Statsketball: Predicting the Mavs' most likely playoff seed, opponent and outcome

After a long hiatus, Statsketball is back! At this point, so far into the season, you've all probably been wondering, "Hey Mavs Moneyball, what gives, we're so close to the playoffs and you guys haven't given us a playoffs update!" Well, guess what: I have all the wonderful playoff information that you need.

At the moment of the publishing of this article -- Thursday, March 5 -- the Mavericks are currently the no. 6 seed in the West, with 40 wins, 22 losses, and a 64.5% win percentage. They're 2.5 games ahead of the no. 7 Spurs, only half a game back of the no. 5 place Clippers, and two games back of Houston for no. 3 seed. The Western Conference playoff picture looks as follows:

So given all of that -- given Dallas' win percentage, how they've performed as a full team with Rondo, given the other team's changes, and given each team's schedule for the rest of the season -- where is Dallas most likely to end the season? What's Dallas' most likely path to the playoffs, and what's going to happen when they get there?

Luckily, I've done all the hard math for you, and my description thereof follows. If you don't particularly care, feel free to skip it!

I ran a simulation of the rest of the season roughly 200 times to tell me what the likely playoff picture would be at the end of the year, what the Mavericks' odds of pulling any given opponent would be, and what the odds of beating any given opponent would be.

The whole goal of the model in prior years was to estimate a team's likelihood of beating an average team and then leveraging that against everyone else, but thanks to this wonderful study by the guys at fivethirthyeight.com, I actually had access to that precise, more exact data found by running regular "probably of winning" estimators throughout games. Or, I have the exact, definite probability that a team will beat an average team!

Because that kind of a metric would naturally bias win percentage though, which is a less effective predictor of success than Net Rating long term, I adjusted the probabilities somewhat by (recent) Points Scored More Than Opponent by 100 Possessions before running the simulation.

Finally, in calculating the odds of the Mavericks actually winning any given matchup, I also adjusted (proportionate to the small sample size) the win probabilities in a series for how well the teams matchup by seeing whether the Mavericks consistently under-performed or over-performed relative to Vegas' expectations for that matchup.

So, what's the most likely end to the season? Let's check out what the sim had to say.

On the playoff race and the end of the season

Dallas is the no. 6 seed now, but I found it really interesting that by far the most often, Dallas ended up falling to the no. 7 rather than holding on or climbing the standings.

What I think was most clear as I watched the results pour in is that the weight of the schedule in this late playoff run is just as -- if not more important than -- the actual win probabilities for forming these rankings.

It turns out that the Clippers and Spurs have really easy ends to their seasons, and the Mavericks have a hard finish. That matters a lot when it comes to the likelihood of teams finishing at various seeds, especially when (in very relative terms) every team from the Clippers to the Mavs is roughly as good as one another.

The Mavericks did end up as the no. 6 seed a healthy amount, and very rarely fell to the no. 8. The odds of Dallas getting any given seed was pretty easy to calculate:

Again, there's a decent amount of room for error here, but I think it's pretty telling -- and probably pretty accurate -- that there's a respectable shot of Dallas falling, but also about a 1 in 6 chance of them climbing up a seed or two.

I think it's funny, also, that the West is so packed that the odds over the Mavs getting the no. 4 seed are only marginally smaller than the odds of getting the no. 5.

At any rate, it seems like the Mavericks are zooming towards a finish in the sixth or seventh seed range, pitting them against most likely either the Grizzlies, the Clippers, or the Rockets. But which of those three is Dallas most likely to play? And what's the most likely result of each series?

Funny you should ask...

Matchup odds

Dallas vs Memphis: Odds of Matchup (33.8%), Odds of Victory (29.5%)

The Mavs' most likely opponent is Memphis, which I actually didn't expect: Memphis is almost locked into the no. 2, with a relatively marginal chance of either dropping or climbing, so I thought the Clippers or Rockets -- who Dallas could face at basically any seeding -- would be more likely. That's not the case.

In some ways a higher chance of getting Memphis is a good thing; the Grizzlies obliterated Dallas once, but the Mavericks have otherwise performed above expectations against the Grizz, and Memphis' system has historically struggled to cover Dirk's shooting.

