"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, March 17, 2010

A Numbers Look at Washington

Time to break this game down by the numbers with some predictions and recommendations for success. I'm not going to get into where Washington is good (turnover margin, pace, and offensive rebounding), and by now you should know the details on Marquette too. If you want the stories, take a few minutes and read the media update from this morning.

I boiled down the stats for each team (using only Top 100 Pomeroy opponents), and here are the predictions for each team.

MU prediction


Red is bad. The prediction according to the numbers is that Marquette will be worse than average on our field goal percentage, turnovers committed, and offensive rebounding. We are predicted to get to the line much more than normal. Taking this away from the pure stats, and this makes sense. Against an opponent that plays decent defense, forces a lot of turnovers, and commits a lot of fouls, MU should be impacted. Add it all up, though, and our overall efficiency is predicted to be worse than normal. Not so good.

UW prediction


The color code is still red = bad for MU, but it's set against Washington's season average. Unfortunately, Washington is predicted to shoot better than average on eFG% and do better than average on offensive rebounding, but turn the ball over more than they normally do and not get to the line as much. Again, taking it up a level, MU's defensive eFG% isn't that good, we force a lot of turnovers, we are not that strong at defensive rebounding, but we do a great job of limiting opponent free throws.

The bottom line

Unfortunately, the prediction is that Washington wins this game with an efficiency of 103.8 versus our efficiency of 98.7. Adjusted for pace, that's a 69-66 victory for UW. Boo!

Luckily, there is a ton of sensitivity in the prediction. How is it that the Warriors will be able to turn this around and win the game?

The most important three keys to the game

  • #1 - Shoot well (no kidding). On specific numbers, MU needs to get an eFG% of 51% or more. That’s right at season average. MU should fire up a bunch of threes and make them. Marquette is 19-2 when shooting over 49% from eFG%. Our only two losses were @WVU and @VU in that list.
Here is the specific concern... Washington does a good job limiting three point accuracy.

However, the Pac10 is not a three ball conference. Consider this. Marquette is #7 in the country at making three point shots. Want to know who is best in the P10? Cal (#49), and then ASU (#62). Not only that, but teams don't take a lot of three pointers in the P10. Only one team in the P10 takes more three pointers than MU. That’s ASU (#9 overall). Even we don't take that many three pointers overall (#103 nationally). The key point is that UW has hardly faced a team that throws up as many threes as us, and hasn't even come close to a team that makes as many as we do (soft rims, soft rims, soft rims).

Also, please stop taking the long two-point jumper (the McNeal dribble-shoot) jumper. It's the most inefficient shot in basketball. Even when it goes in, it's a bad shot. Please. stop.
  • #2 - Don’t let Washington get easy shots. Hold them to an eFG% of 45%.
Highly unlikely, mostly because our defense isn’t that good (because we’re short), although we are consistent. It's probably easier to say, don’t let Washington get over 50% and try to win other areas.
  • #3 - Offensive Rebounding – If MU gets an OR% of 33% (meaning they get an offensive rebound on 1/3 of their missed shots), that increases their chances of winning significantly.
Marquette averages 32 rebounds per game. The prediction is that MU only gets 7 offensive rebounds against Washington. If Marquette could only get 11 total offensive rebounds, that would shift to a MU victory.

Here is the big concern. Want to know how many offensive rebounds MU had in the BET? Twelve. Over three games. The big concern is that Buzz will follow with his BET strategy (where our OR% was never higher than 15%) in an approach to prevent Washington from getting points on the break. Let's hope for some clutch offensive rebounds by Lazar or JFB.

Secondary importance for the victory

The following set of keys are important for victory, but are far less sensitive to the overall margin of victory.
  • #4 - Protect the ball at a turnover rate of 13%.
Washington forces a lot of turnovers. They thrive on ball pressure, but realize that the Huskies don't force as many turnovers as Louisville, Syracuse, or even Marquette does. MU has protected the ball extremely well this year, and done better than 13% nine times this year, including against Syracuse.

This may be the key to the game according to Buzz’s gameplan. Honestly, I really like our chances with matchup, compared to when we face teams like Pitt, UConn, and Cincy. Those teams never force turnovers. In other words, if protecting the ball is a strength, each turnover is much more costly against teams that never force turnovers.

  • #5 - Keep UW at an OR% of 27%
The Huskies are predicted to get 14 offensive rebounds. This is definitely a strength of their team. Of course, realize that here are teams in the BIG EAST that are better at offensive rebounding (West Virginia, Providence, Louisville, Cincy, UConn, Villanova, and Syracuse). UW would be the eighth best offensive rebounding team in the BIG EAST. We need to keep UW to less than ten overall offensive rebounds.
  • #6 - Turn Washington over at a rate of 25%
This is really unlikely, and speaks to where the sensitivities start to come into play. A better bet is to just make sure that we win the turnover margin battle.

Summary

Even though the base prediction is that Washington will win by three, there is a ton of sensitivity in that analysis. Here's how MU will win
  • Take and make threes. Lots of them.
  • Limit UW’s break
  • Keep the Offensive rebound margin close. Don’t get crushed on offensive rebounding 14-7! Keep it closer to 10-10 each
  • Win the turnover margin battle
The bottom line is that this game will probably boil down to a single possession. (What else is new for Marquette?) In situations like that, every little thing matters. Watch the careless turnovers and root for every offensive rebound.

6 comments:

James said...

Numbers are number. Make 3's, stay out of foul trouble - win game.
We're not going to let some wimpy Pac-10 team beat us, right!? We play Big East basketball 'round here.

John said...

You have to have a game plan. Determination and effort will get you so far. If the initial plan isn't working, you also need to realize you need to make changes and make them correctly on the fly.
GO MARQUETTE!

John said...

Oh yea, awesome analysis again.

Andy said...

Nice analysis, guys. I haven't been reading this blog all season, so I don't know if you've done this before, but I'd imagine (just looking at the numbers) that we should have lost quite a few games this season that we wound up winning? Is that a correct assessment?

Rob Lowe said...

Andy, not quite. Taking out all of the cupcake games, MU played 26 games against conf opponents or Top 100 ratings. Our record against those opponents according to the math should have been 15.9 - 10.1. Our actual record was 15-11 in those same games.

We lost a few games we should have won and won a few games we should have lost. Overall, we are slightly under where we should be according to the numbers.

JasonRubenstein said...

UW fan here again. 26 games against Top 100?!? Holy moley.

Here's hoping we stop the three and you get into foul trouble.

Is it 4:20pm (PST) yet???