Michael Jordan -- Complete Playoff Shooting Data

[Djoker]

Merry Christmas to y'all!

I'm proud to say that we now have shooting data for Michael Jordan's entire playoff career.

A lion's share of the work for the 1988-1992 period was done by Dipper13 here. He had three instances of missing data (total of 35 FGA) that I filled in -- 1988 Round 1 Game 3 (missing the entire game 25 FGA), 1990 Round 1 Game 4 (missing the first half 9 FGA), and 1991 Round 1 Game 1 (missing 1 FGA).

In addition to completing the 1988-1992 dataset, I logged the entire 1985, 1986, 1987, 1993, 1995 and 1996 postseasons.

1985

Spoiler:

1986

Spoiler:

1987

Spoiler:

1988

Spoiler:

1989

Spoiler:

1990

Spoiler:

1991

Spoiler:

1992

Spoiler:

1993

Spoiler:

1995

Spoiler:

1996

Spoiler:

1997

Spoiler:

1998

Spoiler:

Career Playoff Summary

Spoiler:

Finals Only

Spoiler:

And here are a few individual games too.

63 Points - 1986 Game 2 at Boston -- PLAYOFF RECORD

Spoiler:

At Rim: 5/9 FG (55.6%)
In Paint (Overall): 11/21 FG (52.4%)
Mid-Range: 11/20 FG (55.0%)
3 Point: 0/0 FG

38 Points - 1988 Game 3 at Cleveland

Spoiler:

At Rim: 5/8 FG (62.5%)
In Paint (Overall): 2/6 FG (33.3%)
Mid-Range: 5/11 FG (45.5%)
3 Point: 0/0 FG

45 Points - 1996 Game 4 at Orlando

Spoiler:

At Rim: 2/3 FG (66.7%)
In Paint (Overall): 1/3 FG (33.3%)
Mid-Range: 10/13 FG (76.9%)
3 Point: 3/4 FG (75.0%)

[Special_Puppy]

The midrange shooting stands out to me. It was a literal majority of his shot attempts and he converted at a pretty amazing clip of 45.6% on massive volume.

[Peregrine01]

The mid-range gets a lot of flack now but back in that era it made sense. With how packed the paint was back then, it probably made sense to stop and pop. And given how important the possession game was back then, it made sense to play in a way that maximizes possessions and limits turnovers. The 2nd Bulls 3-peat was really built off of offensive rebounding and turnover economy on the offense end and stifling perimeter defense on the other end.

Jordan playing the way he did helped optimize the Bulls in those key areas.

[tsherkin]

Merry Christmas to y'all!

I'm proud to say that we now have shooting data for Michael Jordan's entire playoff career.

Damn, that's a lot of work. Thanks for posting this!

[falcolombardi]

65% at rim
39% 3-10 feet
46% 10-23 feet

Ngl while they are obviouslt very great numbers they demithify jordan being a goat tier rim finisher (we knew this but is nice to have the data) or a clear goat mid range scorer

[tsherkin]

65% at rim
39% 3-10 feet
46% 10-23 feet

Ngl while they are obviouslt very great numbers they demithify jordan being a goat tier rim finisher (we knew this but is nice to have the data) or a clear goat mid range scorer

Worth measuring that against the defenses he was facing on average. 80s Boston, the Pistons. 90s Knicks, Heat, Jazz. There's a LOT of high-end defense there, more than there were soft teams.

[capfan33]

65% at rim
39% 3-10 feet
46% 10-23 feet

Ngl while they are obviouslt very great numbers they demithify jordan being a goat tier rim finisher (we knew this but is nice to have the data) or a clear goat mid range scorer

Worth measuring that against the defenses he was facing on average. 80s Boston, the Pistons. 90s Knicks, Heat, Jazz. There's a LOT of high-end defense there, more than there were soft teams.

Even relative to era I'm not really sure Jordan faced significantly tougher defenses than other all-time greats. Also wouldn't include the Jazz in that, even in 97 they were only 10th, not noteworthy in the grand scheme of things. Also would echo falco, fantastic numbers especially considering the volume, but nothing out of this world.

[tsherkin]

Even relative to era I'm not really sure Jordan faced significantly tougher defenses than other all-time greats. Also wouldn't include the Jazz in that, even in 97 they were only 10th, not noteworthy in the grand scheme of things. Also would echo falco, fantastic numbers especially considering the volume, but nothing out of this world.

They were a physical team with a large-ish frontline and some good perimeter defenders. Their O was better than their D, obviously, and they didn't stand up to teams with Ewing or Mourning sure, but they were no pushovers.