It's interesting to note that Memphis' odds of winning the series in 7 is actually below their odds of winning in 6, too. Meaning, IF Dallas can get to a 7th game against Memphis, their odds of winning the series skyrocket to much closer to 50-50. And, while Dallas' 30 percent chance of winning seems low, keep in mind that that's three times out of every 10, not that unlikely, and those are fairly decent odds over a series against a team as awesome as Memphis has been all year.

Matchup Notes

• In prior years, Dirk has had a disproportionately huge positive influence on the court versus off against Memphis, but this season, Dirk is shooting way below his career TS% against the Grizzlies and he has been a massive net negative on the court rather than off (by -27.9 points per 100 poss per nba.com). If Memphis has figured out how to successfully guard Dirk, Dallas is in trouble.

• Devin Harris and Chandler Parsons seem to be the key to beating Memphis: they're the only two players on the team who while they're on the court Dallas is competitive, and when either is taken off the court against Memphis the team plummets (to the tune of almost -20 points per 100 possessions). Parsons' playmaking skill against a scrambled defense is exactly what Memphis struggles with, and he's hitting his wide open shots against them -- 66 percent TS%.

• Monta's interesting against Memphis because he personally seems to struggle -- all his advanced (and counting) stats dip heavily against Memphis -- but the team does exponentially better with him on the floor. If he can find a way to brave Memphis' smothering defense, he may find himself being the catalyst for Mavs success.

Dallas vs Los Angeles Clippers: Odds of Matchup (21.6%), Odds of Victory (24.9%)

The Clippers are -- short of the Warriors -- the worst case scenario for Dallas. The Mavs still have a roughly one in four chance of winning that series, which are far better odds than they had against the Spurs last year, for example, and offer plenty of hope.

Still, the Mavs have been decimated by the Clips in every game so far this year and they just have no idea whatsoever how to handle DeAndre Jordan, or Jamal Crawford. Even if the Clippers are not the best team that Dallas could end up with, though they've been criminally underrated all year in most respects. They're the still the worst case scenario for Dallas.

It's unfortunate that it's also the second most likely outcome.

Matchup Notes

• Dallas has a defensive rating of 122 against the Clippers, and a defensive rebounding percentage of only ~71 percent. That means, the Clippers are getting almost 30% of the Mavs' boards as offensive rebounds. Dallas just has absolutely no response to the rebounding tandem of DeAndre Jordan and Blake Griffin, since even Tyson can't stop them and contest against perimeter threats.

• Dallas' starting lineup is a collective -25 points per 100 possessions against LAC. Rajon Rondo is a -45. Monta Ellis is a -36. The Mavs perform exponentially better with every single starter on the bench against the Clippers, per nba.com's stats page.

• The Mavs actually assist and pass extraordinarily well against the Clippers -- they assist on 68% of made shots -- but they can't convert that into points or good shots. They log an offensive rating of a smidge less than 105, and a below average TS% against the Clippers. You have to imagine that some of that is luck that will get corrected.

Dallas vs. Houston: Odds of Matchup (18.5%), Odds of Victory (61%)

The Rockets are the only team that the Mavs are favored against by my model, which may seem odd, but let me explain.

The Rockets, for one, have been a mite overrated for a while. Their insanely hot start gave them a nice cushion for a lot of the year, and though they've picked it back up behind a top five defense even without Dwight Howard, this is still a team that's not as good as either its record or point differential over the season. They're also a high variance team, which tends to be risky heading into the playoffs.

On top of that, the Mavericks have outplayed Vegas expectations in every single game against the Rockets thus far, indicating that, at least to an extent, the Mavs have a natural edge in what is already a close matchup.

Matchup Notes

• The Mavericks actually play a grind it out game with the Rockets: both Dallas scores and allows fewer than 100 points per possessions against Houston. When you think "Dallas vs Houston" you don't typically imagine a defensive slugfest, but here we are.

• Perhaps unsurprisingly, Chandler Parsons, Devin Harris, and Al-Farouq Aminu are the biggest on/off positives for Dallas against the Rockets, probably due mostly to the effectiveness of bench units containing all three players. Parsons in particular gets his revenge-game on, with the team's second highest Offensive Rating behind Rondo (?) and a TS% of over 60%.

