Yeah, +/- used to be a major metric used in hockey, but modern analytics showed it isn't all that great a predictor of success. Since +/- has fallen out of favor with statisticians, it gets a lot of hate as a "junk stat", but I do think the criticism is a bit overblown (I'm saying this as a stats person). While efficiency metrics are far better when analyzing both individual performance and given line performance, +/- can give some insight into the intangibles that are benefitting the team. When you compare those with higher +/- lines and marginal or worse efficiency metrics with those who have lower +/- lines with higher efficiency metrics, you can see certain things jump out. Things like isolated moments of very high danger shot selection, or counterattack, or better limiting of high efficiency shots despite getting your head bashed in from a modern metrics standpoint. And +/- does have some tribal insight into team chemistry. Sometimes pairing lower individual efficiency players can provide better tangible results even if from a statistical efficiency standpoint this simply should not happen. Math is funny that way.Honest question - any idea how much the staff uses +/- during a game to decide who plays down the stretch? Are they looking at the stats, or is it just a gut feeling thing? My impression of +/- is that it's very noisy, even averaged across many games, so I'm kind of hoping that it (+/- over 32 minutes or so) is not the determining factor for who's playing in crunch time...
That said, this game in my opinion had one of the more severe cases of +/- leading to some poor conclusions I've seen recently. I'll do a full workup later to explain, but the long and the short of it is that TSJ was awful efficiencywise- quite frankly the worst on the court, and Quincy while getting a few rebounds in his 16min had poor individual offensive efficiency (0.88ppp) and largely disappeared (contributed in only 12% of possessions!). Coleman also had a poor game offensively (0.88ppp) but did at least contribute defensive. That brings us to Justin Harmon, who was far and away the +/- darling of our team, and looking at anything +/- related, you'd assume he was far any away our best player. Except he wasn't.
Illinois was considered +6.8 on offense and +3.9 on defense when Justin was on the court, the highest of any player on both sides of the ball. His +/- leads to the conclusion that he put the team on his back and carried his teammates to victory, but then you look at his efficiency and box stats: 0.79ppp, 4rbs, 1ast, and no other stats. Justin was the worst on the team offensively in terms of efficiency! He also contributed next to nothing statwise defensively, and in fact, when he was on the court, he personally gave up double digit points to the guys he was defending on fairly high offensive efficiency for them.
So what in the hell happened? How could someone with such bad individual efficiency metrics look so great with +/-? It makes no sense. But then you look at the usage stats and suddenly things make sense. Justin was on the court when Ty took over the game and Marcus was playing his best ball. And who was off the court in the majority of minutes Justin played? TSJ and Quincy. So basically, Justin rode the Tydal wave to good +/- without having to suffer from the poor play of TSJ or uninvolved play of Quincy.
Point is, +/- can indeed give some insight, but you also have to be careful with the conclusions you make because of it. Based on the lineups Brad was throwing out there, it makes sense he was looking at +/-, but I think he really overvalued it with Justin Harmon, who was more in the right place at the right time over directly contributing to it. I'll add charts of all this later when I'm not at work, but I hope it provides some cursory insight.