Obviously a lot of what we’re seeing in the rotation right now is Darko trying to see what he’s got in different players, as well as putting guys in positions to grow and develop. Not everything is done with winning in mind, that was clearly part of his mandate when he was brought in.

But when we’re thinking though about what the actual, competitive rotation could be, I think there are some general truths that have emerged:

• The starters are generally winning their minutes against the other team’s best units. Some nights the matchups may be more or less in their favour, but they’re solid.
• GTJ generally does well in minutes with any four starters, but also struggles when asked to exceed that role (with one exception being when he’s with Porter).
• Any Barnes + bench minutes tends to lose their minutes, sometimes badly. (Again, I appreciate that there’s a development agenda behind them, but this post is talking about optimal winning lineups.)
• Dick and Koloko out right now kinda simplifies the rotation which makes it a good time to experiment with what an optimal rotation would be. (I single out Dick minutes a lot below and I don’t mean to dump on him, I’m still a big fan but the reality is that any attempt to use net ratings to analyze lineups means accepting that Dick is almost always a huge negative net impact right now.)
• McDaniels is not a rotation player at this point.
• Right now, most nights we’re weighing how well the starters, or starters + GTJ minutes are, against how bad the Barnes + bench units are.
• If a top 6 pick is off the table as the team hovers around .500, and no major trades emerge, at some point there will be an expectation to start to focus on wins more than they currently are.

One of my assumptions at the beginning of the year (and I think a lot of people felt this way) was that you wanted to break apart Siakam’s and Barnes’ minutes, so that you always had one of them on the court, but they each had significant time without one-another. I’m definitely questioning that assumption now though.

Lately, a wrinkle that Darko has been using is to overlap Barnes and Siakam in such a way that we’ve seen Barnes+Siakam+bench toward the end of the 1st and 3rd, and those minutes have been really strong in a small sample size. I feel like this might be the key to optimizing the rotation from a competitive standpoint: in an 8 or 9 player rotation, you’ve never got a four bench player unit, plus Siakam and Barnes anchor the three bench player units, which can beast against almost any opposing bench-heavy unit.

The problem with this is that it means that you’ve got to have more time without one of Barnes or Siakam on the court. These units are rare in non-garbage time. Of 826 minutes of Raptors basketball, just 53 minutes have been without Siakam or Barnes on the court. Just 17 minutes have at least one of Poeltl, OG, and Schroeder without Siakam or Barnes, and no units that fit that description that have more than 2 minutes together. Units that have both Dick and McDaniels account for over 50% of the non-Barnes/Siakam minutes, so as you can imagine any attempt to analyze those minutes means they are atrocious net ratings if you include units with that pair, and meaningless sample sizes if you exclude them.

But I think non Siakam/Barnes should be something to experiment with; if for no other reason, Siakam or Barnes are going to miss some games at some point and those units are going to need to get minutes then. So it seems prudent to start experimenting now in small doses. On paper, what makes sense is something like a Schroder+Poeltl+OG+GTJ+Porter unit. Structurally, this is really traditional and should be familiar to the players… not a complex offense. It is combining the team’s best P&R pair with three shooters, and this unit should also be able to hold their own defensively against other transitional units; or at least have a chance to create enough offense to make up for defensive shortcomings. Obviously Porter’s availability is limited, but he also seems to do really well regardless of the number of minutes, so playing 10 minutes a night in these sorts of transitional lineups might do well. But he still might not be an every-night guy even in that sort of role, and finding who the other guy who can step into that role is will take some experimenting.

Beginning and ending each half with the starters, the substitution pattern might be something like: (obviously doesn’t account for matchups, injuries, foul trouble, etc.)

• GTJ in, Barnes out (around 6 minutes, as the current rotation does).
• Porter in, Siakam out. (this is the Poeltl+Schroder+shooters lineup… you need them to handle about 2-3 minutes to give Siakam/Barnes enough rest here).
• Barnes and Flynn in, Schroder and Poeltl out.
• Siakam and Achiuwa in, OG and Porter out. (At this point, the lineup is Siakam/Barnes/Flynn/Achiuwa/GTJ. This would be a typical beginning-of-second/fourth-quarter unit and would have an extended shift together. Some nights you might leave Porter in if you think he can manage the minutes).
• Schroder and Boucher in, Barnes and Achuiwa out.
• Barnes and Poeltl in, Siakam and Flynn out.
• Siakam in, Boucher out. (all starters to finish the half)

In this substitution pattern, the keys are that you’re buying a few early minutes of rest for both Siakam and Barnes at the same time with that Schroder, Poeltl, and shooters lineup, and then you can have Siakam+Barnes+bench, or (Siakam/Barnes)+2 starters units the rest of the way. No one-starter+bench units anywhere, or even a Barnes/Siakam + one other starter + bench unit.

The roles of Porter, Achiuwa, and Boucher are somewhat interchangeable based on what the team needs on a given night, but Porter is clearly the most impactful of the three and shooting is particularly necessary when out there with OG, Poeltl and Schroder. Once his shot comes around, Gradey can fit into this mix as well.

This wouldn’t have a huge change in anyone’s total minutes, nor does it significantly change the number of minutes the starters play together. The biggest difference to anybody’s minutes/substitutions is that now, instead of Siakam playing all of the first and third, and a long rest at the start of the 2nd and 4th, you’re breaking him up into 3 shifts per half, similar to what Barnes is currently doing. Rather than a 12 minute shift, six minute rest, six minute shift, Siakam would instead be doing something like 8 minute shift, 4 minute rest, 5 minute shift, 2 minute rest, 5 minute shift. This isn’t ideal for his own energy level; sometimes it will mean taking him out while he’s hot and maybe those nights you just revert to the current substitution pattern; he’s a player that stays hot once he gets hot. But I think there’s a potential big upside to the way the rotation works.

However, this is only possible if you can get about 3 minutes per half of that Schroder+Poeltl+OG+shooters lineup. I’d love to see that lineup in action here and there and just see if there’s something there… or at least it should be explored in practice. There’s some lineup data to back this up. We haven’t seen any minutes for the above lineup, however Schroder+Poeltl+Siakam+GTJ+Porter has been hugely effective +64 net in an 11 minute sample size across 3 games (still small, but not insignificant… by far the best ratings of any unit to play 10+ minutes), and Schroder+Poeltl+GTJ minutes have been really good overall, except when Dick is one of the other players. I think it warrants a deeper look.

Thoughts on this rotation? How would you make an optimal competitive rotation out of the current lineup without significantly increasing anyone’s minutes?