APBRmetrics

[ishmael]

Long time reader, first real post (love the forums!). OK, so it is still summer and we need new topics of conversation. Here is some fodder for the old timers and stat geeks out there!

Last night I was having a conversation with my Dad and the topic came up "How many shots did Bill Russell actually block per game?" Now I know that the NBA did not officially start keeping tracks of blocks per game until 1973, but this does not mean it is impossible to go back and figure out how many shots Bill Russell blocked. We have video of complete playoff games from that era. We also have data on modern players.

Here are the top 10 all time leaders in playoff blocks (not bpg) from basketball-reference.com:
Name -- Last name -- Playoff blocks -- Playoff games -- Playoff blocks per game -- Regular season career bpg
Hakeem -- Olajuwon -- 472 -- 145 -- 3.26 -- 3.1
Tim -- Duncan -- 380 -- 138 -- 2.75 -- 2.5
Dikembe -- Mutombo -- 238 -- 93 -- 2.56 -- 2.8
David -- Robinson -- 312 -- 123 -- 2.54 -- 3
Shaquille -- O'Neal -- 433 -- 198 -- 2.19 -- 2.5
Patrick -- Ewing -- 303 -- 139 -- 2.18 -- 2.4
Kareem -- Abdul-Jabbar -- 476 -- 237 -- 2.01 -- 2.6
Julius -- Erving -- 239 -- 141 -- 1.70 -- 1.5
Robert -- Parish -- 309 -- 184 -- 1.68 -- 1.5
Kevin -- McHale -- 281 -- 169 -- 1.66 -- 1.7

If I run those two variables (bpg career and playoffs) as a regression I get an R^2 and coefficient above .76 -- pretty good. If I ran that for all players all seasons post 1973, I bet the R^2 would be more like .90 or higher.

So my question is this: has anyone ever sat down and gamecharted the ESPN classic games and actually counted Bill Russell blocks?

This seems like something that would be fairly easy to do and provide a fun (and possibly useful) stat. The only players in NBA history to average greater than 3 blocks per game for their career are:
Mark Eaton
Manute Bol
Hakeem Olajuwon

My Dad claims that there is no way Russell averaged less than 4. How many Russell full games are available on tape and who has enough time to count blocks? If someone else does that, I will be happy to provide more in-depth analysis on blocks per game by all players and also season to season correlations (probably even stronger than career numbers!).

(X-posted to Realgm)

[Conan the Librarian]

I have no idea about his NBA totals, but Bill Russell himself estimates that he may have blocked as many as 15 shots a game in college. That's pretty staggering, but not that surprising as no one had ever seen a blocked shot before Russell, so nobody had any idea how to avoid it or draw contact or what have you.

[thref23]

The only relevant links I can find are -

http://72.14.253.104/search?q=cache:obBqgNFy1HYJ:baseballcrank.com/archives2/basketball/+%22bill+russell%22+blocked+shots+%22per+game%22+estimate&hl=en&ct=clnk&cd=11&gl=us

http://72.14.253.104/search?q=cache:dvDwX58C0QkJ:entertainment.howstuffworks.com/bill-russell-at.htm+%22bill+russell%22+blocked+shots+%22per+game%22+estimate&hl=en&ct=clnk&cd=1&gl=us

[Mike G]

Russell averaged only 2.7 fouls per game, while playing 42 mpg.
No player (going over 1000 minutes) has managed to get more than 1.6 blocks per foul.
Just 25 players are known to have managed 1.0 Blk/PF for a season.

[FrontRange]

Mike,

Given that Chamberlain never fouled out of game, I suspect fouls were called much differently in the 60's than in the 70, 80, and 90s. Do you have any test (perhaps rebounds per PF) to see if a great center like Wilt or Russell might be off the scale on PF/blocks?

Also do you need to make an adjustment for the higher number of shots per foul in the 60's compared to the other era's when blocks were recorded? Off the top of my head I would guess Russell had 30% more chance per game than Bol and I doubt that fouls were 30% higher.

[Mike G]

Maybe it was like blocking a punt (or FG) in football: if you get a piece of the ball, you may go ahead and run over the kicker.

In Russell's day, the ratio of FTA/FGA ranged from 34-38%. It's been 30-33% since the late '80s. Ratio of PF/FGA was 23-27%; lately it's 25-29%.

PF/team/game was 25-27 in Russell's time, 21-23 in the last 17 years.

I'm not convinced that with 30% more FGA there are 30% more block opportunities. If your opponent has a great shotblocker, your strategy
might be to take the first available shot; even hurrying down the floor to shoot before he can get back.

The only way Russ (or Wilt) got over 4 Blk/G is if they were (a) superhuman beyond anyone that has come along since (which I don't
believe), or (b) if the level of their competition was so lame they just couldn't adjust. If B is it, then I'd have to say they might have had
'inflated' totals that they would not replicate had they been born 40 (or 20) years later.
_________________
40% of all statistics are wrong.