The perimeter shooting IS out of this world on the volume he was producing (45.6%, remember), and 65% at the rim in-era was pretty good. His short game is the part that's underwhelming, and you'll note it represented 12.6% of his total volume.

Food for thought.

[lessthanjake]

65% at rim
39% 3-10 feet
46% 10-23 feet

Ngl while they are obviouslt very great numbers they demithify jordan being a goat tier rim finisher (we knew this but is nice to have the data) or a clear goat mid range scorer

I dunno—I’d say that that playoff mid-range scoring looks pretty great to me. We can roughly derive a similar number for other players by using some back-of-the-napkin math on BBREF and taking a weighted average of a player’s playoff FG% on shots from 10-16 feet and 16ft-3P, weighted by what percent of their shots they took from those categories. Dirk Nowitzki is someone who sticks out to me as a GOAT-level mid-range shooter. And Dirk’s playoff mid-range FG% is essentially exactly the same as Jordan’s (just below 46%), and that’s with less volume than Jordan. Durant’s is slightly higher (around 47%), but again with lower volume. Steph Curry’s is also a little higher (should be right above 46%), but his mid-range volume isn’t very high at all. Kobe’s playoff mid-range FG% was around 42%, so that’s a good deal lower. Steve Nash was one of the greatest shooters ever, but had a playoff mid-range FG% of around exactly 45%, while obviously on much lower volume. [EDIT: Note that, as per my post just below this, there’s also good reason to believe that Jordan’s midrange FG% would be higher using this BBREF method than it is in Djoker’s data, since the two methods are a little different and in 1997 and 1998 where we have both sets of data, Jordan does better in the BBREF method than in Djoker’s data].

I think there’s a very good argument that Jordan’s combination of mid-range FG% and volume makes him the GOAT mid-range scorer, and I think this data backs that up—even if the percentage itself is not unprecedented (and so perhaps you can plausibly say it doesn’t show him being the “clear goat mid range scorer”)

That said, I think with Jordan, just looking at the mid-range FG% and volume doesn’t quite tell the whole story about why his mid-range game was so good offensively. A big reason it was so good is that Jordan could get to his spots and get his shot off with very minimal numbers of turnovers. This was a really big deal. For instance, if Jordan makes 46% of his mid-range shots and Durant makes 47% of his mid-range shots, you’d still expect to have a more efficient offense with Jordan shooting mid-range shots because Jordan is going to turn the ball over significantly less in the process of getting those shots. That’s true for all the above-mentioned guys except maybe Dirk.

Another thing worth noting is that I think the calculus of mid-range shots was different in Jordan’s era than in the more recent eras that the above-discussed guys played in, in a way that I think makes Jordan’s numbers even more impressive. More recent eras have had more spacing (with that additional spacing obviously getting more and more pronounced the more recent we get). That makes it easier to get a good look on a jumper. But it also makes offensive rebounding less likely. So I think there’s a bit of trade-off here, where, all else being equal, we’d expect players in more recent eras to have a higher FG% on their mid-range shots, but to have that balanced out in terms of overall team offensive efficiency by teams in more recent eras getting fewer offensive rebounds on those shots. In that context, Jordan’s mid-range FG% looks even more impressive IMO.

[lessthanjake]

.

Thanks for this! This is great stuff!

One question I have is about how we think this maps onto BBREF data.

In my above post, I tried to derive similar numbers for other players, using a weighted average of BBREF’s 10-16 foot and 16ft-3P field goal percentages for those players. It occurs to me that what’s being measured here is subtly different, since “Paint (w/o Rim)” is not *exactly* the same as BBREF’s 3-10 foot category. Specifically, “Paint (w/o Rim)” is going to include some shots further than 10 feet out that are pretty straight on and close to the FT line (since the FT line is further than 10 feet out), while it is not going to include some shots that are less than 10 feet near the sides of the basket (since the paint extends less than 10 feet out in either direction). Of course, I realize why you tracked using this instead—since it’s surely *way* easier to figure out which category the shot should go into using your method.

