWAY from our KenPom profile a bit, as the chinks in our offensive armor (too many 3's, not enough FT's) so far this year are EXACTLY matched up with things that Duquesne's D has done well (force tough shots, commit few fouls) so far this year. This is somewhat worrisome.
  

• Synthesis: our offense may end up looking as impotent as ever outside of our transition baskets (or defense-caused transition game is no doubt a huge part of our high 2pt FG%, and the source of our highlight reel dunks, too)... Duquesne will make us take less-than-premium shots, and won't just send us to the line for a ton of freebies, either. But if we stay patient, don't get into one of our UNM-esque spells of hoisting stupid 3's (keep our 3FGA at 30% or less of our total FGA, even if Fabrizius is back in the line-up), loosen up for a little more one-on-one with our superior athletes driving to the hoop (and getting to the line as Duquesne's smaller interior players are forced to foul), and hit the offensive boards nice and hard, we should still be able to squeeze out more than enough points to win.
  

• Duquesne's offense profiles as slightly below average: they are a horrible 3pt shooting team (one of the bottom 25 teams in the nation), yet for some reason they still take 37% of their FGA from three-land (that's the 78th highest percentage in the nation, but STILL less than what Dayton has averaged). This seems the defining characteristic of an otherwise nondescript offense. It could once again tie into Duquesne being under-sized, relying heavily on the perimeter game and transition baskets forced by their tenacious defense. The fact that they are one of the 100 most-blocked teams in the nation probably backs up this notion that they don't really have the front-line size to convert easy buckets in the paint out of their half-court sets.
  
  
• Our defensive profile is strong, and improving... a month ago, we were an average defensive team according to KenPom's metrics (leaving us to usually rate in the 70s or 80s); now, we're a top level defensive team (and rated in the 50s, even after another loss at UNM). This is because of our steadily improving EffFG% and our always formidable number of steals and rebounds. Just do what we do, and all should be well: nothing in the numbers indicates that Duquesne has any kind of kryptonite to weaken our D. On the contrary, the numbers suggest that our strengths and their weaknesses correspond, and could result in an afternoon of offensive futility for them.
  

• The end result could be a game not unlike Ball St.: if we bring that same 40 minutes of killer defense, Duquesne is a team that will have trouble scoring 50 against us. Or even 45. So let's bring that killer defense, let's augment a possibly impotent half-court game with some transition buckets off that D, and even if Duquesne is superior BSU on the defensive end and holds us in the 50s (a possibility, since BSU kept us to just 60), we can still win comfortably. That's the formula to victory tomorrow, in my eyes; well, to a convincing and heartening victory. We simply shouldn't lose in ANY situation... but if we bring that defense, there really oughta be enough offense to win a game by double digits and with little drama. Even if you're a compulsive gambler type who cares more about beating the spread than about winning the game. [Note to those sickos: I'd bet the under, if I were you. If UD/DU combine for much more than 115 points, something will have had to go horribly, horribly wrong for one or both teams.]
  

Thus ends my first public attempt at using my "Pomeroy-to-English" Dictionary. KenPom takes his little numbers, and spits out the following analysis of the Dayton/Duquesne match-up: "Dayton 88% favorite; game tempo 68." Which is all well and good, but I like my Word Picture lots better. I hope you do, too.

Now: to wait and see if that picture ends up bearing any resemblance to reality, or if my dictionary needs tweaking before my Highly Anticipated Podcast Showdown versus Ken Pomeroy (allegedly coming soon to an mp3 near you)! Till next time, kids....


Rick