clock menu more-arrow no yes mobile

Filed under:

Game 3 Analysis: On Norm, Toronto’s bench, and the superlative play of Kawhi

After the Raptors clawed out a Game 3 victory, what have we learned about Toronto’s play and how it could turn the series against the Milwaukee Bucks?

Milwaukee Bucks v Toronto Raptors - Game Three Photo by Gregory Shamus/Getty Images

We haven’t touched base in this space since Game 1 of Raptors-Bucks, so for this one we’ll take a look at some emerging trends and positive signs from the past two games, which is far easier to do after the Raptors won Game 3.

What have the Raptors learned about the Bucks in three games? Can they avoid another Game 2-style blowout? What’s working and what’s not working for Toronto? Let’s figure this out.

Keys to the Game

Over the past two games, which includes that ugly Game 2, here are your on-court net rating leaders for the Raptors.

Player | Net Rating
Lowry: -0.5
Powell: -3.5
Leonard: -4.3
Siakam: -4.3
Gasol: -5.3

One thing should not surprise you. Kyle Lowry, as ever, is the key to the team’s success.

And that’s what made the end of Game 3 so difficult — Lowry was +15 in 31 minutes on the night — and at the time he fouled out, the team was -10 in the 11 minutes he had sat to that point. Lowry likely would have sat even less than that if he had not been in foul trouble most of the game, but still the idea of closing the game with a narrow lead, let alone playing two overtimes without Lowry, was not an appetizing idea. It was an incredible effort by the rest of the team to find a way to close out without him.

The other key that has been developing is one the Milwaukee Bucks must be getting awful tired of. Norman Powell, mired in a terrible playoff run, arguably the worst player on the court for the Raptors for the first two rounds (yes, worse even than Fred VanVleet, in spite of their relative shooting numbers), has miraculously transformed into a solid rotation piece for the Raptors. In the playoffs. Against the Bucks. Again.

Powell has once again played like a borderline star against the Bucks. In this series, he is averaging 22 points, 5 rebounds and 3 assists (with fewer than two turnovers) per 36 minutes on 56 percent from the field and 53 percent from 3 point range. That’s 22 points on 17 possessions and a 70% TS% while carrying a 22% usage rate. Powell can’t average those numbers over a larger sample — that true shooting percentage is simply unsustainable. But if he can maintain something close to that over the rest of this series, he could once again be a big difference in a Milwaukee playoff series.

Still, That Bench... Still

You might wonder, though. Powell has put up decent numbers earlier in the playoffs as well, but posted brutal on-court team numbers, far worse even than Serge Ibaka and VanVleet, the two other bench players who have struggled so much these playoffs. So what’s changed?

Well, a big part of it is Powell has been given some minutes where he is the only bench player on the floor — the sort of look where Toronto’s bench players in general have succeeded. Also, Norm has played fewer of his minutes in those problematic lineups where all three bench players are on the court together. Here are the most used Raptors lineups with at least one bench player in this series. These are small samples (only the four lineups with more than five minutes played are shown) but they demonstrate a larger point we’ll look at with a larger sample in a minute.

Lineup | Minutes | Net Rating
VanVleet-Green-Leonard-Siakam-Gasol: 14 MP, +12 RTG
Lowry-VanVleet-Powell-Leonard-Ibaka: 11 MP, -21 RTG
Lowry-Powell-Leonard-Siakam-Ibaka: 8 MP, +25 RTG
Lowry-VanVleet-Powell-Siakam-Ibaka: 6 MP, -32 RTG

Those small samples are hardly proof of anything, except that this is a continuing trend. Play all three bench players together, and no matter who fills the other two spots, it’s going to likely mean losing those minutes. Play one or two in lineups that make sense around them, and you can start to see some success.

For a larger sample and more evidence of this, here are some splits for the entire post-season, looking at minutes played all together, and then in pairs and individually away from the other bench players.

Combination | Minutes | Net Rating
VanVleet-Powell-Ibaka: 155 MP, -3 RTG
VanVleet-Powell: 37 MP, -21 RTG
VanVleet-Ibaka: 58 MP, -6 RTG
Powell-Ibaka: 18 MP, +10 RTG
VanVleet alone: 52 MP, -8 RTG
Powell alone: 18 MP, +58 RTG
Ibaka alone: 85 MP, +15 RTG

Some interesting trends there. In general, it seems like VanVleet’s struggles are dragging down any unit he is in. His minutes should probably be limited — but it’s worth noting that for the most part, VanVleet is also not dramatically losing those minutes, except when paired with Powell. So replacing him with another guard further down the depth chart likely just means the team does even worse in those minutes. Better to bet on VanVleet rediscovering his shot, and in the meantime keep him and Powell separated — both work better as the lone non-starting perimeter player in a lineup.