So what I’m curious about is whether we’d expect “Midrange” as defined here to be above or below what we’d find if we combined BBREF’s 10-16 and 16-3P categories. To try to answer that, I compared Jordan’s own numbers in the years we have both sets of data for him (i.e. 1997 and 1998). In the 1997 playoffs, the number we’d derive using the BBREF method would be a 46.3% FG%, while your data has a 43.3% midrange FG%. In the 1998 playoffs, the number we’d derive using the BBREF method would be 40.3%, while your data has a 39.8% midrange FG%. So, at least for Jordan, your data seems to be lower than what BBREF would show if we combined the 10-16 and 16-3P categories. I’m not *sure* if this carries over to other players, since it might just reflect idiosyncrasies for Jordan himself in terms of what spots he was best from (or perhaps it’s just noise). But I guess my baseline assumption is that this method probably slightly undersells a bit how good Jordan would look if we instead tallied “midrange” as being 10ft-3P instead of it being defined as 2-point-shots that aren’t in the paint.

[Djoker]

.

Thanks for this! This is great stuff!

One question I have is about how we think this maps onto BBREF data.

In my above post, I tried to derive similar numbers for other players, using a weighted average of BBREF’s 10-16 foot and 16ft-3P field goal percentages for those players. It occurs to me that what’s being measured here is subtly different, since “Paint (w/o Rim)” is not *exactly* the same as BBREF’s 3-10 foot category. Specifically, “Paint (w/o Rim)” is going to include some shots further than 10 feet out that are pretty straight on and close to the FT line (since the FT line is further than 10 feet out), while it is not going to include some shots that are less than 10 feet near the sides of the basket (since the paint extends less than 10 feet out in either direction). Of course, I realize why you tracked using this instead—since it’s surely *way* easier to figure out which category the shot should go into using your method.

So what I’m curious about is whether we’d expect “Midrange” as defined here to be above or below what we’d find if we combined BBREF’s 10-16 and 16-3P categories. To try to answer that, I compared Jordan’s own numbers in the years we have both sets of data for him (i.e. 1997 and 1998). In the 1997 playoffs, the number we’d derive using the BBREF method would be a 46.3% FG%, while your data has a 43.3% midrange FG%. In the 1998 playoffs, the number we’d derive using the BBREF method would be 40.3%, while your data has a 39.8% midrange FG%. So, at least for Jordan, your data seems to be lower than what BBREF would show if we combined the 10-16 and 16-3P categories. I’m not *sure* if this carries over to other players, since it might just reflect idiosyncrasies for Jordan himself in terms of what spots he was best from (or perhaps it’s just noise). But I guess my baseline assumption is that this method probably slightly undersells a bit how good Jordan would look if we instead tallied “midrange” as being 10ft-3P instead of it being defined as 2-point-shots that aren’t in the paint.

Why not compare to data on NBA.com which breaks it down by both distance (same as B-Ref) but also by zone like I did here? You're right that I did it by zone rather than distance because it's tough to see even with rewinding (and I did rewind quite a lot while tracking) if a jumper was taken from say 9 feet or 10 feet away from the basket.

Here is Jordan's data for the 1996-97 postseason from NBA.com and you can see the splits both by zone and by distance:

https://www.nba.com/stats/player/893/shooting?Season=1996-97&SeasonType=Playoffs

Note the weird quirk on NBA.com (only in 1997) where some midrange shots are in fact worth 3 points because of the shorter 3pt line! I label only 2pt midrange shots as midrange and the 3pt midrange as 3pt shots.

[homecourtloss]

65% at rim
39% 3-10 feet
46% 10-23 feet

Ngl while they are obviouslt very great numbers they demithify jordan being a goat tier rim finisher (we knew this but is nice to have the data) or a clear goat mid range scorer

Yes, nothing really stands out here relative to what we already knew — if anything, they’re a little bit underwhelming, especially the playoff numbers.

If anyone has ever gone and done any tracking work as a few of the posters here have, we know that the scoring numbers were boosted by a lot and a lot of touch fouls on the perimeter, especially in the later years that boosted the scoring efficacy. We already knew this because of the rEFG% numbers that were underwhelming relative to the mythologizing.

[lessthanjake]

.

Thanks for this! This is great stuff!

One question I have is about how we think this maps onto BBREF data.

In my above post, I tried to derive similar numbers for other players, using a weighted average of BBREF’s 10-16 foot and 16ft-3P field goal percentages for those players. It occurs to me that what’s being measured here is subtly different, since “Paint (w/o Rim)” is not *exactly* the same as BBREF’s 3-10 foot category. Specifically, “Paint (w/o Rim)” is going to include some shots further than 10 feet out that are pretty straight on and close to the FT line (since the FT line is further than 10 feet out), while it is not going to include some shots that are less than 10 feet near the sides of the basket (since the paint extends less than 10 feet out in either direction). Of course, I realize why you tracked using this instead—since it’s surely *way* easier to figure out which category the shot should go into using your method.