• Monta doesn't have impressive stats against the Rockets, but the Mavericks shoot almost nine percent better, pass two percent better, and rebound five percent better with him on the court as opposed to off. Interestingly, Rajon Rondo's personal stats are worse, but the team has a similarly drastic improvement in shooting, passing and rebounding ... but Rondo still have a very negative plus/minus while Monta does not.

Dallas vs. Portland: Odds of Matchup (12.3%); Odds of Victory (45.5%)

This outcome fits with conventional wisdom, I think: games against the Blazers this year have seems to be pretty much 50-50 affairs, and at a 45%-55% split over a whole 7 game series, we're looking at something pretty darn similar here.

I think it's hard to argue that the Blazers haven't been the better team this year, and so they should probably be favored somewhat, but if this unlikely series were to somehow come to fruition -- if the Mavs got the 5 and Portland got the 4, or vice versa -- this would be a fun, hyper-competitive, and unpredictable series. This is probably one to hope for.

Matchup Notes

• Despite the fact that the Mavericks have been (and should be) competitive with the Blazers, the Mavericks' per 100 possessions point differential against the Blazers is almost -10, and the best individual player against the Blazers is Parsons clocking in at -0.6. There's not a single Maverick with a positive Net Rating against the Blazers, and Dallas is shooting an abysmal 47 percent TS% against Portland.

• The Blazers are the one playoff team that Dirk has excelled against. He's shooting an above average 56 percent TS% against Portland, and the team as a whole has rebounded, shot, and passed by over 10 percent better in every category when he's on the floor as opposed to off. All of this despite Dirk's usage against Portland being noticeably higher than against most other West playoff teams.

• While Dirk excels against Portland, Monta has struggled. He's shooting a truly horrific 43.7 percent against Portland on a substantial number of shots, while barely registering more assists than turnovers. Most of this reverses in the fourth quarter: Dallas has a fourth quarter Net Rating of 8.2 in the fourth against Portland, and Monta's percentages explode.

Dallas vs. Golden State: Odds of Matchup (8.7%); Odds of Victory (5%)

This is basketball Ragnarok; this series would be the end days. Either: the Warriors come in and completely outclass the Mavs, making them look like a sad, total joke ... or the no. 8 seeded Mavs upset the no. 1 seeded Warriors who seem predestined to win it all, in a brutal and terrific reversal of 2007.

Part of me is rooting for this series to happen if only because Dallas winning would probably be the most satisfying basketball thing I've ever seen. Sadly, there's only a nine percent chance of drawing the Dubs, though in all practical senses this is a series that Dallas wants to avoid at all costs.

Matchup Notes

• The Warriors are the better team by a lot

• The Mavs are the worse team by a lot

• The Warriors are an incredible, extraordinary team

• The Warriors will probably win this series

• But if they didn't, how awesome would that be? Pretty awesome right?

Dallas vs. San Antonio: Odds of Matchup (4.6%); Odds of Victory (48%)

No one wants to see the Spurs as their first round matchup in case they just happen to find their next gear, but the truth is the Spurs have been somewhere in between "better than people think they are" and "not as good as people expect them to be."

San Antonio has actually been pretty darn good since getting the whole team back from injury-- and probably good enough to make the second round of a playoff series against most opponents -- but "pretty darn good" isn't going to cut it in the long haul for them.

It's not a shock then that this matchup looks to be roughly an even affair, because in a way, these are similar teams. They're both storied teams with championship aspirations who are also currently underachievers, pulled down by the struggles of their star point guard, but with a high, dangerous ceiling.

Matchup Notes

• If I have to watch Ginobli flop for another whole series and then go bonkers in a game 7 again, I swear to God...

• The Mavericks only have one player with a massively positive Net Rating while he's on the court who's so important that Dallas' point differential per 100 possessions plummets into the negatives while he's off the court ... Chandler Parsons. Have you noticed that he seems to be important everywhere? Monta's impact on the team against the Spurs is hugely influential here as well.

• The Mavericks have yet to play a fully healthy Spurs team, but satisfyingly, Manu Ginobli has the second worst impact of anyone on the Spurs against Dallas so far, behind only their struggling point guard, Tony Parker.