Efficiency by half

[Bumped.  Whatever Tom Izzo is doing at halftime, he should go ahead and do it before the game starts.  My theory on the phenomenon is in the first two paragraphs after the jump in Sunday's game recap.

Also, hockey fans should check out SpartanDan's comprehensive rundown of what need to happen for MSU to make the NCAA Tournament. -KJ]

As we are all aware, this year's MSU team has a tendency to fall behind by big margins in the first half of games and then make a valiant comeback effort in the second half. In the past few games against quality opponents those comeback efforts have fallen just short. I thought it would be interesting to look at our efficiency and our opponents' efficiency by half for every Big 10 game. Statsheet.com provides box score data by half but not tempo free stats by half, so I had to calculate possessions myself in a spreadsheet. This was an educational experience, as when I did the calculation using the common formula suggested by Ken Pomeroy, which is FGA-OR+TO+(.475xFTA), I got disparate possessions for each team, often with differences approaching 4 possessions per half. I asked KJ and John Gasaway how to reconcile discrepancies, and both said to average the estimated possession totals for both teams to get total possessions, so that's what I did. Without further ado, here are the MSU and opponent efficiency averages by half:

We are definitively a second half team. In general we play much better offense in the second half of games. Our defense is somewhat worse, but the net effect is a much better efficiency margin in the second half of games compared to the first half. In the second half we look like a true Big 10 title contender, with an efficiency margin comparable to OSU, Purdue, and Wisconsin. In the first half we look like Minnesota or Illinois.

In order to be sure that this effect is not being fueled by a few outliers due to furious comebacks from large deficits, I threw out the Wisconsin, Purdue, and OSU games, and the pattern of improved offense in the second half still holds. We play better offense in the second half of games, on average, and better defense in the first half of games as well. Throwing out those three games does change our first-half efficiency margins significantly, however. Without those games we look like a Big 10 contender in both halves, with a first half efficiency margin of .14, fueled by very good defense (average .89 opponent points per possession) and OK offense (average 1.02 PPP). Our second half features excellent offense (an average 1.16PPP) and mediocre D (1.03 Opponent PPP). We get it done with D in the first half of games and offense in the second half, but in both cases our efficiency margin is in double digits if you throw out our three big losses.

Looking specifically at those three games, we had a total offensive and defensive collapse in one of them (Wisconsin), a total collapse on offense in the OSU game, and a total defensive collapse in the first half against Purdue. We did not play particularly good D in the first half against OSU, allowing them 1.08PPP, but it was fantastic compared our offense, which averaged only .72PPP. The D we played in the first halves against Purdue, where we allowed 1.41!!! points per possession, and Wisconsin, where we allowed 1.35PPP in the first half, was really bad. Our first half offense against Purdue was not atrocious but was worse than normal. The Wisconsin game was the only game where we played absolutely terrible basketball in both halves. I guess if there is a pattern it is that our normally stout first-half D deserted us in all three games.

This analysis doesn't go very far toward explaining why we are playing so badly in the first halves of games against quality opponents lately. One could argue that we looked better than we really are due to a soft early schedule, but I'm not sure that's the problem. Granted, we did not get Purdue or OSU until recently, but they hadn't played us (a team that looked, statistically, like a title contender) either. We had Wisconsin, Minnesota (twice) and Illinois as well as Michigan and Northwestern (twice) on our early schedule. Also notable was who we hadn't played yet - no games against PSU or Indiana, the two worst teams in the league. That's one top shelf team and six games against decent teams, vs two games against a basement dweller (Iowa). We didn't play a murderers row but it wasn't a particularly soft early schedule. We just seem to have a knack for coming up small in the first half of big games. If we can break that habit we may be able to turn things around.