So what I’m curious about is whether we’d expect “Midrange” as defined here to be above or below what we’d find if we combined BBREF’s 10-16 and 16-3P categories. To try to answer that, I compared Jordan’s own numbers in the years we have both sets of data for him (i.e. 1997 and 1998). In the 1997 playoffs, the number we’d derive using the BBREF method would be a 46.3% FG%, while your data has a 43.3% midrange FG%. In the 1998 playoffs, the number we’d derive using the BBREF method would be 40.3%, while your data has a 39.8% midrange FG%. So, at least for Jordan, your data seems to be lower than what BBREF would show if we combined the 10-16 and 16-3P categories. I’m not *sure* if this carries over to other players, since it might just reflect idiosyncrasies for Jordan himself in terms of what spots he was best from (or perhaps it’s just noise). But I guess my baseline assumption is that this method probably slightly undersells a bit how good Jordan would look if we instead tallied “midrange” as being 10ft-3P instead of it being defined as 2-point-shots that aren’t in the paint.

Why not compare to data on NBA.com which breaks it down by both distance (same as B-Ref) but also by zone like I did here? You're right that I did it by zone rather than distance because it's tough to see even with rewinding (and I did rewind quite a lot while tracking) if a jumper was taken from say 9 feet or 10 feet away from the basket.

Here is Jordan's data for the 1996-97 postseason from NBA.com and you can see the splits both by zone and by distance:

https://www.nba.com/stats/player/893/shooting?Season=1996-97&SeasonType=Playoffs

Note the weird quirk on NBA.com (only in 1997) where some midrange shots are in fact worth 3 points because of the shorter 3pt line! I label only 2pt midrange shots as midrange and the 3pt midrange as 3pt shots.

Good point—I’d not realized NBA.com had splits that were the same as what you tallied. The downside of that is just that they only seem to have season-by-season numbers rather than career numbers, so it does take a bit of time to tally up career numbers for someone (though obviously absolutely nothing in comparison to how much time it took you to tally all this for Jordan!). That said, I did just tally that for Dirk, and he had a 45.6% playoff midrange FG%. That’s exactly the same as Jordan, but it’s on notably lower volume. Which I’d say looks pretty great for Jordan, since Dirk is definitely a contender for GOAT midrange shooter.

[Peregrine01]

65% at rim
39% 3-10 feet
46% 10-23 feet

Ngl while they are obviouslt very great numbers they demithify jordan being a goat tier rim finisher (we knew this but is nice to have the data) or a clear goat mid range scorer

I dunno—I’d say that that playoff mid-range scoring looks pretty great to me. We can roughly derive a similar number for other players by using some back-of-the-napkin math on BBREF and taking a weighted average of a player’s playoff FG% on shots from 10-16 feet and 16ft-3P, weighted by what percent of their shots they took from those categories. Dirk Nowitzki is someone who sticks out to me as a GOAT-level mid-range shooter. And Dirk’s playoff mid-range FG% is essentially exactly the same as Jordan’s (just below 46%), and that’s with less volume than Jordan. Durant’s is slightly higher (around 47%), but again with lower volume. Steph Curry’s is also a little higher (should be right above 46%), but his mid-range volume isn’t very high at all. Kobe’s playoff mid-range FG% was around 42%, so that’s a good deal lower. Steve Nash was one of the greatest shooters ever, but had a playoff mid-range FG% of around exactly 45%, while obviously on much lower volume. [EDIT: Note that, as per my post just below this, there’s also good reason to believe that Jordan’s midrange FG% would be higher using this BBREF method than it is in Djoker’s data, since the two methods are a little different and in 1997 and 1998 where we have both sets of data, Jordan does better in the BBREF method than in Djoker’s data].

I think there’s a very good argument that Jordan’s combination of mid-range FG% and volume makes him the GOAT mid-range scorer, and I think this data backs that up—even if the percentage itself is not unprecedented (and so perhaps you can plausibly say it doesn’t show him being the “clear goat mid range scorer”)

That said, I think with Jordan, just looking at the mid-range FG% and volume doesn’t quite tell the whole story about why his mid-range game was so good offensively. A big reason it was so good is that Jordan could get to his spots and get his shot off with very minimal numbers of turnovers. This was a really big deal. For instance, if Jordan makes 46% of his mid-range shots and Durant makes 47% of his mid-range shots, you’d still expect to have a more efficient offense with Jordan shooting mid-range shots because Jordan is going to turn the ball over significantly less in the process of getting those shots. That’s true for all the above-mentioned guys except maybe Dirk.