[Conan the Librarian]

[bchaikin]

The only way Russ (or Wilt) got over 4 Blk/G is if they were (a) superhuman beyond anyone that has come along since (which I don't believe), or (b) if the level of their competition was so lame they just couldn't adjust.

here's something i posted to the old APBR discussion group back in 11/2005 about this topic:

-------------------------------------------------------------------------------------

here's an easy calculation i've used for trying to estimate blocked shots prior to 73-74 for the nba for players like chamberlain and russell:

for this i'm assuming those two were not the major shot blockers manute bol, mark eaton, and shawn bradley were because each of these players were 7'4" and above, and exerted most of their effort on defense only. they could have been but i'm guessing they weren't. bol routinely blocked 10.0% to 10.5% of all opponents FGA when he was on the floor, eaton 8% - 10% his first four seasons in the league (6% - 7% his next 3 seasons), and bradley 7% - 8% his first 8-9 years in the league. defensive specialists like tree rollins and george johnson both blocked at or close to a high of 8% of all opponents shots in a season, but were usually closer to 6% - 7%...

but more complete centers, but still very good shot blockers, like david robinson and hakeem olajuwon blocked at most 6.0% - to 6.5% of all opponents shots in a single season, and over their best 10 year stretch, olajuwon averaged about 5.5% BS and robinson 5.2% BS per 100 opponent FGA...

olajuwon's best 10 year stretch of shot blocking came from 84-85 to 93-94, when the NBA averaged 87.5 FGA/g, and robinson's best 10 year stretch (actually 11 as he missed the 96-97 season) was 89-90 through 99-00 when the league averaged 83.1 FGA/g. both players averaged 37 min/g during this time...

now for the 10 year stretch of 1959-60 to 1968-69, when both chamberlain and russell played pretty much most of their time in the league, the NBA averaged 102.7 FGA/g (105.0 FGA/g the first half of the decade and 100.9 FGA/g the 2nd half) due to a faster game pace. i would consider chamberlain more of an offensive player like olajuwon/robinson, and russell's pts/min scored are close to those of tree rollins for rollins's first 3 seasons in the league (rollins was at 0.34 - 0.36 pts/min, russell at 0.37 - 0.43 pts/min but in a faster paced game), so i'd consider russell more of a defensive specialist. so to calculate blocked shots for someone like robinson or olajuwon in a good season of 6% BS would look something like this:

(6 BS) / (100 FGA) x 83.1 FGA/g x (37min/g played) / (48min/g) x 82 games = 315 blocks

and these two players did in fact block 300+ shots in a season three times each. so if chamberlain, who played 47 min/g, and in a much faster paced game, blocked at most 6% of the opponents FGA in a season like robinson or olajuwon he could have blocked as much as:

(6 BS) / (100 FGA) x 102.7 FGA/g x (47min/g played) / (48min/g) x 82 games = 495 blocks

or close to 500 blocks in a good season, or about 6 blocked shots per game, because he played so many minutes. if he did in fact block as much as 500 shots in a season it'd be easy to see a high of 20-25 blocks in a single game occuring..

as for russell if he blocked at most say 7% of all opponents FGA like a good (not great) tree rollins or george johnson year, but still much less than the best shot blocking rates of mark eaton or manute bol, and playing 43 min/g as he did, he could block:

(7 BS) / (100 FGA) x 102.7 FGA/g x (43min/g played) / (48min/g) x 82 games = 528 blocks

or close to 530 blocks in a good season, again with 20-25 blocks in a single game being a realistic high mark for BS in a single game...

again i can't realistically see either chamberlain or russell blocking shots at the rate that bol or eaton did, but if they did at rates similar to robinson, olajuwon, or possibly someone like tree rollins (and i'm guessing chamberlain/russell were each among the best shot blockers of their era), then because they each played so many minutes, i think 500 BS in a season for either of them in a good season is a reasonable guesstimate...

even if each blocked shots at the rate of just 5% (5 BS per 100 opponent FGA), because they played so many minutes and in a faster paced game compared to today or even the past 20 years their numbers would still be close to 400 blocks:

chamberlain
(5 BS) / (100 FGA) x 102.7 FGA/g x (47min/g played) / (48min/g) x 82 games = 412 blocks

russell
(5 BS) / (100 FGA) x 102.7 FGA/g x (43min/g played) / (48min/g) x 82 games = 377 blocks

or close to 4.5 - 5.0 bs/g as a conservative guesstimate...

[FrontRange]

Much as I hate to admit it, I find Bob's evidence much more compelling than Mike's this time. (Well actually I like to see Bob get people riled up).

Out of curiosity Bob, would you consider a counter arguement to Eaton or Bol representing an unachievable rate for Russell? My understanding is that he revolutionized defense and given the average height in the league (and I assume less athletic players), couldn't his relative shot blocking ability approached the top of all time? Anybobdy ever chart any of Russell games if available to see what he rate might have been?