Ibaka has thankfully seemed like a viable option off the bench, and although he also should have his minutes limited next to the VanVleet where possible, at least his minutes beside the starters and beside Powell have gone well.

This does further highlight just how valuable a functional VanVleet would be to the Raptors. As of late, he’s shown positive signs playing with the starters in this series, but the biggest impact we could see from FVV is likely just making some shots — and taking better ones. Far too often we see VanVleet trying to create offense as the primary ball handler. This mostly leads to him putting up tough, contested looks inside the arc, as demonstrated by his 2-for-9 line from there so far this series. VanVleet is -26 in his 68 minutes played in this series — eliminate those nine inside shots (redistribute them elsewhere at about 1 point per possession) and have him hit 5-of-11 threes (45%) instead of two, and his place in this series becomes more viable (he’d be only -12 across three games). This is all hopes and dreams, but these little changes in efficiency in shot selection and simple shot making can really swing games very quickly.

If no such correction occurs, the Raptors may need to once again shrink VanVleet’s minutes, even with the risk that comes with overloading the starters. And the more they have to do that, the more valuable Powell’s performance becomes — and the more glaring the absence of OG Anunoby becomes as well.


Kawhi Leonard had a pretty impressive Game 3, overall, though you wouldn’t think he was a real game changer from his line. He did put up a lot of points, but took a lot of possessions to do so. Between shooting only 44 percent from the floor (a reasonable number for a perimeter player but far from what we have come to expect from Leonard) and turning the ball over five times, he only put up his 36 points by using up 33 offensive possessions. That’s not a efficient night for Kawhi, not by a long shot.

But what do they say about Kawhi? Never is it simply said that he is one of the best players in the game — no, it’s always that he is one of the best two-way players in the game. Leonard’s defense comes and goes, with the huge scoring burden he has to carry, but we do see spurts of dominance, especially when it matters.

Here we had a quasi-elimination game: if the Raptors go down 3-0 in the series, their season is essentially over. And Kawhi posted an 88.9 on-court defensive rating, second best on the team, while piling up an incredible 52 minutes across two overtimes. In the second half, that dropped to a 76 DRTG. In overtime, he had an 84 DRTG. These are all very good numbers.

But they don’t begin to describe the effect Leonard has when he is locked in defensively. In Game 3, Kawhi was on the court for 105 defensive possessions. Across all 105 possessions, his direct check scored 12 points. Twelve. His direct check finished the possession 29 times. So that’s 12 points on 29 possessions. And his direct check created one assist in those 105 possessions. One. Overall, Kawhi’s defensive line was 12 points on 5-of-24 shooting, including 0-of-7 from three point range, with three turnovers to one assist.

Kawhi defended Giannis on 41 of those possessions. Giannis shot 2-of-12 and scored four points on those 41 possessions, while his team managed only 35 total points (an 85 ORTG). Next up, 20 of Kawhi’s defensive possessions came against Middleton. In those 20 possessions, the Bucks managed a 95 ORTG (still bad, but better). But Middleton produced none of those points — he was 0-for-3 from the floor and had no assists. The Bucks essentially completely removed Middleton from the offense rather than try to score against Kawhi.

In the series as a whole, Kawhi has held Giannis to nine points on the 60 possessions they’ve matched up. Better yet, unlike with other players, where the Bucks (and other teams in these playoffs) avoid using the player Kawhi is matched up to, Milwaukee has thus far played into that matchup — they may not have an option with the way their offense is built. Giannis is actually using eight percent more offense when guarded by Kawhi than he typically does, in spite of his losing that individual matchup thus far, leading to the Bucks scoring 22 points per 100 possessions fewer than expected on those possessions.

Contrast that with Middleton, who Kawhi has guarded 99 times. He has held Middleton to only 13 points in those 99 possessions, but the Bucks have scored at their typical pace in those possessions. Ultimately, Middleton’s scoring hasn’t driven the Bucks’ scoring this series, and so that’s a matchup you can leave to your lesser defenders (though it is quite an embarrassment of riches to describe Pascal Siakam, Danny Green and Kyle Lowry as lesser defenders).

As this series continues, watch for the Raptors to use this matchup option more and more. Possessions by Kawhi on players other than Giannis may be a waste of his talents. The question may ultimately become one of fatigue — how long can Leonard keep this up, and for how much of each game? I personally would love to find out.

All stats per