Another thing worth noting is that I think the calculus of mid-range shots was different in Jordan’s era than in the more recent eras that the above-discussed guys played in, in a way that I think makes Jordan’s numbers even more impressive. More recent eras have had more spacing (with that additional spacing obviously getting more and more pronounced the more recent we get). That makes it easier to get a good look on a jumper. But it also makes offensive rebounding less likely. So I think there’s a bit of trade-off here, where, all else being equal, we’d expect players in more recent eras to have a higher FG% on their mid-range shots, but to have that balanced out in terms of overall team offensive efficiency by teams in more recent eras getting fewer offensive rebounds on those shots. In that context, Jordan’s mid-range FG% looks even more impressive IMO.

What strikes me is the consistency - he was >45% from mid-range on huge volume for basically the entire stretch from 88-97. In the playoffs where games often come down to a few possessions, consistency and counters matter a lot and the problems of variance rears its ugly head. That you can get the ball to Jordan in the mid-post with the shot clock running down and get a consistent 45% fadeaway with low turnover probability is a huge advantage that a lot of other teams didn’t have. The Jazz series were great examples - when the games got really mucked up Malone had a hell of a time trying to score in isolation. Jordan was just a lot more dynamic and robust.

[trex_8063]

Very nice to have this, thanks!

A point of clarity: is "paint w/out rim" to be taken very literally (literally in the paint, and not a function of shot distance)?

e.g. a 9-foot shot on the baseline [near the block, just outside the key] will be counted as "mid-range", but a 14-foot shot just in front of the FT-line would be counted as "paint w/out rim".

[1993Playoffs]

Honestly expected a bit better, but considering the volume it’s good

[Djoker]

Very nice to have this, thanks!

A point of clarity: is "paint w/out rim" to be taken very literally (literally in the paint, and not a function of shot distance)?

e.g. a 9-foot shot on the baseline [near the block, just outside the key] will be counted as "mid-range", but a 14-foot shot just in front of the FT-line would be counted as "paint w/out rim".

Yes that's right. The paint w/o rim is literally the paint without the restricted area.

[Djoker]

Due to popular demand, here is the playoff midrange data for some legends. MJ looks pretty good!

Player: FG/FGA (%); FGA/game

Michael Jordan: 1040/2279 (45.6%); 12.7
Demar Derozan: 208/559 (37.2%); 8.9
Dirk Nowitzki: 563/1230 (45.8%); 8.5
Kobe Bryant: 759/1854 (40.9%); 8.4
Kevin Durant: 601/1238 (48.5%); 7.3
Chris Paul: 374/743 (50.3%); 5.0
Steve Nash: 232/517 (44.9%); 4.3
Stephen Curry: 225/487 (46.2%); 3.3

[capfan33]

Even relative to era I'm not really sure Jordan faced significantly tougher defenses than other all-time greats. Also wouldn't include the Jazz in that, even in 97 they were only 10th, not noteworthy in the grand scheme of things. Also would echo falco, fantastic numbers especially considering the volume, but nothing out of this world.

They were a physical team with a large-ish frontline and some good perimeter defenders. Their O was better than their D, obviously, and they didn't stand up to teams with Ewing or Mourning sure, but they were no pushovers.

The perimeter shooting IS out of this world on the volume he was producing (45.6%, remember), and 65% at the rim in-era was pretty good. His short game is the part that's underwhelming, and you'll note it represented 12.6% of his total volume.

Food for thought.

I mean sure they were good but they aren't in the same realm as the Pistons or Knicks who were legit ATG defenses, especially in 98.

And based off the data I would feel reasonably comfortable calling him the GOAT midrange shooter, mostly due to volume, but Dippers data originally had him at like 52%, which I always thought was probably too high, but maybe my expectations were unrealistic. And would love peer data on the rim finishing but regardless after 91 his rim finishing merely looks good as opposed to excellent, and honestly worse than I would've expected. Based off this, I don't really think he has an argument as the best slasher ever.

[lessthanjake]

One thing I’d be interested in is how often different players get fouled on their mid-range shots. I don’t think this is data that’s really accessible anywhere (and it would take prohibitively long to actually count it up for players), so this is just an aspirational wish. I definitely don’t know for sure but my sense is that MJ drew more fouls on his mid-range shots than most other mid-range greats did. Which is something that matters since it definitely affects the efficiency of players in the mid-range. It’s probably not a *huge* factor since I doubt anyone was drawing a ton of fouls on midrange shots, but I guess I don’t really know how big the range is between players on that.