Fan Duel Toronto Raptors

Playing with Data: Raptors Had Tightest Games in NBA

There's not really a point to this except, hey, thanks for some fun ball.

It was an incredibly entertaining Toronto Raptors season. That much is obvious, considering they blew away expectations and made an improbable run to an Atlantic Division title, coming within a point of the second round of the playoffs (and oh, that still hurts, badly).

Prior to the season, I had theorized that the Raptors were likely to play in a lot of close games since they figured to be a middle-of-the-pack team. Here’s a snippet:

The logic is this: great teams should blowout bad teams, beat mediocre teams by less and show parity with other great teams. Bad teams should be blown out by great teams, lose by less to mediocre teams and show parity with other bad teams.

Mediocre teams, then, should play closer games with both great and bad teams and show parity against other mediocre teams.

This is basically the crux of Pythagorean Win Theory, whereby point differential is the best predictor of future performance. Consider the point differentials for different groups of teams based on win totals:

55+6.5
46-552.75
36-450.4
26-35-3.2
26--6.85

It turns out that the more mediocre the teams, the closer the point differential, which is what we would expect.

The leap of faith here is that a close total point differential also means close games. The results from 2012-13 aren’t dramatic, but let’s look at that table again and include the number of games each team had that finished within a five-point margin of victory or defeat:

WinsPoint DiffGms +/- 5
55+6.520.3
46-552.7521
36-450.423.1
26-35-3.221.9
26--6.8521

Again, it’s not extreme, but the best and worst teams seem to play about two fewer ‘close games’ than mid-level teams.

Basically, mid-level teams play in more close games. As it turned out, however, the Raptors were 48-34, not quite the middle of the pack, and had the number nine average margin of victory at 3.24 (17th in absolute terms).

But still, thanks to some crazy fourth-quarter comebacks and a never-say-die, Undertaker-before-my-client-Brock-Lesnar-ended-the-streak-like ability to continue to fight back, the Raptors ended up playing more “clutch” minutes than almost any other team during the regular season.

Only the Washington Wizards, who played 242 minutes that qualified as “clutch time” (the game within five points in the last five minutes of the game), topped the Raptors’ mark of 212. What’s more, no team entered “clutch time” more often than the Raptors, who did so in 54 of their 82 games, four more occasions than any other team. Even leaving aside that the Raptors were generally terrific in the clutch (+8.2 points per 48 minutes), we were treated to more must-see basketball than maybe any other fanbase. Again, two-thirds of their games were within five points at some point in the final five minutes. MLSE owes us all some baby aspirin to help with our blood pressure.

TeamGPMIN+/-+/- per48
Washington Wizards46242-43-8.53
Toronto Raptors54212368.15
Charlotte Bobcats4821071.60
New Orleans Pelicans50204-11-2.59
Golden State Warriors43200153.60
Portland Trail Blazers471988721.09
Atlanta Hawks49196112.69
Miami Heat43189246.10
Orlando Magic42182-86-22.68
Dallas Mavericks481816015.91
Phoenix Suns50181-7-1.86
Boston Celtics49177-98-26.58
Cleveland Cavaliers42176-3-0.82
Milwaukee Bucks40173-97-26.91
Brooklyn Nets3817110.28
New York Knicks39171-22-6.18
Oklahoma City Thunder4316892.57
Memphis Grizzlies4916610229.49
Chicago Bulls371654613.38
Houston Rockets421626218.37
Sacramento Kings42162-56-16.59
Minnesota Timberwolves41160-91-27.30
Los Angeles Clippers431515818.44
Denver Nuggets43148-18-5.84
Utah Jazz40146-42-13.81
Indiana Pacers371405217.83
Philadelphia 76ers34136-6-2.12
Detroit Pistons39134-32-11.46
Los Angeles Lakers38115-8-3.34
San Antonio Spurs301135021.24

And, of course, the Raptors made the playoffs and had a very tight series with Brooklyn, one I can no longer bring myself to talk about.
clutchminteam
That’s pretty cool. End of article. I made my point. We got lucky with some awesome basketball.

Of course, there are more ways we can show this and since it’s late May and why the hell not, let’s have a look.

As mentioned, the Raptors ranked 17th in absolute average game margin, meaning their average game was slightly closer than the median and average team but not appreciably so.
avgmarginbytm
That tells us a bit but not a great deal. For example, the following outcomes could all get you roughly in the range that the Raptors shared with the Cleveland Cavaliers: 72 one-point games and 10 20-point games, or 24 four-point games and 58 three-point games, or 40 10-point games. Same average, far different story in terms of how entertaining the season was.

So, let’s instead look at the league’s spread of outcomes and compare it to the spread that the Raptors experienced.
marginsprdlg
What this clearly shows is that the Raptors played a greater percentage of their games at lower margins than the league on average (though the curve lines have been drawn incredibly crudely and the tails don’t complete). For a better look, we’ll put the margins in bins rather than individual outcomes.
spreadtwo
Just in case you had any modicum of respect left for my curve-drawing from above, I destroyed it here. However, what becomes clear from this more “zoomed in” look at the margin spreads is that the Raptors didn’t actually play a greater proportion of their games to a very narrow finish compared to league average, but it was their absence of blowouts that stood out. Nearly two-thirds of Toronto’s games finished with a single-digit margin, compared to just a hair over half of all games league-wide.

And just for further comparison, here’s the same graph but also including Charlotte and Philadelphia, the team’s with the league’s lowest and highest average game margins, respectively.
spreads3
Again, it shows just how blowout-proof the Raptors were, ranking lower than even the Bobcats, who basically played their entire season to a draw, in blowout games.

So, hey, it wasn’t just fun for narrative purposes, we also happened to get generally more entertaining games than other teams. Hooray for graphs.