[Mike G]

Actually, I like Bob's reasoning as well. But I've already brought up a couple of possible restraints on his extrapolations.
We aren't even that far apart (4.0 max vs 5.0 max)

One is the Blk/PF ratio. Even shotblocking specialists max out at 1.60 (Bol, at his youngest and quickest). Five blocks per 2.0 PF (2.50) is
unapproached in ensuing decades. For guys who also were world-class rebounders, it's like supposing Olajuwon and Robinson could either (1)
double their shotblocking rate without fouling more, or (2) cut their
fouling in half while still vying for the league lead in blocks and rebounds, or (3) something in between, but equally impressive.

Maybe my imagination is just limited, but I can't really picture that an athlete can be twice as coordinated, strong, and otherwise capable as
DRob or Dream. If Russell or Wilt blocked and rebounded like the best ever, and almost never fouled, it must have been a different league.

Wilt's 50.4 PPG, while playing 48.5 mpg, shooting just over .500 TS% (in 125-123 games) is not as impressive (to me) as several of Jordan's
scoring seasons. Karl Malone had a season that was equivalent or better. Oscar's best seasons are not quite as good as Magic's best, when
scaled to their environments.

And no one has offered evidence that it's as easy to block a given % of shots in a run-n-gun league as it is when the offense sets up plays.
There's less need to, when FG% are around .400.
_________________
40% of all statistics are wrong.

[Carlos]

But if a faster pace means more blocks, shouldn't it lead to more fouls, too?

[deepak_e]

Historically when we look at big men, do fouls correlate better with rebounding or blocks? If its about the same, then I think that could Mike's argument as we already know that Wilt/Russ were dominant rebounders without fouling a lot.

[bchaikin]

Well actually I like to see Bob get people riled up...

doesn't take much for some does it?...

would you consider a counter arguement to Eaton or Bol representing an unachievable rate for Russell?

no, not at all. the above is a conservative guesstimate. they could easily have blocked more. from the few celtics games i have on tape from the early/mid 1960s i've counted russell with 7-8 blocks and more in each, and some of these games are missing parts of quarters or entire quarters alltogether. do i think russell and chamberlain blocked shots at a similar high rate (BS per opp FGA) to eaton and bol's best? no. but could they have - sure...

My understanding is that he revolutionized defense and given the average height in the league (and I assume less athletic players), couldn't his relative shot blocking ability approached the top of all time?

russell averaged 44-45 min/g during the height of his career, and typically played close to every celtics game during a season. imho i don't think there is any question he could have blocked 400-500 shots a year during these regular seasons...

Actually, I like Bob's reasoning as well. But I've already brought up a couple of possible restraints on his extrapolations. We aren't even that far apart (4.0 max vs 5.0 max)

sorry but my 5% BS (4.5 - 5.0 bs/g) above was a conservative minimum guess, with the 6%-7% BS (6+ bs/g) being more of an average guess...

One is the Blk/PF ratio. Even shotblocking specialists max out at 1.60 (Bol, at his youngest and quickest). Five blocks per 2.0 PF (2.50) is unapproached in ensuing decades. For guys who also were world-class rebounders, it's like supposing Olajuwon and Robinson could either (1) double their shotblocking rate without fouling more, or (2) cut their fouling in half while still vying for the league lead in blocks and rebounds, or (3) something in between, but equally impressive.

is there some credence to this bs/pf ratio? such that only shot blockers commit fouls? fyi since BS were first tracked (73-74) there's been over 8 rebs for every 1 BS, consequently i'd bet far more fouls were committed in going for rebounds than in going for blocks. have you checked the reb/pf ratio? because if you do you'll see that chamberlain, russell, and nate thurmand had the very highest one season reb/pf ratios (ratios of 7 to 16), and this was during the 1960s (ben wallace and moses malone had ratios as high as 6 after 1980). if they had - by far - the highest reb/pf ratios, in other words were able to get vastly more rebounds per foul committed versus the players of today, i don't see why they wouldn't have the highest bs/pf ratios...

Maybe my imagination is just limited, but I can't really picture that an athlete can be twice as coordinated, strong, and otherwise capable as DRob or Dream. If Russell or Wilt blocked and rebounded like the best ever, and almost never fouled, it must have been a different league.

you don't have to imagine that russell/chamberlain/thurmond had seasons with much higher reb/pf ratios than olajuwon and drob. its a fact...

And no one has offered evidence that it's as easy to block a given % of shots in a run-n-gun league as it is when the offense sets up plays.

in 90-91 the denver nuggets scored 120 pts/g and gave up 131 pts/g. they averaged 10 more possessions per game (115.5 poss/g) than the next fastest paced team (gsw, 105.5 poss/g), yet that next fastest paced team averaged only 12 more possessions per game than the slowest team in the league that year (det, 93.4 poss/g). but that nuggets team had the 3rd most shots blocked of all 27 teams that season...

[rlee]

I watched the xmas day 1968 national telecast of LAL-Pho & counted 23 blocks by Wilt.

I carried that # around in my head since then hoping to find corroboration & then recently found a 1/27/69 Sports Illustrated story by Frank Deford in which he recounts that in that game Wilt blocked 23 shots!!

[HoopStudies]